November 10, 2025 • 60 mins
Matthew Bertram brings Rob Neumann on to map how private LLMs, PIM/DAM, and AI-driven search turn messy B2B catalogs into findable, persuasive product experiences that convert. We push leaders to move now on the LLM land grab, unify entities, and align teams to become AI first.
• unifying brand, author, and entity signals for clarity
• why every business needs an e-commerce path
• findability before conversion and product page SEO
• persona-based descriptions with private LLMs
• PIM/DAM as the enrichment layer above ERP
• AI search for exact matches and smart substitutes
• channel-aware inventory and safety stock reduction
• SKU and vendor normalization for white label margin
• CAD, specs, and E‑E‑A‑T for trust and rankings
• email fatigue, SMS prompts, and postcard nudges
• urgency of the LLM land grab and authority hardening
Guest Contact Information:
Website: robneumann.com
LinkedIn: linkedin.com/in/robertneumann
More from EWR and Matthew:
Leave us a review wherever you listen: Spotify, Apple Podcasts, or Amazon Podcast
Free SEO Consultation: www.ewrdigital.com/discovery-call
With over 5 million downloads, The Best SEO Podcast has been the go-to show for digital marketers, business owners, and entrepreneurs wanting real-world strategies to grow online.
Now, host Matthew Bertram — creator of LLM Visibility™ and the LLM Visibility Stack™, and Lead Strategist at EWR Digital — takes the conversation beyond traditional SEO into the AI era of discoverability.
Each week, Matthew dives into the tactics, frameworks, and insights that matter most in a world where search engines, large language models, and answer engines are reshaping how people find, trust, and choose businesses. From SEO and AI-driven marketing to executive-level growth strategy, you’ll hear expert interviews, deep-dive discussions, and actionable strategies to help you stay ahead of the curve.
Find more episodes here:
Follow us on:
Facebook: @bestseopodcast
Instagram: @thebestseopodcast
Tiktok: @bestseopodcast
LinkedIn: @bestseopodcast
Connect With Matthew Bertram:
Website: www.matthewbertram.com
Instagram: @matt_bertram_live
LinkedIn: @mattbertramlive
Powered by: ewrdigital.com
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:00):
This is the unknown secrets of internet marketing.
Your insider guide to thestrategies top marketers use to
crush the competition.
Ready to unlock your businessfull potential.
SPEAKER_01 (00:12):
Let's get started.
Howdy, welcome back to anotherfun-filled episode of the
Unknown Secrets of InternetMarketing.
I am your host, Matt Bertram.
I have been going through anidentity crisis based on Entity
SEO.
I'm known as Matthew Bertram.
I'm known as Matt Bertram.
The podcast is known by UnknownSecrets of Internet Marketing,
Best SEO podcast as a variety ofother things.
(00:34):
And I've been going through thelast two weeks some work to try
to unify those entities online.
And we have a common direction.
So stay tuned.
I know you have heard about mechanging the name of the
podcast.
We're going to just stick withuh the best SEO podcast because,
well, we all know that's what itis.
And um as uh the discussioncontinues to go further towards
(00:57):
LLMs and visibility uh in thoseLLMs online.
I wanted to bring somebody onthat somebody true, near and
dear to my heart, because he's amentor of mine.
He's been my a mentor of minefor a lot of years, and he was
doing AI even before it's cool.
It's not any reflection on hisage.
Um, he has been he has uh beendoing machine learning and
(01:22):
structured data for uh a longtime.
And now LM's hit the scene andeverything's cool, and he's
right in the middle of it all.
He's brought me in on a numberof projects.
I am honored to have the one,the one only, the Rob Newman.
Rob, welcome to the show.
SPEAKER_00 (01:36):
Hey, thank you so much for having me, Matt.
And um, it it really is apleasure to be here.
Uh, truthfully, I learn as muchfrom you as hopefully I've been
able to impart to you.
But um, this is a huge topicright now, just huge.
SPEAKER_01 (01:53):
So yeah, and and and to credentialize you a little
bit, uh, you built one of thefirst email automation software
companies back in the day.
You've sold multiple agencies.
You've recently sold your lastagency and you're doing big uh
e-commerce implementations thatare utilizing AI as well as uh a
variety of other things, andyou've worked with some of the
(02:14):
biggest brands in the world.
So um, you know, I'm I'm so gladthat we're able to work again
together on another project.
Thank you for calling meSaturday mornings and uh
starting to uh get get get allthis kind of spun up and and um
I'm glad where it sits now.
And so I just thought for theaudience, uh everybody
(02:36):
listening, um, I talk a lotabout uh B2B marketing and
what's going on with LMs.
And you've really brought LLMsinto the e-commerce component.
I think that there's uh a lotthere.
And um, you know, Pim, Dan, wedid a podcast a long time ago if
people want to go check it outon what that is.
(02:58):
Um, and I've been doing a numberof series around uh Amazon
marketing and uh ad testing andand utilizing AI.
And I thought you could reallykind of round out uh some of
that knowledge on on how we'reutilizing it.
So um one of the biggest thingsthat happened pre-COVID to kind
(03:19):
of tee and uh tee this wholething up and set the table is I
I interviewed um really somebodyfrom the Google Webmaster team,
okay, the spam team, uh, thathas written a lot of the code.
And and they came on and said tome that every business needs to
consider setting up ane-commerce cart on their
(03:42):
website.
Okay, and this is comingdirectly from Google, and this
is pre-COVID.
Okay.
And this was something that youand I have had a lot of
conversations about on how uheverything's coming together to
one point, right?
Like um, an example of this isyou can't have a LinkedIn
(04:03):
profile and uh uh a Facebookprofile if you're graduating
that that are two separatepersonas, right?
You got to unify that persona,you got to bring that together.
If you're doing online marketingum and you're you're selling
online or you're sellingoffline, 75% of that customer
journey happens online, and youwant to reduce as much friction
(04:25):
as you can to get people to buyonline.
So everybody needs to startthinking like an e-commerce
company.
SPEAKER_00 (04:32):
Without a doubt, right?
People's buying habits havecompletely changed.
Let's just think about ourselvesin a consumer experience, right?
I walk into Best Buy, I check iton Amazon.
I just make sure my price isright.
Um, it's a it's a continuousprocess.
But you know, given that this isSEO, I wanted to stick this in,
(04:54):
right?
You're spending money on ads,SEO, a lot of work as a business
to get people to come to yourwebsite.
My job picks up from there.
You've got people on yourwebsite, but are you actually
making money off of them, right?
So conversion is huge.
Shopping cart conversion ishuge.
(05:15):
But I like to say, look, ifyou're not finding, you're not
selling.
And a lot of what we use AI foris to make sure that the
products, the services, thethings that you want to sell
actually appear for your clientswhen they're searching for
something and they wantsomething.
Because in B2B, if they can'tfind it, in their mind you don't
(05:38):
have it.
Whether you do or not, they areonto your competitor and moving
right along.
And that's a huge benefit um uhfinancially.
SPEAKER_01 (05:48):
Well, I I can tell you like product page SEO is
incredibly important.
A lot of people are just usingthe manufacturer, depending on
what you're selling, right?
Like we can talk about a lot ofdifferent use cases, but you
know, I look for okay, I don'tknow, whatever it is.
If I'm doing this, I want it toactually say that it does that
(06:09):
in the product description tomake sure I'm buying the right
thing.
And if it doesn't have it, eventhough I think it does, and
something else says it and it'sa little bit more, even though
it might even be like very, verysimilar.
I I'm gonna err on the sign oflike, okay, this says this
component or whatever it is,fits with that.
And and and that falls in rightwith what you're saying of well,
(06:31):
if you're doing marketing, um,you got to be able to
differentiate, and that productneeds when people are searching
for it to rise to the top, to beone of the choices in the
selection process.
You also open up the can ofworms of you know, UX and the
customer journey on page, whichis probably arguably the most
(06:52):
important thing that needs tohappen, which is where SEO,
right?
You said SEO drops off, CRO kindof picks up, or there's a big
overlap there.
But when clients are typicallylooking for SEO or CRO, they're
really saying the same thing.
I want, I want a customer, Iwant you to put a customer
together with a bow on it, and Iwant you to deliver that to me.
(07:14):
And I don't care what multimodalyou're doing, like give me new
customers is essentially whatI'm asking for from uh most
people that are looking forthese services.
SPEAKER_00 (07:26):
Without a doubt, and and a point you just made is
really important, right?
You said I want it to say what Iwant it to do, or you know, it
describes exactly what I want todo.
And there are a few points thatare really important to that.
The first is in search, givingonly one perfect answer is not
(07:49):
as good as giving a few answers.
One, exactly what they'relooking for, but two or three
that are similar, right?
Or that are bought with, right?
Amazon is great at this.
Amazon is a great at a lot ofstuff, but you it actually has a
(08:09):
higher satisfaction score togive them one perfect and then
three options, and I like themstacked on the side that are one
is cheaper, one is you know, adifferent brand, one is
something that always getsbought with it.
Those things actually sparkinvestigation where people will
(08:31):
buy will one at a better ratebuy something, and two, it
raises your shopping cart value.
SPEAKER_01 (08:40):
Yeah, I mean the cognitive load is is really
something that people um try tocram too much information in
there.
Even Amazon, again, Amazon'sgonna test it, they'll figure it
out.
They just move to like, here'sall the choices of everything
you bought, you know.
And so if I'm buying a piece oftechnology, it's like, do you
want bananas too?
And I'm like, like it's toomuch, right?
(09:00):
And and I I just click next andI actually don't like the
current experience.
Um, they'll figure it outthrough testing.
But to your point, when you giveme three recommendations that
are tied into what I'm buying,and this is bought with that,
and it's something I canconsider, I usually will click
on that.
I'll usually look at that, andand it's not too much to
(09:22):
process, right?
And um on the website, um a lotof what we're working on right
now on on a separate project isokay, what is that customer
journey?
What are the steps that they'regoing through to help make a
buying decision?
And and what happens on mostsites and why the bounce rate's
so high is you have thesemassive menus, okay, that give
(09:47):
it so many options that youdon't know where you're taking
them.
Like if you're trying to getthem to buy something, all these
other things are just takingthem away from that.
And many times they're gonna getdistracted, they're gonna leave
the website, they're never gonnacome back, and that's why you
have a horrible bounce rate.
SPEAKER_00 (10:04):
Well, you were right, and then let me come back
to because I know you wanna youwant to talk about some of the
technical and AI things we do,right?
So, one of the things you alsosaid is does it say it?
Does it say yes?
Well, that is dang important.
You're right, and we have sometools that we utilize to make
(10:24):
sure it gets said, right?
And it's and it's more it andB2B is catching up, but it's not
there yet, right?
It isn't, but it has to be.
The expectation of people isthat there's no difference
between if I want to buysomething personally or for
(10:46):
business.
Now, one of the ways that we dothat is we use AI and some
private LLMs that are industryspecific, right, to enrich the
product information, right?
And we actually can do that intwo ways, right?
Um, one is just what are thedescriptions?
(11:08):
What are the things that tietogether?
What are the matching products?
What are those descriptions?
What material and partner itgets large amounts of data, but
it has to say what you'relooking for.
The second thing we use it foris voice.
With AI today, in for example,one of the products we might use
(11:29):
is called a product informationmanagement system, or B2B,
right?
Is we can not only take a sliverof product description and make
it useful to marketing and SEO,but we can also change the
voice.
I can take that same thing andtell it to describe this to a
(11:53):
five-year-old, describe this toum uh a buying agent, describe
this to an engineer, anddescribe this to call it the
general public, which might begood for SEO or whatever, or a
fifth one, right?
Generating specific keywords.
We can we can use thepersona-driven side and the
(12:17):
description side to create someincredibly rich uh data.
SPEAKER_01 (12:23):
So to to kind of unpack this a little bit more,
uh, a lot of uh clients thatthat listen or listeners are are
B2B companies, and I do a lot ofthings in B2B selling.
Uh ERP system, right?
So uh ERP system is is whateverybody's familiar with on
maybe inventory management,product management, whatever.
(12:44):
PIM sits on top of that and andit has kind of some machine
learning components, which wereused before LLMs with uh like
you you had to clean the dataup, you had to label the data,
and and then um you could usethese LLMs to to to get multiple
product descriptions, to uhextend the the uh the the
(13:08):
category or the definitionaround it, and it could speak to
more types of people.
Um and so uh, you know, if youjust have the the the ERP
system, you you're gonna getreally basic data.
You're gonna get just reallybasic labels.
And you you wanna you wannaexpand that reach and you wanna
(13:30):
expand uh understanding of howpeople are searching and what
they're lurking for, and you addthat AI or intelligence layer to
that, and and now there's morechances to get a sale because
you're answering the questionbetter, and it also is
understanding more about whatyou're looking for and who it's
trying to reach, and it givesyou a better chance to surface
(13:52):
um the right products for them.
SPEAKER_00 (13:55):
Absolutely, and you know, what I do never sounds as
good as the stuff you do, Matt.
I mean, you get all passionate,and I'm sitting here in the
background doing all the datawork and trying to put this
stuff together, but you'reexactly right.
You know, we've been doing thisfor a few years, several number
of years, right?
(14:16):
Where it started out as machinelearning, the the grandfather of
AI, right?
But being able to going through,and we started with cleansing
data, ordering data, creatingmore data and some structures,
it has now moved into exactlywhat you've described, right?
We have full-on AI models, andwe use both public models,
(14:39):
right?
So that it's understandable,they're they're well used, but
we also have created privatemodels, and they're important
for a couple of reasons.
One, a number of B2B companieshave um, you know, data they'd
rather not not have outpublicly, right?
We have we have a uh a companyright now in common, right,
(15:01):
that's in pharmaceuticalmanufacturing, and they have you
know secrets that they don'twant out there.
So we've created a LLMspecifically around their
chemistry, around the thingsthat they do, and we can search
it and serve up informationwithout revealing the secrets,
(15:22):
yet be able to pull in all ofthe information you're talking
about and do all the things thatwe do to give someone the right
answer.
Because that's actually, youknow what's great is it's just
like today.
How many people, you know, Iactually just want an answer
from Google, I don't want fivewebsites to go look at.
(15:43):
And what has happened?
And you talk about this, Matt,right?
I want an answer.
Just give me the freaking ah,that's the answer I'm looking
for.
Beautiful.
Move on.
I don't need to go visit yourwebsite.
SPEAKER_01 (15:57):
And and so, how for everyone listening, how what Rob
and I do kind of matches is I'mfocused on the public-facing
view of surfacing in thosepublic-facing LLMs.
Rob's maybe taken some of thosepublic surfacing or
public-facing LLMs, been able toput it on a separate server, MPC
server, a data lake, building adata mark, labeling the data,
(16:20):
and then you can recall thatinformation that's private, and
then it serves up inside aportal on the back end.
So it's taking um thefront-facing marketing piece and
now operationalizing it from ane-commerce piece.
And then there's an even greaterpiece of an expansion of the
business leverage or the AIapplied leverage that you can
(16:43):
get to apply it to other piecesin the business.
We're just talking aboutspecifically um the the product
catalog or um you know uh uhinformational search of a
proprietary library orproprietary database.
And it's fine-tuned from a youknow a chat bot standpoint to to
give you the answer.
(17:03):
And and again, that nextexpansion for the like VPs of
innovation that are listening,um, you can apply that to all
your data governance um to beable to recall data, to get
trends and and how this uhaffects what we're doing uh on
this shared project is you cansee the trends of what people
(17:24):
are buying, how often, the timethey need to refill and buy
more, and you can suggest thesethings to them to to help them
stay on top and to pull uhforward maybe some of that
demand as well as to just helpmake their life uh easier to
keep stuff in stock, whichbecomes really, really helpful
from a projection standpoint,from a uh data usage standpoint,
(17:48):
from a customer standpoint.
Um, these things are kind ofkeeping you out of the ditch.
So there's there's all theseadvantages right around um
purchasing orders and uh to toB2B companies that are listening
uh to be able to implement thison a supply chain side.
That's what we're getting a lotof uh interest in, Rob, is a lot
(18:08):
of supply chain logisticcompanies are are are seeing the
use for like uh AI layer to helpproject um a lot of these um uh
coordination uh and and deliveryschedules and uh uh fallout
rates in say oil and gas.
And so um it's it's justincredible what it what it's
(18:29):
able to do.
SPEAKER_00 (18:31):
It is, but I'm gonna back it down a tier because B2B
is behind, right?
So, you know, what if you're amedium-sized company and you're
selling through multiplechannels, you're selling to some
distributors, you may be sellingthrough Amazon or or whatever,
right?
Your Shopify store a lot ofchannels, TikTok store and your
(18:53):
Shopify store and and and andand right, yeah.
One of the key problems to solvefor these companies is
inventory.
Yes.
How do I know how much inventoryI need to sell on Amazon versus
my Shopify?
What if I have more than onewarehouse?
So, what they wind up doing isactually a very old-fashioned
(19:13):
solution.
They put in safety stock, whichis extremely important.
And if I allocate safety stockincorrectly to one channel and
the other channel sells out,I've got a bunch of stocks
stranded over here that I couldhave sold, right?
So using AI and using you knowgood connectivity, good data, it
(19:37):
can actually manage all thechannels at one time.
And we do that in in PIM.
And the reason we do that iswe're able to work directly with
the ERP, apply some AI, look atthe channels, manage and even be
predictive in the selling.
Now, Amazon can be helpful withthat, but I know a lot of
(19:57):
customers that are doing safetystock across all of their
channels, and you don't need to,and it's expensive.
SPEAKER_01 (20:04):
Well, one of the things that you and I were
talking about the before thiscall, uh, to to kind of build on
that concept was, you know, ifyou're a supplier, right?
If you're a supplier, you'retypically, you know, you're
you're either repackaging orreselling other people's stuff,
or if you're buying somethingand you're configuring something
(20:25):
or whatever, um, that ERPsystem's not going to do a
really good job of recognizingall these different product
orders and POs and um you knowproduct labels.
And if you're trying to resellthat as your own product, right?
You're you're you're you'reyou're you have a markup or uh
you're you're on our arbitragingwhat's going on and you're uh
(20:48):
buying that thing from a bunchof different sources, okay?
They're gonna have their own POnumbers and product label
numbers.
Uh, and you got to be able tounify that to your point on the
product supply side, uh, andthen present your own product
label number.
You can't do that really well.
It's a very manual process withthe ERP system.
(21:09):
And a lot of people are banddating and gerry rigging that
up.
And I just wanted to flesh thatout a little bit more because
that's where the the the PIM andthe dam come in is you can run
all those APIs uh into that AIlayer, it can figure it out and
tell you um what's at where,what store, what it is, and it
(21:30):
and it streamlines that productso you can have just in time um
delivery where you're notcarrying all that overhead.
Um, and I think that that'ssuper powerful because I know
that there's people listeningright now that they're using
those ERPs and they're like,there's got to be a better way
to do this, and they're tryingto, you know, uh band-aid up a
(21:54):
bunch of different solutions.
SPEAKER_00 (21:55):
Sure.
Well, ERP was, of course,primarily set up as a financial
system, right?
And it's really good at it.
It will, it will, it will handleyour finances, it's designed to
plan your resources, but in theend, right?
Um, and you threw a lot ofinformation out there, Matt.
But in the end, right, what arewe trying to do?
(22:17):
We're trying to sell product,and that may come in many
flavors.
It could come from multiplevendors.
You may even have a white labelproduct that you do of your own.
You may have um the opportunityif you're a distributor, you
probably sell multiple brandsthat do the same thing.
Relating all of that is hard.
(22:40):
It's a lot of data.
PIM can make that so mucheasier.
You can sit there and do it byhand in an ERP and start
matching stuff up, or you caningest it, connect it, and use
AI to begin to put all thesethings together and then deliver
it to your AI-driven search.
It becomes a whole lot easier.
(23:02):
And it gives you a lot offeatures that people have not
been able to do before.
I'll give you uh an example fromone uh plumbing client.
We were able to take all of thatinformation, create all the
relationships between Delta,Kohler, whatever the brands
were, their own white labelbrand and off brand, and a lot
(23:23):
of data.
And in the end, we were able tosurface one exactly what the
customers, if you put in KohlerValve 1, 2, 3, you will get a
Kohler Valve 1, 2, 3.
But you will also get theirwhite label exact product that
matches it, which they make moremoney off of.
(23:44):
It's a cheaper product, butwe're maximizing their profit.
And then we'll give you someother options.
This data is truly valuable.
SPEAKER_01 (23:53):
Okay, so I want to break one thing apart and and
dig in a little bit more of whatyou just said.
So, okay, so we have like ourwebsite and and you know,
there's an ecosystem associatedwith that.
We have it connected to an ERPsystem.
Now we're talking about a a PIMor a DAM or a hybrid of that on
(24:15):
pulling in data and API.
Um, is what you're talking aboutnow adding an AI search layer on
top of that, or is that stillpart of PIM and Dan?
And like, what are thedifferences between plugging
something like that in to giveyou even more granular search to
give people the absolute rightanswer of what they're looking
for?
Because that's some morehorsepower.
SPEAKER_00 (24:37):
It is, and it actually requires both.
So okay, I just wanted to makesure we get to we may get to one
as we begin to be able to usemore and more unstructured data
to do these things, but we haveto just like any LLM, right?
We have to have the descriptorsin there for the AI to find
what's common, right?
(24:58):
If all I do is put in screw andnail, right, they're not going
to be able to connect it ashardware.
So uh, you know, it it doesn't,it doesn't always work that way.
Now, what we do is we take allof that data, we do some
structure inside of the PIM, anda PIM is the product, the dam is
(25:20):
media, digital assets.
That's where that comes from.
And what we're able to do isenhance that data inside of the
PIM, give it all the terms thatthe AI will need to search and
look for.
And then when we serve that up,the AI-driven search is able to
just go to town.
(25:41):
I've got your materials, yoursize, your product, your
everything that goes into thathas been enriched to allow the
full search capabilities acrosseverything.
So it's two steps right now.
SPEAKER_01 (25:56):
Yeah.
So what what's just I there's alot of people out there, right?
And I I just the term uhinformation's the new oil,
right?
Has been talked about for a longtime.
And you had to do you had tohave a a data uh analyst team,
you had to spend months like toslow down like the data
(26:17):
labeling.
Like you're building, likethere's a ton of people
listening that have data marksand they have tons of data and
they haven't done anything withit.
I would love to hear from youwhat are some different use
cases if they start to implementPIM, DAM, some AI-driven search
on on to let's just open up theminds of what are some of the
(26:39):
things that they could do withthis data.
I'd love you to just rattle offa few uh high value use cases
where you could make more moneydoing these sort of things.
You can leverage it.
SPEAKER_00 (26:48):
Yeah, when you've got data buried across wherever
it is, right?
Being able to relate data andserve data are the two
beginnings of better revenue andtotal cost of ownership, right?
And the way you can do that isby beginning to tie that
(27:09):
information to what's actuallyhappening on the website.
You can serve it and you know,and and actually test out what
appetite is there for thisproduct data and information and
products.
You can use it to make yourproducts more enticing, more
descriptive, provide moreinformation to different
(27:33):
personas and buyer groups.
And then let's not forget,right, once we've invested in
all this data, long tailmarketing and long tail find is
really important, right?
Why does your website care ifyou know um so many clients,
right?
They have the top 20% of theirvolume on the website, not 100%.
(27:54):
80% is buried in catalogs or asalesperson's mind.
So what happens?
Salesperson retires, and a lotof guys are beginning to retire
in in these manufacturing anddistribution businesses, and
you've lost it.
The e-commerce and search doesnot care.
Put all million up, and the onetime that somebody buys it, you
(28:19):
are way, way ahead.
It also helps in your search.
So, in the end, just from a fewideas, right?
You're getting more people toyour website, you're getting
more sales off your website, andthen you're also going to be
able to create more data tounderstand your customer better.
What's the customer journey onthese kinds of things?
Did we miss something thatsuddenly turns out to be
(28:41):
extremely important right now inengineering or in uh in a
buyer's group?
The only way you can find outthis data is let's expose it and
run some reports.
Let's find out.
There's so many opportunitiesthat um that you can do with it.
And I'll I'll give you a reallife example, right?
(29:02):
So we had a tubing company,right?
Very, very boring industry,right?
They might say.
And yet it's critical, right?
They sell a lot of tubes, theythey're all different sizes,
right?
And what they realized is therewas a that by just putting up a
few things they don't normally.
(29:24):
One, their CAD drawings.
CAD drawings are typically filedaway somewhere.
Putting up their CAD drawings isa true blueprint of what it's
going on, the real information.
And then secondly, using AI tomake their detailed product um
(29:46):
PDF sheets just readable.
Traffic to the website went uptwelve percent.
Twelve.
Just making all this exposed.
Right, went up 12%, and thebuying went up three percent,
three percent pure revenue, moreclients, more revenue.
(30:10):
Nicely just exposing the data,which the IT does not care, it
does not care how many things Istuff in it, it doesn't care how
many calls you make from it.
It was you know a one-timeeffort, create a process, and
now I'm making more money.
SPEAKER_01 (30:28):
You know, this is the hard thing because AI can be
applied to almost every aspectof the business to help improve
it in some way.
If we drill down to well, AIlearnings, and we're talking
about like CRO conversion rateoptimization, like understanding
that customer journey becausethe buyer behaviors changed and
(30:50):
people want more of that proof.
You you started, I startedthinking about eat expertise,
authority, trust, andexperience.
That's what you were doing isyou were proving that out more.
So people said, Oh, this is whatI'm gonna get.
This is how it's gonna do.
It's explained.
The LLMs understand it, thepeople understand it.
It's a win-win because thatprocess is happening.
(31:11):
A lot of people are like, I'mgonna save all that until I get
them on the phone, but nowthey're missing the deal because
they're not even in the topthree choices or whatever the
selection is because they'realready out of the running
because someone's deselected andunraised their hand from using
them.
And so, I mean, you look atthese bounce rates, you look at
(31:34):
these click-through rates,they're minuscule.
You you gotta get that numberup.
And when they get to yourwebsite, you got to understand
what content are they missing?
Where did they get stuck?
Do they get stuck cardabandonment, right?
That's a huge, huge issue.
Why do they put it in there?
What's going on with that?
I would love to hear a littlebit more about how you look at
conversion rate optimization,heat mapping, um, and and how
(31:58):
you're starting to make some ofuh those decisions because yeah,
SEO gets them there.
We have there there's a a aconversation that has to happen,
and and really, you know, 50% ofthat informational content is no
longer on the website.
So you got to drill down on themiddle of the funnel and bottom
of the funnel content.
Because if someone's came toyour website, right, they've
(32:20):
gone past the AI overviews orthe LLMs have recommended them
or whatever it is, now they'reon your website.
If you don't get some kind ofinformation, and we got, you
know, we got all the uh pixelsgoing away, right?
And so um you got to capturetheir information and you got to
get them to the next step, andand you got to push them through
that funnel.
(32:40):
And the only way you're gonna beable to do that is with data,
and that's the bigdifferentiator today that you
can you can still analyze isonce they come to your website,
like you know, they're they'rein your territory, right?
SPEAKER_00 (32:53):
Absolutely.
Um, and it it can be a littlemore difficult in the B2B space.
So there are there are a fewthings that that we do, right?
One, don't just provide yourfellow your telephone number.
That's the way to interact withyou.
It sounds funny, but there aretons of clients we start with
where call my salespeople andwe'll get going.
(33:15):
Well, most people don't want totalk to your salesperson until
they're 100% ready to buy.
Yes, not a good strategy.
So, what what do we do?
First of all, we want to be ableto utilize some methods, right?
That in and and they're consumermethods, right?
We we learned them and we knowthese from websites we interact
(33:39):
with all the time.
How would you like to why don'tyou just save a few favorite
parts so that you can come backand look at it later?
No problem.
Give me your email address andI'll save them for you, right?
These are standard things, butyou know what?
Why don't you join our loyaltyprogram and I'll give you points
toward everything you buy?
(33:59):
And if you give me my your emailand phone number right now, I'll
stuff you with uh a hundredpoints just for signing up,
right?
Um, and as you go, so that's onthe very early side.
SPEAKER_01 (34:12):
Quizzes and calculators.
I love quizzes and calculatorsand engaging them to uh but you
but we got to have your email tobe able to give you the answer,
right?
SPEAKER_00 (34:21):
Uh you do, right?
So um, you know, what's youraddress?
I'll calculate, I'll calculateout your your shipping address.
Do you want you know, we usegated content, right?
Yeah, I have really valuablecontent.
Just give me your email address.
Everything you can do to getsomething from the client is
really important, but not everyclient does.
(34:42):
So what happens?
Now, let's say you do have ashopping cart, you know that it
is the most effective way forreordering.
You know that it's the you know,the way sometimes people want to
try out a small buy before theygo call your uh your salespeople
and give you a big purchaseorder.
(35:02):
They just want to order one, nota thousand or a million, right?
Just one.
Let me buy one, let me see whatit looks like.
That happens all the time.
So, what happens?
You get into the uh shoppingcart and then they stop.
It's not high on their prioritylist, it's very high on your
priority list, right?
(35:24):
So, what do you try to do?
First of all, right, we'retrying to get their information
up front.
Hey, create a shopping cart.
Okay, to create a shopping cart,let's create a quick account.
Where am I sending it?
What's your name?
What's the company, right?
It and we create all that.
You've got it sitting here inyour shopping cart, and now I've
(35:45):
stopped.
Well, what here's two thingsthat that surprise me now.
Yeah, you send emails, of courseyou do.
Hey, you stopped something inthe shopping cart.
What I was very surprised by iswe encountered a marketing
company and they convert thatand they send automatic
(36:06):
postcards.
I love that.
SPEAKER_01 (36:09):
I was like, just direct shipment, yeah.
SPEAKER_00 (36:11):
Just a direct here's a postcard, and by the way, I'll
give you 10, 5, 20.
They actually look at what's inthe cart.
Can I ship them a 10 coupon?
SPEAKER_01 (36:23):
And that's that's so that the the drop shipping to
set something like that up.
I actually had a conversationwith a client about that last
week.
Is let's just put a tag in theCRM if you want to drop ship a
postcard to them that lookshandwritten, and you can have a
couple different options.
You could even put, you know, uhlike a a tier up from that, send
(36:45):
them cookies, whatever.
And and the conversion rate'sgonna be so much higher because
you're building reciprocity andyou're hitting them offline
because there's so much noiseonline, people get distracted,
you know.
SPEAKER_00 (36:56):
And oftentimes, well, sometimes they don't use
the right email address.
I don't know why, but sometimesthey don't, right?
So that's true.
There are methods around it,particularly on like gated
content.
If you're selling information asyour product, right?
You can you can insist on a goodemail because that's how you're
(37:19):
sending access to theinformation.
But if it's a low value, youknow, sign up to get something,
you know, often you know, we'llsee you know, junk12345 at
gmail.com uh as the thing, andtherefore is that really
working?
Where in the shopping cart, ifthey want it, they will give you
(37:39):
the right address.
Absolutely.
SPEAKER_01 (37:42):
Yeah, I I think I think adding the text automation
to that, like I mean, I'mstarting to get a ton of spam
text, but I still need to deletethem and clean out my text
messages.
Absolutely, but text it has awhatever 98.9 or 99.8 percent
open rate still.
Uh email, it people are justdeaf by email at this point.
(38:03):
Like, we're having to set up Robwith a lot of our clients, Slack
chat because they're not gettingour emails, um, because they're
overloaded with emails, they'reworking on whatever they want to
work on.
So here's a specific form ofcommunication to talk to.
So, like, I mean, emails gettingtougher and tougher.
I mean, now you can use AI to,of course, clean out your emails
(38:25):
and you know, sort by thedifferent categories you're
using it for, but email email'stough these days.
Uh, it's not always the mostreliable, but if they get it in
the mail, right, they're gonnapick it up and see it.
So, I I love that example.
That's a great example.
SPEAKER_00 (38:40):
Yeah, everyone looks at their mail, even the junk
mail.
You still look at it.
Is it junk?
SPEAKER_01 (38:47):
Yeah, you gotta, you gotta, you gotta define it, you
gotta categorize it.
SPEAKER_00 (38:49):
Yeah, you gotta define what is it, right?
Um, I I think your point aboutemail is is spot on, right?
And that is it's so full, it'sso completely um uh uh overused.
I mean, I watch my my thenumber, the red number grow on
(39:10):
my email.
And then sometimes I just gothrough okay, anything older
than this, gone, right?
So it it's tough to get throughthe clutter.
SPEAKER_01 (39:23):
So let's let's bring this back and and kind of bring
it back to to marketing andclose it out.
Uh I think a lot of people arehearing, okay, AI is gonna
affect business, right?
It's gonna have a positiveimpact.
Where can I see that impact?
How can I apply that?
I think we went through a numberof steps on how to do that.
I think that Pim and Dam, if youhave an ERP system, you should
(39:45):
be really critically looking atthat.
How can I get uh uh AI to applyto different categories?
And that could branch out in alot of different ways.
CRO and SEO are connected, okay?
I think they're attached at theat the hip.
A lot of times when people wantone service or another, they
really want the same thingbecause they want to just open
up that funnel and and tightenthat funnel and bring that
(40:07):
together and unify that data.
And so, you know, if you'reusing LLMs to make the website
better for a public facing,right?
So you got you got behind thebehind the wall and they set up
an account kind of on that backend when they log in.
But but a lot of the uh look andfeel is that first-time cut
(40:29):
customer interaction, gettingthem into the right customer
journey in the right channel.
Um, you got to have thatinformation publicly facing the
LM's got to be able to read it.
You got to have a uh a site map,you got to have a uh organized
and logical categorization ofyour your product category.
You got to have gooddescriptions to speak to
different people.
One of the things we didn't talkabout is um, you know, I like to
(40:52):
use like this profile because Idid it many times.
Um, but these different types ofpeople want that information,
like you said, given to them,right?
Or some people might want to digmore and get that research, or
some people might want to uhvideo uh or or audio or voice
hear it, right?
They I just I want it just spoonfed to me, right?
Some people want to submit uh uhuh email and have you contact
(41:16):
them.
Some people want to sign up fora calendar, some people want to
download a form and talk tosomebody, somebody people want
to call a 1-800 number.
And if you're not addressing allthese different modalities, you
are missing customers, you aremissing uh uh distribution
channels on marketing.
(41:37):
Um, you need to be able tosplinter this information and be
able to push it out multipleways to bring everything back to
your main source hub.
And you know, something that weprobably don't have time to go
into too much, but hopefully itteases it up is uh we need to
understand your company as anentity.
(41:59):
We need to define that.
We need to define your keyplayers of who the authors are.
All these things are theinformation architecture, which
is really what you specialize inand kind of how we work together
is is mapping that out andmatting mapping the site
architecture out.
And that's how how you and I areworking together really well uh
(42:22):
on some of these projects,because you need different
voices at the table, differentexperts giving different
opinions of looking at itdifferently from design, from
the back end, uh, from themarketing, from you know how the
LLMs are picking it up.
And it's all well, marketing andpositioning and um well,
(42:46):
information architecture.
That's what Google's trying todo, uh, is understand that's
what all these LLMs are tryingto do.
They're trying to figure out theinformation, trying to organize
it and layer in a way that youcan recall that information uh
really easily.
We're talking about likeinformation recall, is is
really, I think, the the the bigor data recall.
There's a whole industry aroundthat.
(43:07):
And um, that's what we'retalking about, whether it's on
the website or before thewebsite, on the website, or in
the website.
And and I think that um thatgets underserved sometime.
And and these are complex issuesthat that you need to unify a
lot of data.
And I I think just the executionof companies is they've they've
(43:32):
spent a bunch of money on on oneor two different things, and it
and they just are 85% of the waythere.
And they just need someone, Rob,like you to like help connect
that data together.
And then the circuit startsrunning, and it's really, really
quite powerful.
Um, or or someone like you, youknow, and and I I just think
that uh this is where things aregetting missed.
(43:54):
Like they've they've built thedata lakes, they've labeled a
lot of the data, and it's justsitting there, and they're not
getting that the the insightsthey're looking for.
They haven't gone the rest ofthe way.
And that last 10 or 15 percentis the hardest way to do it.
And once you do that, that thatcircuit gets completed, and and
(44:14):
now you get all these benefitsof the you know, uh the the AI
machine learning, LLMs, whateveryou want to call it.
SPEAKER_00 (44:21):
I agree, and the and the thing is look, we've talked
about removing silos forever,right?
SPEAKER_01 (44:27):
Oh my gosh, yeah.
SPEAKER_00 (44:28):
But leaving your information internal and not
utilizing it, right, is anothersilo.
We have to break down the dataof the company to the consumer.
One of the things you talk aboutquite a bit is you know putting
a stake in the future marketing,right?
(44:48):
The LLMs will rule all search,they will, they will be giving
an answer, you know, and and somany companies have developed
out so much intellectualproperty and they have so much
that just isn't displayed.
Part of this is stick the flagin the ground for the future,
(45:08):
because what we're doing rightnow lays the groundwork for
years of future success becausethe LLMs are looking now,
they're looking for, and youknow this better than I do, but
they're looking for authority,they're looking for depth of
content, right?
They're that we're no longerlooking at high-level marketing
(45:29):
words.
They want you to demonstratedeep knowledge and being able to
expose some of the things Italked about, whether it's an
engineering drawing to um to PDFinformation and getting it in
such a way that you're properlystaking your flag and saying,
look, we know tubing better thananyone else.
(45:52):
It's important, right?
You you think it's not, and thatthere's lots of competition.
But the people who stake who getthere first and show the depth
of competency have a huge, hugeadvantage.
SPEAKER_01 (46:07):
That is the unknown secret that I don't think people
are aware of, that the trainingdata in these LLMs, they've just
opened up like the World WideWeb from their kind of training
data that they're built on.
And the LLMs are still trying todecide what gets uh saved in
their long-term data in thetraining data sets that they
(46:28):
continue to reference.
And the more and more that youget recalled and referenced, the
harder and harder it is gonna beto dislodge you as that
authority or that expert in theLOM data set because every
recall does that with Google.
You know, there's a little bitmore play.
These LLMs are gonna solidifyand in 18 months, it is gonna be
(46:52):
you know extremely hard to dothat.
And the people that are plantingthat flag in the ground now and
and building those libraries andthose data sets and becoming the
the go-to recall expert forthese things are gonna win.
And people are gonna be spendinga lot more in the future uh to
(47:13):
just get a piece of that becausethe this this is this is uh the
time that people need to likeget with it and and take action.
Um, I just did a uh article onthis of a kind of like the LM
land grab, and and I think it isabsolutely so true.
(47:34):
Um, and in some of thepresentations, you know, I mean,
just ask ask any LLM, they'lltell you the same thing.
Like this is where we're attoday, and this is a monumental
shift on how business is doneand where it's going.
And I'm seeing the jump rates ofof usage to to your point.
(47:55):
Uh yeah uh Google was thelibrarian, okay.
And you would go to Google andthey would give you 10 blue
links.
Here's where you go to goresearch it.
Now people just want thatanswer.
Give me that answer, and you godo the research, synthesize the
data, and give it to me.
As people start to use that,they don't go backwards, okay?
(48:16):
They they only move forward, andwe're seeing you know, huge
amounts of increase in this.
And as people get more and morecomfortable with this, they're
they're using it more and moreoften.
And and and this is the future,it is not a fad.
Um, and and uh any business thatwants to be forward thinking
(48:38):
needs to needs needs to startjumping on this.
And one of the things you'retalking about is that push to
becoming an AI first company,which there's a couple steps to
make that jump.
Amazon didn't do it overnight.
There's a sequence of steps thatyou have to make.
And I feel like a lot of people.
SPEAKER_00 (48:54):
Oh, I like that.
I I gotta trademark that, Matt.
AI first.
An AI first company.
I love it.
But I but you're right.
And here's the thing thatunfortunately people will think,
right?
Oh, AI makes it so easy, sofast.
15 months from now, I can stillget on the in on that 18-month
(49:16):
land grab.
Unfortunately, the answer is no.
Uh, you cannot wait.
The amount of time, right?
Part of it is the amount of timeyou're sitting in the next 18
months.
18 months is a flash, it will gofast.
And as you're thinking about2026, it's better to start
(49:38):
earlier than later.
If you're going, if you're goingto be serious about owning this
space, you need to start rightaway because it takes time to
set it up properly, it takestime for it to be ingested
properly, and it will take timefor these LLMs to decide, right,
that your data is what theyreally want to uh quote in the
(50:02):
future.
SPEAKER_01 (50:03):
And you got to get your IT team on board.
You need to get you you got toget IT and security are gonna
be, you know, uh, they they needto be rowing in the same
direction because uh the biggeryour company is, uh, the more
time it's gonna be taking to geteverybody on board with this, to
make those changes, to get thoseapprovals.
(50:24):
And what you're talking about isyou need to be showing up now,
okay?
Because every time that somebodyelse shows up, they're getting
their moats getting bigger andbigger and bigger and bigger.
That's right.
And if you're at a companythat's gonna take a quarter or
two to get this implemented orapproved, you got to start
pushing on this now.
(50:45):
So so you even have a shot atthis.
But the nimble companies wherewhere everybody's on board and
is roan in that same directionand understands this and has
this education, that's why I'mgetting called in for a lot of
talks, is to to to get everybodyon board with this, to
understand what's happening, toknow that we have to take action
(51:06):
now.
Like the next couple of quartersare really, really critical.
Um, and I think that the gapsare gonna really widen between
the people that get it and doit.
I I have on the opposite end ofthe spectrum, Rob, like one
client that it's kind of pushinguphill.
Um, this was about six to ninemonths ago.
(51:27):
They uh are owned by anothercompany that's owned by another
company, so they've got allthese approvals going up.
They got purchased.
And their stance is we don'tunderstand AI.
We want and and the contentcreation and and SEO, these
these are the marketing salesnow.
HR is about to be affected.
Like these are the things thathad this disruption first.
(51:50):
They said no content, like wewant a hundred percent zero AI
content, okay.
And they were pushing backcontent we were sending them
when we actually had a subjectmatter expert writing it.
And where do you think theseLLMs, Rob, were trained?
Where do you think theinformation they were trained on
actual humans writing content?
(52:11):
So it's very hard to get to ahundred percent.
Like, we were actually having totwist and change the content to
get that to submit it.
That's where uh uh education umat the executive level needs to
happen because this needs tohave champions all the way
through.
You got VPs of innovation, yougot CROs, but like everybody's
(52:32):
got to be on board with this andunderstand it because a lot of
people are afraid of it and arepushing back or are worried
about their jobs.
I have told we've implemented AIheavily in our company over the
last two and a half years.
And we just gave people morereach, we gave better
productivity, we we gave peoplethe the access to do more
(52:55):
things.
It actually wasn't limiting umand and like, hey, we need to
get rid of these jobs.
It was actually how we can domore, deliver more, deliver
better quality work.
Um, but it's all about thatperception.
And if you're thinking AI isgonna replace you, it is.
If you're at a company andyou're worried about that, you
(53:15):
need to understand how they arebecause somebody that that uh AI
is not gonna replace you yet.
These agents have a little bitmore time before you turn out
full agentic AIs, and we'veheard some horror stories about
that.
But somebody leveraging AI andunderstands how to use these
systems will absolutely replaceyou if you're not understanding
it.
SPEAKER_00 (53:35):
And that's the truth, no doubt about that.
And I would I would take it upone level, Matt.
This is a CEO issue, this isstrategy, this is vision, this
is the future of the company forthe next five to 10 years,
right?
IT and marketing alone are notgoing to drive these decisions.
(53:57):
You and I had it, we were at aat a customer and we were
talking to him, and I wassitting next to the CEO, and he
sat there and he said, We haveto do this, right?
He got it.
He he he just was like, I don'thave a choice.
If I want to be there in thefuture, I have to do this.
(54:18):
And the good news for them,right, is they have they have
information, it has to be putout there, we have to make the
land grab, but they're actuallyin decent shape.
But the CEO totally got it, andsuddenly, guess what?
Oh, the CFO, the CMO, the the CCRO and the CTO were all like,
(54:41):
yeah, we got to do this.
SPEAKER_01 (54:43):
Well, that was that's rare though, that you get
all of them in a room and youget everybody on the same page
and they understand it.
And and that's where this uhexecutive level advisory really
comes in to get them on the samepage.
I I know after that thatspecific presentation you're
talking about, we were asked tolike leave the room.
(55:04):
So I had the CFO and the CEO andthe head of IT starting to talk.
We had all the other executivesthere in the mic, could y'all
please leave and talk about thisoutside?
Um, that that's what you have todo.
You gotta get everybody on boardto push from the top down and
the bottom up.
So it's a top-down approach anda bottom up, and you got to hear
(55:26):
what the people in the trenchesare hearing.
And it and it and it can't bedisjointed because the last
thing that I'll say about thisis what is happening, the
transformation in happening isgonna be so fractured.
Okay.
There's gonna be a huge tail onthis that businesses are gonna
operate the same way.
Some people are gonna implementit in some areas, but the
(55:47):
companies that leverage thisthat can move forward towards
this AI first company are gonnacrush other companies.
I have I did a podcast a couplepodcasts ago.
I'd encourage anybody to golisten to it.
It is a driver's ed company,okay?
That is a true AI first companythat they are operating in such
an archaic industry that I wouldput bets on this that this will
(56:11):
be the biggest driver's edcompany in the next three to
five years in the country, handsdown.
I would put quite a bit of moneyon how they're operating and how
they're competing with it.
And and you should just listento how they're implemented.
I'm gonna be doing moreinterviews with AI first
companies.
Uh, and and I would justencourage everyone to go educate
(56:34):
yourself on this or or reachout.
Um, we are gonna be doing awebinar on this very soon.
Um, we've really focused on LLMvisibility and and how you can
leverage an AI layer in yourbusiness.
So um I would encourage you tocheck it out.
Rob, um, you're doing a lot ofthis stuff, you're implementing
a lot of these processes.
(56:54):
How do people find out moreabout your work, what you're
doing, check you out, uh, findout more about your company?
SPEAKER_00 (57:02):
Well, thank you.
Because I was just writing downsome of the stuff you were
saying.
I learned from you every time.
I'm like, I gotta remember that.
I gotta follow up.
But but thanks, Matt.
So so first of all, I am onLinkedIn.
Um, it's in slash Rob Newman.
My my name is up there, it'sspelled funny.
(57:23):
Uh, you can find us atnetformic.com slash US.
We're part of a multinationalnow, which is kind of cool to be
a partner in that.
Um, we handle the entire UnitedStates North America region.
Uh, you can find me uh anywhereMatt is because I follow him
(57:43):
around and uh I'm just trying tolearn from him.
But no, in all seriousness, anduh and it's uh and it's
rob.newman at netformic.com.
SPEAKER_01 (57:53):
So everybody go check out what Rob's doing.
If you need uh a dam or a PIMlayer to your ERP system, um
there's also a lot of other usecases we didn't cover.
I know that there was a lot tounpack.
You can go back and listen tothis at like 0.75 or whatever um
uh to to hear it again.
Uh I know I talk fast.
(58:14):
Um, I listen to everything atlike 1.5 minimum.
I that's all I can handle.
Um, but uh, you know, therethere's a lot of great value
here.
Hopefully you got uh a lot outof it.
If you did, please like, share,or follow.
Shaiko, we used to use the wordShaiko.
Please leave us a review on thepodcast.
Um, a lot of people are leavingus a review on our uh EWR uh uh
(58:38):
GNB.
I do appreciate that, but uhleave it on the podcast if it's
about the podcast.
Um, you know, if you like whatI'm doing, uh leave me a
LinkedIn review.
I would appreciate that too.
Leave it on me as an individual.
I think entities are really,really important and to separate
those out.
Uh, if you like to hear more,you like to hear this, reach out
to me.
(58:58):
I'm gonna put together aselection of playlists.
We are moving into YouTube, uh,best SEO podcasts.
We're we're everywhere.
I know we are on internetmarketing secrets too, but I'm
gonna fold that back into BestSEO podcast because I I think we
need to really get clarity anddrill down on this LM visibility
um and and going too wide.
(59:20):
Um, all those things are aconsideration in these
frameworks of AI discoverabilitythat that I'm putting together.
So you can check out some of ourframeworks.
We're gonna be uh launchingsome.org sites to help educate
the public uh on some of thesethings because I just think it's
so important.
Um, and so uh thank you so muchfor listening to the end.
(59:40):
Uh, if you would like to be aguest on our show, we do have um
uh a guest option on best SEOpodcast.
We're gonna be revamping thatsite.
Uh until the next time.
My name is Matt Bertram.
This is the Unknown Secrets orBest SEO podcast.
Um, bye bye for now.
This is the unknown secrets of internet marketing.
Your insider guide to thestrategies top marketers use to
crush the competition.
Ready to unlock your businessfull potential.
SPEAKER_01 (00:12):
Let's get started.
Howdy, welcome back to anotherfun-filled episode of the
Unknown Secrets of InternetMarketing.
I am your host, Matt Bertram.
I have been going through anidentity crisis based on Entity
SEO.
I'm known as Matthew Bertram.
I'm known as Matt Bertram.
The podcast is known by UnknownSecrets of Internet Marketing,
Best SEO podcast as a variety ofother things.
(00:34):
And I've been going through thelast two weeks some work to try
to unify those entities online.
And we have a common direction.
So stay tuned.
I know you have heard about mechanging the name of the
podcast.
We're going to just stick withuh the best SEO podcast because,
well, we all know that's what itis.
And um as uh the discussioncontinues to go further towards
(00:57):
LLMs and visibility uh in thoseLLMs online.
I wanted to bring somebody onthat somebody true, near and
dear to my heart, because he's amentor of mine.
He's been my a mentor of minefor a lot of years, and he was
doing AI even before it's cool.
It's not any reflection on hisage.
Um, he has been he has uh beendoing machine learning and
(01:22):
structured data for uh a longtime.
And now LM's hit the scene andeverything's cool, and he's
right in the middle of it all.
He's brought me in on a numberof projects.
I am honored to have the one,the one only, the Rob Newman.
Rob, welcome to the show.
SPEAKER_00 (01:36):
Hey, thank you so much for having me, Matt.
And um, it it really is apleasure to be here.
Uh, truthfully, I learn as muchfrom you as hopefully I've been
able to impart to you.
But um, this is a huge topicright now, just huge.
SPEAKER_01 (01:53):
So yeah, and and and to credentialize you a little
bit, uh, you built one of thefirst email automation software
companies back in the day.
You've sold multiple agencies.
You've recently sold your lastagency and you're doing big uh
e-commerce implementations thatare utilizing AI as well as uh a
variety of other things, andyou've worked with some of the
(02:14):
biggest brands in the world.
So um, you know, I'm I'm so gladthat we're able to work again
together on another project.
Thank you for calling meSaturday mornings and uh
starting to uh get get get allthis kind of spun up and and um
I'm glad where it sits now.
And so I just thought for theaudience, uh everybody
(02:36):
listening, um, I talk a lotabout uh B2B marketing and
what's going on with LMs.
And you've really brought LLMsinto the e-commerce component.
I think that there's uh a lotthere.
And um, you know, Pim, Dan, wedid a podcast a long time ago if
people want to go check it outon what that is.
(02:58):
Um, and I've been doing a numberof series around uh Amazon
marketing and uh ad testing andand utilizing AI.
And I thought you could reallykind of round out uh some of
that knowledge on on how we'reutilizing it.
So um one of the biggest thingsthat happened pre-COVID to kind
(03:19):
of tee and uh tee this wholething up and set the table is I
I interviewed um really somebodyfrom the Google Webmaster team,
okay, the spam team, uh, thathas written a lot of the code.
And and they came on and said tome that every business needs to
consider setting up ane-commerce cart on their
(03:42):
website.
Okay, and this is comingdirectly from Google, and this
is pre-COVID.
Okay.
And this was something that youand I have had a lot of
conversations about on how uheverything's coming together to
one point, right?
Like um, an example of this isyou can't have a LinkedIn
(04:03):
profile and uh uh a Facebookprofile if you're graduating
that that are two separatepersonas, right?
You got to unify that persona,you got to bring that together.
If you're doing online marketingum and you're you're selling
online or you're sellingoffline, 75% of that customer
journey happens online, and youwant to reduce as much friction
(04:25):
as you can to get people to buyonline.
So everybody needs to startthinking like an e-commerce
company.
SPEAKER_00 (04:32):
Without a doubt, right?
People's buying habits havecompletely changed.
Let's just think about ourselvesin a consumer experience, right?
I walk into Best Buy, I check iton Amazon.
I just make sure my price isright.
Um, it's a it's a continuousprocess.
But you know, given that this isSEO, I wanted to stick this in,
(04:54):
right?
You're spending money on ads,SEO, a lot of work as a business
to get people to come to yourwebsite.
My job picks up from there.
You've got people on yourwebsite, but are you actually
making money off of them, right?
So conversion is huge.
Shopping cart conversion ishuge.
(05:15):
But I like to say, look, ifyou're not finding, you're not
selling.
And a lot of what we use AI foris to make sure that the
products, the services, thethings that you want to sell
actually appear for your clientswhen they're searching for
something and they wantsomething.
Because in B2B, if they can'tfind it, in their mind you don't
(05:38):
have it.
Whether you do or not, they areonto your competitor and moving
right along.
And that's a huge benefit um uhfinancially.
SPEAKER_01 (05:48):
Well, I I can tell you like product page SEO is
incredibly important.
A lot of people are just usingthe manufacturer, depending on
what you're selling, right?
Like we can talk about a lot ofdifferent use cases, but you
know, I look for okay, I don'tknow, whatever it is.
If I'm doing this, I want it toactually say that it does that
(06:09):
in the product description tomake sure I'm buying the right
thing.
And if it doesn't have it, eventhough I think it does, and
something else says it and it'sa little bit more, even though
it might even be like very, verysimilar.
I I'm gonna err on the sign oflike, okay, this says this
component or whatever it is,fits with that.
And and and that falls in rightwith what you're saying of well,
(06:31):
if you're doing marketing, um,you got to be able to
differentiate, and that productneeds when people are searching
for it to rise to the top, to beone of the choices in the
selection process.
You also open up the can ofworms of you know, UX and the
customer journey on page, whichis probably arguably the most
(06:52):
important thing that needs tohappen, which is where SEO,
right?
You said SEO drops off, CRO kindof picks up, or there's a big
overlap there.
But when clients are typicallylooking for SEO or CRO, they're
really saying the same thing.
I want, I want a customer, Iwant you to put a customer
together with a bow on it, and Iwant you to deliver that to me.
(07:14):
And I don't care what multimodalyou're doing, like give me new
customers is essentially whatI'm asking for from uh most
people that are looking forthese services.
SPEAKER_00 (07:26):
Without a doubt, and and a point you just made is
really important, right?
You said I want it to say what Iwant it to do, or you know, it
describes exactly what I want todo.
And there are a few points thatare really important to that.
The first is in search, givingonly one perfect answer is not
(07:49):
as good as giving a few answers.
One, exactly what they'relooking for, but two or three
that are similar, right?
Or that are bought with, right?
Amazon is great at this.
Amazon is a great at a lot ofstuff, but you it actually has a
(08:09):
higher satisfaction score togive them one perfect and then
three options, and I like themstacked on the side that are one
is cheaper, one is you know, adifferent brand, one is
something that always getsbought with it.
Those things actually sparkinvestigation where people will
(08:31):
buy will one at a better ratebuy something, and two, it
raises your shopping cart value.
SPEAKER_01 (08:40):
Yeah, I mean the cognitive load is is really
something that people um try tocram too much information in
there.
Even Amazon, again, Amazon'sgonna test it, they'll figure it
out.
They just move to like, here'sall the choices of everything
you bought, you know.
And so if I'm buying a piece oftechnology, it's like, do you
want bananas too?
And I'm like, like it's toomuch, right?
(09:00):
And and I I just click next andI actually don't like the
current experience.
Um, they'll figure it outthrough testing.
But to your point, when you giveme three recommendations that
are tied into what I'm buying,and this is bought with that,
and it's something I canconsider, I usually will click
on that.
I'll usually look at that, andand it's not too much to
(09:22):
process, right?
And um on the website, um a lotof what we're working on right
now on on a separate project isokay, what is that customer
journey?
What are the steps that they'regoing through to help make a
buying decision?
And and what happens on mostsites and why the bounce rate's
so high is you have thesemassive menus, okay, that give
(09:47):
it so many options that youdon't know where you're taking
them.
Like if you're trying to getthem to buy something, all these
other things are just takingthem away from that.
And many times they're gonna getdistracted, they're gonna leave
the website, they're never gonnacome back, and that's why you
have a horrible bounce rate.
SPEAKER_00 (10:04):
Well, you were right, and then let me come back
to because I know you wanna youwant to talk about some of the
technical and AI things we do,right?
So, one of the things you alsosaid is does it say it?
Does it say yes?
Well, that is dang important.
You're right, and we have sometools that we utilize to make
(10:24):
sure it gets said, right?
And it's and it's more it andB2B is catching up, but it's not
there yet, right?
It isn't, but it has to be.
The expectation of people isthat there's no difference
between if I want to buysomething personally or for
(10:46):
business.
Now, one of the ways that we dothat is we use AI and some
private LLMs that are industryspecific, right, to enrich the
product information, right?
And we actually can do that intwo ways, right?
Um, one is just what are thedescriptions?
(11:08):
What are the things that tietogether?
What are the matching products?
What are those descriptions?
What material and partner itgets large amounts of data, but
it has to say what you'relooking for.
The second thing we use it foris voice.
With AI today, in for example,one of the products we might use
(11:29):
is called a product informationmanagement system, or B2B,
right?
Is we can not only take a sliverof product description and make
it useful to marketing and SEO,but we can also change the
voice.
I can take that same thing andtell it to describe this to a
(11:53):
five-year-old, describe this toum uh a buying agent, describe
this to an engineer, anddescribe this to call it the
general public, which might begood for SEO or whatever, or a
fifth one, right?
Generating specific keywords.
We can we can use thepersona-driven side and the
(12:17):
description side to create someincredibly rich uh data.
SPEAKER_01 (12:23):
So to to kind of unpack this a little bit more,
uh, a lot of uh clients thatthat listen or listeners are are
B2B companies, and I do a lot ofthings in B2B selling.
Uh ERP system, right?
So uh ERP system is is whateverybody's familiar with on
maybe inventory management,product management, whatever.
(12:44):
PIM sits on top of that and andit has kind of some machine
learning components, which wereused before LLMs with uh like
you you had to clean the dataup, you had to label the data,
and and then um you could usethese LLMs to to to get multiple
product descriptions, to uhextend the the uh the the
(13:08):
category or the definitionaround it, and it could speak to
more types of people.
Um and so uh, you know, if youjust have the the the ERP
system, you you're gonna getreally basic data.
You're gonna get just reallybasic labels.
And you you wanna you wannaexpand that reach and you wanna
(13:30):
expand uh understanding of howpeople are searching and what
they're lurking for, and you addthat AI or intelligence layer to
that, and and now there's morechances to get a sale because
you're answering the questionbetter, and it also is
understanding more about whatyou're looking for and who it's
trying to reach, and it givesyou a better chance to surface
(13:52):
um the right products for them.
SPEAKER_00 (13:55):
Absolutely, and you know, what I do never sounds as
good as the stuff you do, Matt.
I mean, you get all passionate,and I'm sitting here in the
background doing all the datawork and trying to put this
stuff together, but you'reexactly right.
You know, we've been doing thisfor a few years, several number
of years, right?
(14:16):
Where it started out as machinelearning, the the grandfather of
AI, right?
But being able to going through,and we started with cleansing
data, ordering data, creatingmore data and some structures,
it has now moved into exactlywhat you've described, right?
We have full-on AI models, andwe use both public models,
(14:39):
right?
So that it's understandable,they're they're well used, but
we also have created privatemodels, and they're important
for a couple of reasons.
One, a number of B2B companieshave um, you know, data they'd
rather not not have outpublicly, right?
We have we have a uh a companyright now in common, right,
(15:01):
that's in pharmaceuticalmanufacturing, and they have you
know secrets that they don'twant out there.
So we've created a LLMspecifically around their
chemistry, around the thingsthat they do, and we can search
it and serve up informationwithout revealing the secrets,
(15:22):
yet be able to pull in all ofthe information you're talking
about and do all the things thatwe do to give someone the right
answer.
Because that's actually, youknow what's great is it's just
like today.
How many people, you know, Iactually just want an answer
from Google, I don't want fivewebsites to go look at.
(15:43):
And what has happened?
And you talk about this, Matt,right?
I want an answer.
Just give me the freaking ah,that's the answer I'm looking
for.
Beautiful.
Move on.
I don't need to go visit yourwebsite.
SPEAKER_01 (15:57):
And and so, how for everyone listening, how what Rob
and I do kind of matches is I'mfocused on the public-facing
view of surfacing in thosepublic-facing LLMs.
Rob's maybe taken some of thosepublic surfacing or
public-facing LLMs, been able toput it on a separate server, MPC
server, a data lake, building adata mark, labeling the data,
(16:20):
and then you can recall thatinformation that's private, and
then it serves up inside aportal on the back end.
So it's taking um thefront-facing marketing piece and
now operationalizing it from ane-commerce piece.
And then there's an even greaterpiece of an expansion of the
business leverage or the AIapplied leverage that you can
(16:43):
get to apply it to other piecesin the business.
We're just talking aboutspecifically um the the product
catalog or um you know uh uhinformational search of a
proprietary library orproprietary database.
And it's fine-tuned from a youknow a chat bot standpoint to to
give you the answer.
(17:03):
And and again, that nextexpansion for the like VPs of
innovation that are listening,um, you can apply that to all
your data governance um to beable to recall data, to get
trends and and how this uhaffects what we're doing uh on
this shared project is you cansee the trends of what people
(17:24):
are buying, how often, the timethey need to refill and buy
more, and you can suggest thesethings to them to to help them
stay on top and to pull uhforward maybe some of that
demand as well as to just helpmake their life uh easier to
keep stuff in stock, whichbecomes really, really helpful
from a projection standpoint,from a uh data usage standpoint,
(17:48):
from a customer standpoint.
Um, these things are kind ofkeeping you out of the ditch.
So there's there's all theseadvantages right around um
purchasing orders and uh to toB2B companies that are listening
uh to be able to implement thison a supply chain side.
That's what we're getting a lotof uh interest in, Rob, is a lot
(18:08):
of supply chain logisticcompanies are are are seeing the
use for like uh AI layer to helpproject um a lot of these um uh
coordination uh and and deliveryschedules and uh uh fallout
rates in say oil and gas.
And so um it's it's justincredible what it what it's
(18:29):
able to do.
SPEAKER_00 (18:31):
It is, but I'm gonna back it down a tier because B2B
is behind, right?
So, you know, what if you're amedium-sized company and you're
selling through multiplechannels, you're selling to some
distributors, you may be sellingthrough Amazon or or whatever,
right?
Your Shopify store a lot ofchannels, TikTok store and your
(18:53):
Shopify store and and and andand right, yeah.
One of the key problems to solvefor these companies is
inventory.
Yes.
How do I know how much inventoryI need to sell on Amazon versus
my Shopify?
What if I have more than onewarehouse?
So, what they wind up doing isactually a very old-fashioned
(19:13):
solution.
They put in safety stock, whichis extremely important.
And if I allocate safety stockincorrectly to one channel and
the other channel sells out,I've got a bunch of stocks
stranded over here that I couldhave sold, right?
So using AI and using you knowgood connectivity, good data, it
(19:37):
can actually manage all thechannels at one time.
And we do that in in PIM.
And the reason we do that iswe're able to work directly with
the ERP, apply some AI, look atthe channels, manage and even be
predictive in the selling.
Now, Amazon can be helpful withthat, but I know a lot of
(19:57):
customers that are doing safetystock across all of their
channels, and you don't need to,and it's expensive.
SPEAKER_01 (20:04):
Well, one of the things that you and I were
talking about the before thiscall, uh, to to kind of build on
that concept was, you know, ifyou're a supplier, right?
If you're a supplier, you'retypically, you know, you're
you're either repackaging orreselling other people's stuff,
or if you're buying somethingand you're configuring something
(20:25):
or whatever, um, that ERPsystem's not going to do a
really good job of recognizingall these different product
orders and POs and um you knowproduct labels.
And if you're trying to resellthat as your own product, right?
You're you're you're you'reyou're you have a markup or uh
you're you're on our arbitragingwhat's going on and you're uh
(20:48):
buying that thing from a bunchof different sources, okay?
They're gonna have their own POnumbers and product label
numbers.
Uh, and you got to be able tounify that to your point on the
product supply side, uh, andthen present your own product
label number.
You can't do that really well.
It's a very manual process withthe ERP system.
(21:09):
And a lot of people are banddating and gerry rigging that
up.
And I just wanted to flesh thatout a little bit more because
that's where the the the PIM andthe dam come in is you can run
all those APIs uh into that AIlayer, it can figure it out and
tell you um what's at where,what store, what it is, and it
(21:30):
and it streamlines that productso you can have just in time um
delivery where you're notcarrying all that overhead.
Um, and I think that that'ssuper powerful because I know
that there's people listeningright now that they're using
those ERPs and they're like,there's got to be a better way
to do this, and they're tryingto, you know, uh band-aid up a
(21:54):
bunch of different solutions.
SPEAKER_00 (21:55):
Sure.
Well, ERP was, of course,primarily set up as a financial
system, right?
And it's really good at it.
It will, it will, it will handleyour finances, it's designed to
plan your resources, but in theend, right?
Um, and you threw a lot ofinformation out there, Matt.
But in the end, right, what arewe trying to do?
(22:17):
We're trying to sell product,and that may come in many
flavors.
It could come from multiplevendors.
You may even have a white labelproduct that you do of your own.
You may have um the opportunityif you're a distributor, you
probably sell multiple brandsthat do the same thing.
Relating all of that is hard.
(22:40):
It's a lot of data.
PIM can make that so mucheasier.
You can sit there and do it byhand in an ERP and start
matching stuff up, or you caningest it, connect it, and use
AI to begin to put all thesethings together and then deliver
it to your AI-driven search.
It becomes a whole lot easier.
(23:02):
And it gives you a lot offeatures that people have not
been able to do before.
I'll give you uh an example fromone uh plumbing client.
We were able to take all of thatinformation, create all the
relationships between Delta,Kohler, whatever the brands
were, their own white labelbrand and off brand, and a lot
(23:23):
of data.
And in the end, we were able tosurface one exactly what the
customers, if you put in KohlerValve 1, 2, 3, you will get a
Kohler Valve 1, 2, 3.
But you will also get theirwhite label exact product that
matches it, which they make moremoney off of.
(23:44):
It's a cheaper product, butwe're maximizing their profit.
And then we'll give you someother options.
This data is truly valuable.
SPEAKER_01 (23:53):
Okay, so I want to break one thing apart and and
dig in a little bit more of whatyou just said.
So, okay, so we have like ourwebsite and and you know,
there's an ecosystem associatedwith that.
We have it connected to an ERPsystem.
Now we're talking about a a PIMor a DAM or a hybrid of that on
(24:15):
pulling in data and API.
Um, is what you're talking aboutnow adding an AI search layer on
top of that, or is that stillpart of PIM and Dan?
And like, what are thedifferences between plugging
something like that in to giveyou even more granular search to
give people the absolute rightanswer of what they're looking
for?
Because that's some morehorsepower.
SPEAKER_00 (24:37):
It is, and it actually requires both.
So okay, I just wanted to makesure we get to we may get to one
as we begin to be able to usemore and more unstructured data
to do these things, but we haveto just like any LLM, right?
We have to have the descriptorsin there for the AI to find
what's common, right?
(24:58):
If all I do is put in screw andnail, right, they're not going
to be able to connect it ashardware.
So uh, you know, it it doesn't,it doesn't always work that way.
Now, what we do is we take allof that data, we do some
structure inside of the PIM, anda PIM is the product, the dam is
(25:20):
media, digital assets.
That's where that comes from.
And what we're able to do isenhance that data inside of the
PIM, give it all the terms thatthe AI will need to search and
look for.
And then when we serve that up,the AI-driven search is able to
just go to town.
(25:41):
I've got your materials, yoursize, your product, your
everything that goes into thathas been enriched to allow the
full search capabilities acrosseverything.
So it's two steps right now.
SPEAKER_01 (25:56):
Yeah.
So what what's just I there's alot of people out there, right?
And I I just the term uhinformation's the new oil,
right?
Has been talked about for a longtime.
And you had to do you had tohave a a data uh analyst team,
you had to spend months like toslow down like the data
(26:17):
labeling.
Like you're building, likethere's a ton of people
listening that have data marksand they have tons of data and
they haven't done anything withit.
I would love to hear from youwhat are some different use
cases if they start to implementPIM, DAM, some AI-driven search
on on to let's just open up theminds of what are some of the
(26:39):
things that they could do withthis data.
I'd love you to just rattle offa few uh high value use cases
where you could make more moneydoing these sort of things.
You can leverage it.
SPEAKER_00 (26:48):
Yeah, when you've got data buried across wherever
it is, right?
Being able to relate data andserve data are the two
beginnings of better revenue andtotal cost of ownership, right?
And the way you can do that isby beginning to tie that
(27:09):
information to what's actuallyhappening on the website.
You can serve it and you know,and and actually test out what
appetite is there for thisproduct data and information and
products.
You can use it to make yourproducts more enticing, more
descriptive, provide moreinformation to different
(27:33):
personas and buyer groups.
And then let's not forget,right, once we've invested in
all this data, long tailmarketing and long tail find is
really important, right?
Why does your website care ifyou know um so many clients,
right?
They have the top 20% of theirvolume on the website, not 100%.
(27:54):
80% is buried in catalogs or asalesperson's mind.
So what happens?
Salesperson retires, and a lotof guys are beginning to retire
in in these manufacturing anddistribution businesses, and
you've lost it.
The e-commerce and search doesnot care.
Put all million up, and the onetime that somebody buys it, you
(28:19):
are way, way ahead.
It also helps in your search.
So, in the end, just from a fewideas, right?
You're getting more people toyour website, you're getting
more sales off your website, andthen you're also going to be
able to create more data tounderstand your customer better.
What's the customer journey onthese kinds of things?
Did we miss something thatsuddenly turns out to be
(28:41):
extremely important right now inengineering or in uh in a
buyer's group?
The only way you can find outthis data is let's expose it and
run some reports.
Let's find out.
There's so many opportunitiesthat um that you can do with it.
And I'll I'll give you a reallife example, right?
(29:02):
So we had a tubing company,right?
Very, very boring industry,right?
They might say.
And yet it's critical, right?
They sell a lot of tubes, theythey're all different sizes,
right?
And what they realized is therewas a that by just putting up a
few things they don't normally.
(29:24):
One, their CAD drawings.
CAD drawings are typically filedaway somewhere.
Putting up their CAD drawings isa true blueprint of what it's
going on, the real information.
And then secondly, using AI tomake their detailed product um
(29:46):
PDF sheets just readable.
Traffic to the website went uptwelve percent.
Twelve.
Just making all this exposed.
Right, went up 12%, and thebuying went up three percent,
three percent pure revenue, moreclients, more revenue.
(30:10):
Nicely just exposing the data,which the IT does not care, it
does not care how many things Istuff in it, it doesn't care how
many calls you make from it.
It was you know a one-timeeffort, create a process, and
now I'm making more money.
SPEAKER_01 (30:28):
You know, this is the hard thing because AI can be
applied to almost every aspectof the business to help improve
it in some way.
If we drill down to well, AIlearnings, and we're talking
about like CRO conversion rateoptimization, like understanding
that customer journey becausethe buyer behaviors changed and
(30:50):
people want more of that proof.
You you started, I startedthinking about eat expertise,
authority, trust, andexperience.
That's what you were doing isyou were proving that out more.
So people said, Oh, this is whatI'm gonna get.
This is how it's gonna do.
It's explained.
The LLMs understand it, thepeople understand it.
It's a win-win because thatprocess is happening.
(31:11):
A lot of people are like, I'mgonna save all that until I get
them on the phone, but nowthey're missing the deal because
they're not even in the topthree choices or whatever the
selection is because they'realready out of the running
because someone's deselected andunraised their hand from using
them.
And so, I mean, you look atthese bounce rates, you look at
(31:34):
these click-through rates,they're minuscule.
You you gotta get that numberup.
And when they get to yourwebsite, you got to understand
what content are they missing?
Where did they get stuck?
Do they get stuck cardabandonment, right?
That's a huge, huge issue.
Why do they put it in there?
What's going on with that?
I would love to hear a littlebit more about how you look at
conversion rate optimization,heat mapping, um, and and how
(31:58):
you're starting to make some ofuh those decisions because yeah,
SEO gets them there.
We have there there's a a aconversation that has to happen,
and and really, you know, 50% ofthat informational content is no
longer on the website.
So you got to drill down on themiddle of the funnel and bottom
of the funnel content.
Because if someone's came toyour website, right, they've
(32:20):
gone past the AI overviews orthe LLMs have recommended them
or whatever it is, now they'reon your website.
If you don't get some kind ofinformation, and we got, you
know, we got all the uh pixelsgoing away, right?
And so um you got to capturetheir information and you got to
get them to the next step, andand you got to push them through
that funnel.
(32:40):
And the only way you're gonna beable to do that is with data,
and that's the bigdifferentiator today that you
can you can still analyze isonce they come to your website,
like you know, they're they'rein your territory, right?
SPEAKER_00 (32:53):
Absolutely.
Um, and it it can be a littlemore difficult in the B2B space.
So there are there are a fewthings that that we do, right?
One, don't just provide yourfellow your telephone number.
That's the way to interact withyou.
It sounds funny, but there aretons of clients we start with
where call my salespeople andwe'll get going.
(33:15):
Well, most people don't want totalk to your salesperson until
they're 100% ready to buy.
Yes, not a good strategy.
So, what what do we do?
First of all, we want to be ableto utilize some methods, right?
That in and and they're consumermethods, right?
We we learned them and we knowthese from websites we interact
(33:39):
with all the time.
How would you like to why don'tyou just save a few favorite
parts so that you can come backand look at it later?
No problem.
Give me your email address andI'll save them for you, right?
These are standard things, butyou know what?
Why don't you join our loyaltyprogram and I'll give you points
toward everything you buy?
(33:59):
And if you give me my your emailand phone number right now, I'll
stuff you with uh a hundredpoints just for signing up,
right?
Um, and as you go, so that's onthe very early side.
SPEAKER_01 (34:12):
Quizzes and calculators.
I love quizzes and calculatorsand engaging them to uh but you
but we got to have your email tobe able to give you the answer,
right?
SPEAKER_00 (34:21):
Uh you do, right?
So um, you know, what's youraddress?
I'll calculate, I'll calculateout your your shipping address.
Do you want you know, we usegated content, right?
Yeah, I have really valuablecontent.
Just give me your email address.
Everything you can do to getsomething from the client is
really important, but not everyclient does.
(34:42):
So what happens?
Now, let's say you do have ashopping cart, you know that it
is the most effective way forreordering.
You know that it's the you know,the way sometimes people want to
try out a small buy before theygo call your uh your salespeople
and give you a big purchaseorder.
(35:02):
They just want to order one, nota thousand or a million, right?
Just one.
Let me buy one, let me see whatit looks like.
That happens all the time.
So, what happens?
You get into the uh shoppingcart and then they stop.
It's not high on their prioritylist, it's very high on your
priority list, right?
(35:24):
So, what do you try to do?
First of all, right, we'retrying to get their information
up front.
Hey, create a shopping cart.
Okay, to create a shopping cart,let's create a quick account.
Where am I sending it?
What's your name?
What's the company, right?
It and we create all that.
You've got it sitting here inyour shopping cart, and now I've
(35:45):
stopped.
Well, what here's two thingsthat that surprise me now.
Yeah, you send emails, of courseyou do.
Hey, you stopped something inthe shopping cart.
What I was very surprised by iswe encountered a marketing
company and they convert thatand they send automatic
(36:06):
postcards.
I love that.
SPEAKER_01 (36:09):
I was like, just direct shipment, yeah.
SPEAKER_00 (36:11):
Just a direct here's a postcard, and by the way, I'll
give you 10, 5, 20.
They actually look at what's inthe cart.
Can I ship them a 10 coupon?
SPEAKER_01 (36:23):
And that's that's so that the the drop shipping to
set something like that up.
I actually had a conversationwith a client about that last
week.
Is let's just put a tag in theCRM if you want to drop ship a
postcard to them that lookshandwritten, and you can have a
couple different options.
You could even put, you know, uhlike a a tier up from that, send
(36:45):
them cookies, whatever.
And and the conversion rate'sgonna be so much higher because
you're building reciprocity andyou're hitting them offline
because there's so much noiseonline, people get distracted,
you know.
SPEAKER_00 (36:56):
And oftentimes, well, sometimes they don't use
the right email address.
I don't know why, but sometimesthey don't, right?
So that's true.
There are methods around it,particularly on like gated
content.
If you're selling information asyour product, right?
You can you can insist on a goodemail because that's how you're
(37:19):
sending access to theinformation.
But if it's a low value, youknow, sign up to get something,
you know, often you know, we'llsee you know, junk12345 at
gmail.com uh as the thing, andtherefore is that really
working?
Where in the shopping cart, ifthey want it, they will give you
(37:39):
the right address.
Absolutely.
SPEAKER_01 (37:42):
Yeah, I I think I think adding the text automation
to that, like I mean, I'mstarting to get a ton of spam
text, but I still need to deletethem and clean out my text
messages.
Absolutely, but text it has awhatever 98.9 or 99.8 percent
open rate still.
Uh email, it people are justdeaf by email at this point.
(38:03):
Like, we're having to set up Robwith a lot of our clients, Slack
chat because they're not gettingour emails, um, because they're
overloaded with emails, they'reworking on whatever they want to
work on.
So here's a specific form ofcommunication to talk to.
So, like, I mean, emails gettingtougher and tougher.
I mean, now you can use AI to,of course, clean out your emails
(38:25):
and you know, sort by thedifferent categories you're
using it for, but email email'stough these days.
Uh, it's not always the mostreliable, but if they get it in
the mail, right, they're gonnapick it up and see it.
So, I I love that example.
That's a great example.
SPEAKER_00 (38:40):
Yeah, everyone looks at their mail, even the junk
mail.
You still look at it.
Is it junk?
SPEAKER_01 (38:47):
Yeah, you gotta, you gotta, you gotta define it, you
gotta categorize it.
SPEAKER_00 (38:49):
Yeah, you gotta define what is it, right?
Um, I I think your point aboutemail is is spot on, right?
And that is it's so full, it'sso completely um uh uh overused.
I mean, I watch my my thenumber, the red number grow on
(39:10):
my email.
And then sometimes I just gothrough okay, anything older
than this, gone, right?
So it it's tough to get throughthe clutter.
SPEAKER_01 (39:23):
So let's let's bring this back and and kind of bring
it back to to marketing andclose it out.
Uh I think a lot of people arehearing, okay, AI is gonna
affect business, right?
It's gonna have a positiveimpact.
Where can I see that impact?
How can I apply that?
I think we went through a numberof steps on how to do that.
I think that Pim and Dam, if youhave an ERP system, you should
(39:45):
be really critically looking atthat.
How can I get uh uh AI to applyto different categories?
And that could branch out in alot of different ways.
CRO and SEO are connected, okay?
I think they're attached at theat the hip.
A lot of times when people wantone service or another, they
really want the same thingbecause they want to just open
up that funnel and and tightenthat funnel and bring that
(40:07):
together and unify that data.
And so, you know, if you'reusing LLMs to make the website
better for a public facing,right?
So you got you got behind thebehind the wall and they set up
an account kind of on that backend when they log in.
But but a lot of the uh look andfeel is that first-time cut
(40:29):
customer interaction, gettingthem into the right customer
journey in the right channel.
Um, you got to have thatinformation publicly facing the
LM's got to be able to read it.
You got to have a uh a site map,you got to have a uh organized
and logical categorization ofyour your product category.
You got to have gooddescriptions to speak to
different people.
One of the things we didn't talkabout is um, you know, I like to
(40:52):
use like this profile because Idid it many times.
Um, but these different types ofpeople want that information,
like you said, given to them,right?
Or some people might want to digmore and get that research, or
some people might want to uhvideo uh or or audio or voice
hear it, right?
They I just I want it just spoonfed to me, right?
Some people want to submit uh uhuh email and have you contact
(41:16):
them.
Some people want to sign up fora calendar, some people want to
download a form and talk tosomebody, somebody people want
to call a 1-800 number.
And if you're not addressing allthese different modalities, you
are missing customers, you aremissing uh uh distribution
channels on marketing.
(41:37):
Um, you need to be able tosplinter this information and be
able to push it out multipleways to bring everything back to
your main source hub.
And you know, something that weprobably don't have time to go
into too much, but hopefully itteases it up is uh we need to
understand your company as anentity.
(41:59):
We need to define that.
We need to define your keyplayers of who the authors are.
All these things are theinformation architecture, which
is really what you specialize inand kind of how we work together
is is mapping that out andmatting mapping the site
architecture out.
And that's how how you and I areworking together really well uh
(42:22):
on some of these projects,because you need different
voices at the table, differentexperts giving different
opinions of looking at itdifferently from design, from
the back end, uh, from themarketing, from you know how the
LLMs are picking it up.
And it's all well, marketing andpositioning and um well,
(42:46):
information architecture.
That's what Google's trying todo, uh, is understand that's
what all these LLMs are tryingto do.
They're trying to figure out theinformation, trying to organize
it and layer in a way that youcan recall that information uh
really easily.
We're talking about likeinformation recall, is is
really, I think, the the the bigor data recall.
There's a whole industry aroundthat.
(43:07):
And um, that's what we'retalking about, whether it's on
the website or before thewebsite, on the website, or in
the website.
And and I think that um thatgets underserved sometime.
And and these are complex issuesthat that you need to unify a
lot of data.
And I I think just the executionof companies is they've they've
(43:32):
spent a bunch of money on on oneor two different things, and it
and they just are 85% of the waythere.
And they just need someone, Rob,like you to like help connect
that data together.
And then the circuit startsrunning, and it's really, really
quite powerful.
Um, or or someone like you, youknow, and and I I just think
that uh this is where things aregetting missed.
(43:54):
Like they've they've built thedata lakes, they've labeled a
lot of the data, and it's justsitting there, and they're not
getting that the the insightsthey're looking for.
They haven't gone the rest ofthe way.
And that last 10 or 15 percentis the hardest way to do it.
And once you do that, that thatcircuit gets completed, and and
(44:14):
now you get all these benefitsof the you know, uh the the AI
machine learning, LLMs, whateveryou want to call it.
SPEAKER_00 (44:21):
I agree, and the and the thing is look, we've talked
about removing silos forever,right?
SPEAKER_01 (44:27):
Oh my gosh, yeah.
SPEAKER_00 (44:28):
But leaving your information internal and not
utilizing it, right, is anothersilo.
We have to break down the dataof the company to the consumer.
One of the things you talk aboutquite a bit is you know putting
a stake in the future marketing,right?
(44:48):
The LLMs will rule all search,they will, they will be giving
an answer, you know, and and somany companies have developed
out so much intellectualproperty and they have so much
that just isn't displayed.
Part of this is stick the flagin the ground for the future,
(45:08):
because what we're doing rightnow lays the groundwork for
years of future success becausethe LLMs are looking now,
they're looking for, and youknow this better than I do, but
they're looking for authority,they're looking for depth of
content, right?
They're that we're no longerlooking at high-level marketing
(45:29):
words.
They want you to demonstratedeep knowledge and being able to
expose some of the things Italked about, whether it's an
engineering drawing to um to PDFinformation and getting it in
such a way that you're properlystaking your flag and saying,
look, we know tubing better thananyone else.
(45:52):
It's important, right?
You you think it's not, and thatthere's lots of competition.
But the people who stake who getthere first and show the depth
of competency have a huge, hugeadvantage.
SPEAKER_01 (46:07):
That is the unknown secret that I don't think people
are aware of, that the trainingdata in these LLMs, they've just
opened up like the World WideWeb from their kind of training
data that they're built on.
And the LLMs are still trying todecide what gets uh saved in
their long-term data in thetraining data sets that they
(46:28):
continue to reference.
And the more and more that youget recalled and referenced, the
harder and harder it is gonna beto dislodge you as that
authority or that expert in theLOM data set because every
recall does that with Google.
You know, there's a little bitmore play.
These LLMs are gonna solidifyand in 18 months, it is gonna be
(46:52):
you know extremely hard to dothat.
And the people that are plantingthat flag in the ground now and
and building those libraries andthose data sets and becoming the
the go-to recall expert forthese things are gonna win.
And people are gonna be spendinga lot more in the future uh to
(47:13):
just get a piece of that becausethe this this is this is uh the
time that people need to likeget with it and and take action.
Um, I just did a uh article onthis of a kind of like the LM
land grab, and and I think it isabsolutely so true.
(47:34):
Um, and in some of thepresentations, you know, I mean,
just ask ask any LLM, they'lltell you the same thing.
Like this is where we're attoday, and this is a monumental
shift on how business is doneand where it's going.
And I'm seeing the jump rates ofof usage to to your point.
(47:55):
Uh yeah uh Google was thelibrarian, okay.
And you would go to Google andthey would give you 10 blue
links.
Here's where you go to goresearch it.
Now people just want thatanswer.
Give me that answer, and you godo the research, synthesize the
data, and give it to me.
As people start to use that,they don't go backwards, okay?
(48:16):
They they only move forward, andwe're seeing you know, huge
amounts of increase in this.
And as people get more and morecomfortable with this, they're
they're using it more and moreoften.
And and and this is the future,it is not a fad.
Um, and and uh any business thatwants to be forward thinking
(48:38):
needs to needs needs to startjumping on this.
And one of the things you'retalking about is that push to
becoming an AI first company,which there's a couple steps to
make that jump.
Amazon didn't do it overnight.
There's a sequence of steps thatyou have to make.
And I feel like a lot of people.
SPEAKER_00 (48:54):
Oh, I like that.
I I gotta trademark that, Matt.
AI first.
An AI first company.
I love it.
But I but you're right.
And here's the thing thatunfortunately people will think,
right?
Oh, AI makes it so easy, sofast.
15 months from now, I can stillget on the in on that 18-month
(49:16):
land grab.
Unfortunately, the answer is no.
Uh, you cannot wait.
The amount of time, right?
Part of it is the amount of timeyou're sitting in the next 18
months.
18 months is a flash, it will gofast.
And as you're thinking about2026, it's better to start
(49:38):
earlier than later.
If you're going, if you're goingto be serious about owning this
space, you need to start rightaway because it takes time to
set it up properly, it takestime for it to be ingested
properly, and it will take timefor these LLMs to decide, right,
that your data is what theyreally want to uh quote in the
(50:02):
future.
SPEAKER_01 (50:03):
And you got to get your IT team on board.
You need to get you you got toget IT and security are gonna
be, you know, uh, they they needto be rowing in the same
direction because uh the biggeryour company is, uh, the more
time it's gonna be taking to geteverybody on board with this, to
make those changes, to get thoseapprovals.
(50:24):
And what you're talking about isyou need to be showing up now,
okay?
Because every time that somebodyelse shows up, they're getting
their moats getting bigger andbigger and bigger and bigger.
That's right.
And if you're at a companythat's gonna take a quarter or
two to get this implemented orapproved, you got to start
pushing on this now.
(50:45):
So so you even have a shot atthis.
But the nimble companies wherewhere everybody's on board and
is roan in that same directionand understands this and has
this education, that's why I'mgetting called in for a lot of
talks, is to to to get everybodyon board with this, to
understand what's happening, toknow that we have to take action
(51:06):
now.
Like the next couple of quartersare really, really critical.
Um, and I think that the gapsare gonna really widen between
the people that get it and doit.
I I have on the opposite end ofthe spectrum, Rob, like one
client that it's kind of pushinguphill.
Um, this was about six to ninemonths ago.
(51:27):
They uh are owned by anothercompany that's owned by another
company, so they've got allthese approvals going up.
They got purchased.
And their stance is we don'tunderstand AI.
We want and and the contentcreation and and SEO, these
these are the marketing salesnow.
HR is about to be affected.
Like these are the things thathad this disruption first.
(51:50):
They said no content, like wewant a hundred percent zero AI
content, okay.
And they were pushing backcontent we were sending them
when we actually had a subjectmatter expert writing it.
And where do you think theseLLMs, Rob, were trained?
Where do you think theinformation they were trained on
actual humans writing content?
(52:11):
So it's very hard to get to ahundred percent.
Like, we were actually having totwist and change the content to
get that to submit it.
That's where uh uh education umat the executive level needs to
happen because this needs tohave champions all the way
through.
You got VPs of innovation, yougot CROs, but like everybody's
(52:32):
got to be on board with this andunderstand it because a lot of
people are afraid of it and arepushing back or are worried
about their jobs.
I have told we've implemented AIheavily in our company over the
last two and a half years.
And we just gave people morereach, we gave better
productivity, we we gave peoplethe the access to do more
(52:55):
things.
It actually wasn't limiting umand and like, hey, we need to
get rid of these jobs.
It was actually how we can domore, deliver more, deliver
better quality work.
Um, but it's all about thatperception.
And if you're thinking AI isgonna replace you, it is.
If you're at a company andyou're worried about that, you
(53:15):
need to understand how they arebecause somebody that that uh AI
is not gonna replace you yet.
These agents have a little bitmore time before you turn out
full agentic AIs, and we'veheard some horror stories about
that.
But somebody leveraging AI andunderstands how to use these
systems will absolutely replaceyou if you're not understanding
it.
SPEAKER_00 (53:35):
And that's the truth, no doubt about that.
And I would I would take it upone level, Matt.
This is a CEO issue, this isstrategy, this is vision, this
is the future of the company forthe next five to 10 years,
right?
IT and marketing alone are notgoing to drive these decisions.
(53:57):
You and I had it, we were at aat a customer and we were
talking to him, and I wassitting next to the CEO, and he
sat there and he said, We haveto do this, right?
He got it.
He he he just was like, I don'thave a choice.
If I want to be there in thefuture, I have to do this.
(54:18):
And the good news for them,right, is they have they have
information, it has to be putout there, we have to make the
land grab, but they're actuallyin decent shape.
But the CEO totally got it, andsuddenly, guess what?
Oh, the CFO, the CMO, the the CCRO and the CTO were all like,
(54:41):
yeah, we got to do this.
SPEAKER_01 (54:43):
Well, that was that's rare though, that you get
all of them in a room and youget everybody on the same page
and they understand it.
And and that's where this uhexecutive level advisory really
comes in to get them on the samepage.
I I know after that thatspecific presentation you're
talking about, we were asked tolike leave the room.
(55:04):
So I had the CFO and the CEO andthe head of IT starting to talk.
We had all the other executivesthere in the mic, could y'all
please leave and talk about thisoutside?
Um, that that's what you have todo.
You gotta get everybody on boardto push from the top down and
the bottom up.
So it's a top-down approach anda bottom up, and you got to hear
(55:26):
what the people in the trenchesare hearing.
And it and it and it can't bedisjointed because the last
thing that I'll say about thisis what is happening, the
transformation in happening isgonna be so fractured.
Okay.
There's gonna be a huge tail onthis that businesses are gonna
operate the same way.
Some people are gonna implementit in some areas, but the
(55:47):
companies that leverage thisthat can move forward towards
this AI first company are gonnacrush other companies.
I have I did a podcast a couplepodcasts ago.
I'd encourage anybody to golisten to it.
It is a driver's ed company,okay?
That is a true AI first companythat they are operating in such
an archaic industry that I wouldput bets on this that this will
(56:11):
be the biggest driver's edcompany in the next three to
five years in the country, handsdown.
I would put quite a bit of moneyon how they're operating and how
they're competing with it.
And and you should just listento how they're implemented.
I'm gonna be doing moreinterviews with AI first
companies.
Uh, and and I would justencourage everyone to go educate
(56:34):
yourself on this or or reachout.
Um, we are gonna be doing awebinar on this very soon.
Um, we've really focused on LLMvisibility and and how you can
leverage an AI layer in yourbusiness.
So um I would encourage you tocheck it out.
Rob, um, you're doing a lot ofthis stuff, you're implementing
a lot of these processes.
(56:54):
How do people find out moreabout your work, what you're
doing, check you out, uh, findout more about your company?
SPEAKER_00 (57:02):
Well, thank you.
Because I was just writing downsome of the stuff you were
saying.
I learned from you every time.
I'm like, I gotta remember that.
I gotta follow up.
But but thanks, Matt.
So so first of all, I am onLinkedIn.
Um, it's in slash Rob Newman.
My my name is up there, it'sspelled funny.
(57:23):
Uh, you can find us atnetformic.com slash US.
We're part of a multinationalnow, which is kind of cool to be
a partner in that.
Um, we handle the entire UnitedStates North America region.
Uh, you can find me uh anywhereMatt is because I follow him
(57:43):
around and uh I'm just trying tolearn from him.
But no, in all seriousness, anduh and it's uh and it's
rob.newman at netformic.com.
SPEAKER_01 (57:53):
So everybody go check out what Rob's doing.
If you need uh a dam or a PIMlayer to your ERP system, um
there's also a lot of other usecases we didn't cover.
I know that there was a lot tounpack.
You can go back and listen tothis at like 0.75 or whatever um
uh to to hear it again.
Uh I know I talk fast.
(58:14):
Um, I listen to everything atlike 1.5 minimum.
I that's all I can handle.
Um, but uh, you know, therethere's a lot of great value
here.
Hopefully you got uh a lot outof it.
If you did, please like, share,or follow.
Shaiko, we used to use the wordShaiko.
Please leave us a review on thepodcast.
Um, a lot of people are leavingus a review on our uh EWR uh uh
(58:38):
GNB.
I do appreciate that, but uhleave it on the podcast if it's
about the podcast.
Um, you know, if you like whatI'm doing, uh leave me a
LinkedIn review.
I would appreciate that too.
Leave it on me as an individual.
I think entities are really,really important and to separate
those out.
Uh, if you like to hear more,you like to hear this, reach out
to me.
(58:58):
I'm gonna put together aselection of playlists.
We are moving into YouTube, uh,best SEO podcasts.
We're we're everywhere.
I know we are on internetmarketing secrets too, but I'm
gonna fold that back into BestSEO podcast because I I think we
need to really get clarity anddrill down on this LM visibility
um and and going too wide.
(59:20):
Um, all those things are aconsideration in these
frameworks of AI discoverabilitythat that I'm putting together.
So you can check out some of ourframeworks.
We're gonna be uh launchingsome.org sites to help educate
the public uh on some of thesethings because I just think it's
so important.
Um, and so uh thank you so muchfor listening to the end.
(59:40):
Uh, if you would like to be aguest on our show, we do have um
uh a guest option on best SEOpodcast.
We're gonna be revamping thatsite.
Uh until the next time.
My name is Matt Bertram.
This is the Unknown Secrets orBest SEO podcast.
Um, bye bye for now.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Are You A Charlotte?
In 1997, actress Kristin Davis’ life was forever changed when she took on the role of Charlotte York in Sex and the City. As we watched Carrie, Samantha, Miranda and Charlotte navigate relationships in NYC, the show helped push once unacceptable conversation topics out of the shadows and altered the narrative around women and sex. We all saw ourselves in them as they searched for fulfillment in life, sex and friendships. Now, Kristin Davis wants to connect with you, the fans, and share untold stories and all the behind the scenes. Together, with Kristin and special guests, what will begin with Sex and the City will evolve into talks about themes that are still so relevant today. "Are you a Charlotte?" is much more than just rewatching this beloved show, it brings the past and the present together as we talk with heart, humor and of course some optimism.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.