March 4, 2025 • 39 mins
About This Episode
Is your eCommerce site search matching the products you sell with what your potential customers are searching for? Or are you leaving visitors frustrated with 0-result search result pages for terms adjacent to what you sell?
In the 15th episode of eCommerce Masters, Host Ethan Giffin (virtually) sits down with Ari Khan, CEO of Hawksearch, about how Conversational AI is changing the way people shop online. Ari shares how smart search tools like Hawksearch can increase conversions and improve the customer experience, ultimately boosting revenue.
Ari explains why site search should be viewed as your online salesperson, guiding customers to the the products that most closely match their search intent with AI-powered recommendations, real-time suggestions, and natural language understanding. With over 70% of online purchases starting with a search, putting the most relevant products in front of potential customers has become eCommerce table stakes. Ari highlights the importance of clean product data, smart merchandising, and accurately measuring results so that merchants can further optimize search performance.
Ari reveals Hawksearch’s latest innovations, like Conversational Search, which engages shoppers with clarifying questions and dynamic recommendations—just like a real salesperson would. This AI-driven approach helps businesses reduce zero-result searches, increase average order value, and keep customers engaged on-site (where margins are their highest).
Want to learn how better on-site search can drive more sales? Listen now to hear Ari and Ethan’s insights and strategies to improve your customer’s eCommerce search experience.
Chapters00:00 - Introduction & Guest Welcome 02:18 - Ari Khan’s Journey into AI & eCommerce 07:01 - The Role of Site Search in Online Sales 12:55 - How Conversational AI is Transforming Search 18:07 - Optimizing Search for B2B & eCommerce Growth 23:50 - The Importance of Clean Data & Merchandising 29:46 - AI-Powered Search & Reducing Zero-Result Queries 34:41 - Bridging Content & Commerce with Intelligent Search 40:08 - The Future of AI in eCommerce & Site Search 44:44 - Closing Thoughts & Where to Learn More
ResourcesHawksearch - https://www.hawksearch.com
Ridgeline Digital - https://www.ridgelinedigital.com
BigCommerce - https://www.bigcommerce.com
Shopify - https://www.shopify.com
Optimizely - https://www.optimizely.com
Salesforce - https://www.salesforce.com
Magento (Adobe Commerce) - https://business.adobe.com/products/magento/magento-commerce.html
Shopware - https://www.shopware.com
Zapier - https://www.zapier.com
HP (Hewlett-Packard) - https://www.hp.com
Bridge Line Digital - https://www.bridgeline.com
AgentForce (Salesforce AI) - https://www.salesforce.com/products/einstein-ai/
About our Guest, Ari KhanPrior to joining Bridgeline, Dr. Kahn was the co-founder of FatWire, a leading content management and digital engagement company. As the General Manager and Chief Technology Officer of FatWire, Dr. Kahn built the company to a global corporation with offices in 13 countries and top industry analyst ratings. After winning many of the Global-2000 as customers and growing annual revenues to more than $40 Million, FatWire was acquired by Oracle in 2011 for $160 Million. After leading that
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(01:00:02):
(Music Playing) Ladies and gentlemen, welcome to the
E-commerce Masters Podcast. I'm your
host, Ethan Giffen, CEO of Groove
Commerce. And I'm excited to be here
today with a very, very awesome guest,
Ari Kahn, president and CEO of Bridgeline
Digital. They've got some really cool
things happening and that we want to talk
about in terms of search and AI that I
(01:00:24):
think you're just going to love. Before
we jump in to meet with Ari, first, I
want to remind you that we're a video
first podcast. So please, please, please,
for the best way to watch is to head on
over to our YouTube channel at Groove
Commerce and give us a like, give us a
share, leave us a comment or subscribe.
And we love, we love getting feedback
(01:00:44):
from you. And without further ado,
welcome to the podcast,
Ari. How are you today?
I'm great, Ethan. Thanks for
having me. Excited to be here.
That's awesome. That's awesome. Well, I'm
really honored to have you here. I know
we've had some great conversations over
the years. And it's cool to finally get
you here live on the podcast to talk
(01:01:06):
about search, AI, and all kinds of things
that are happening in our industry today.
So tell us a little bit about yourself.
How did you get into the e-commerce
industry is what most people
want to know about these days.
Yeah.
Well, my background is computer science.
I've got a PhD in artificial
(01:01:27):
intelligence. I build guidance systems
for the Tomahawk when I was in graduate
school. But after blowing stuff up for a
while, I decided that maybe being
constructive was the right path for life.
In 96, I started my first software
company, Fatwire. And Fatwire was one of
(01:01:48):
the first content management companies.
In fact, content management wasn't even a
term at the time. And we got lucky. Built
IBM's website, Motorola's website, just a
couple of kids in a garage practically.
Built that business
up. Sold it to Oracle.
And one of the things that I recognized
(01:02:11):
after selling that company and reflecting
upon it for a while was
at the end of the day,
CEOs and companies have two problems--
revenue and everything else.
If you can solve that revenue problem,
you can find a customer. And online
revenue is where it was at. And
e-commerce still had a long way to go. So
(01:02:34):
I came into Bridgeline. Actually, not the
founder. I became the largest shareholder
and ultimately the president and the CEO
with really a vision of helping companies
grow their online revenue by breaking
revenue into its three constituent
components. The first one for online
(01:02:57):
revenue is traffic. How many
people come to your website?
The second one is conversion. What
percentage of those visitors buy
something? And the third one is order
size. How much does each one spend? You
multiply those three components together,
that is your revenue. And MarTech,
marketing technology, is a saturated,
(01:03:19):
overpopulated space with 10,000 products
that most of them are trying to solve one
of those three problems. But none of them
really are bringing it all together and
looking at the broader end goal of how do
you bring these things in to solve that
revenue problem? So we started buying
(01:03:39):
some companies, companies that would fall
into the traffic bucket. We bought a
company called Wurank, for example.
Companies that fall into conversion. Hawk
Search being one of those. Companies that
fall into the order size. Hawk Search
actually solves that, Orchestra and
(01:04:01):
Unbound do as well as Celebros. And
bringing all of these products together
to be able to solve that ultimate problem
for customers. Then Hawk
Search became a runaway success.
And it started winning customers left and
right, driving tons of
value for those customers.
(01:04:22):
And all of a sudden, large language
models and all this great stuff with AI
matured overnight, fell right into the
sweet spot with our customers' need to
help their customers find the products
that they want. It started integrating
all of that into Hawk Search. And last
year alone, we released eight new
(01:04:42):
products. All large language model
AI-based products and are driving tons of
business for those customers, especially
in the B2B space, where there are so many
lagers that are able to leapfrog past all
of the feeling around in the dark stuff
that the B2C companies have kind of
(01:05:03):
figured out and get right to the point of
doing business now online.
That's great. That's great. Yeah, I mean,
it's been interesting to see the
portfolio evolve of all the different
products over the years and just see Hawk
just really start to grab hold and start
to push that forward. I think people--
in terms of--
(01:05:25):
one of the things that I think AI is
creating, especially with marketers, is
just a real understanding of how search
can impact every aspect of what you're
doing. And I think people thought about
SEO as bigger, different, maybe more
technical. But site search is one of the
(01:05:46):
key pieces that can have a direct impact
on conversion rate, on average AOV, all
of those things, right outside of the
box. Like, I mean, and--
yeah, so go ahead. I'm sorry.
Absolutely. Site search is your
salesperson. Your online store
(01:06:06):
salesperson is the search dialogue. More
than 70% of all online purchases go
through search. And search is not just a
one-way conduit anymore where people just
enter a keyword that describes what they
want and hope to get the right result
back. Search is actually interactive. So
(01:06:27):
with Hawk Search, we ask clarifying
questions. We provide summaries of
results. We anticipate what somebody is
typing and immediately start providing
results to them without making them spell
the whole thing out. We transform their
request, which might have typos or
(01:06:48):
improper language, into the right type of
search results. Search has to be very
smart. It has to be interactive. It has
to be an intelligent salesperson, or you
will lose business. Search is the conduit
between your product
catalog and your customer.
Well, you know, it's interesting because
(01:07:08):
machine learning has been around for
decades, right? Like, people have used
leverage machine learning within their
products within search. But it's just
literally the evolution of this long-form
language being able to be put into a
search box, understand its intent,
understand misspellings. It used to be,
well, maybe you could misspell a single
(01:07:29):
word, and you would be able to understand
that, or maybe have a synonym or some
other type of trigger there. But it's
that evolution of native language, long
queries, and being able to understand
intent and provide back recommended
results from that has been very, very
(01:07:50):
game-changing overall.
Yes, understanding that intent is a
fundamental change. It's a paradigm shift
in the way that people interact with the
site. So prior to the large language
model coming in to site search, you
search with keywords. And what that
means, fundamentally, is that you know
the product that you want to buy.
(01:08:12):
And OK, you might be pecking around at
the right keyword and stuff, but you know
what you want. The difference is now, you
don't need to know what product you want
to buy. You need to know the problem you
want to solve. I want to go hiking in the
Appalachian Trail. OK, great. Now the
site, rei.com, can come back with, you
know, here are some hiking shoes and
(01:08:33):
hiking poles. Oh, I didn't even know I
needed those things. Or ask clarifying
questions. When are you going hiking? Are
you going hiking in January? Are you
going hiking in July? All right, now
you're interacting with this, but you're
interacting by telling it what your goal
is, not what you want.
(01:08:54):
Well, you know, I talk to directors of
marketing. I talk to VPs of sales,
especially as we're kind of speaking to
manufacturers and distributors that are
maybe not as mature in terms of their
understanding of commerce and all the
available technology around that. And
we're helping them look and try to solve
(01:09:14):
problems. And they still have significant
amounts of zero return queries that come
out of their site search, right? And so
there's generally no silver bullet in
this business to increase somebody's
sales immediately. And I say there's two
(01:09:34):
times. One is if there's a bug in the
checkout. The second is if they don't
have their site search properly set up,
or if they don't have a proper site
search to begin with. And we'll find
people with five or six out of 10 of
their top queries still with zero results
or very minimal results coming back
there. And that's a significant amount of
(01:09:54):
revenue that is walking away at that
point if people can't find what they're
looking for. And it's not necessarily
their fault. They're not as educated.
They may not be as mature in their
understanding of that. And it's fixable.
But when we can come in and implement
that site search and determine what are
all of the components of that
(01:10:17):
to set that up and configure it properly,
it can have an immediate impact on
revenue. Because those people are three
to five times more likely to buy. Those
that engage with site search are three to
five x more likely to buy. I think that
their A of E is generally 25% to 50%
higher at times than a non-- than a
(01:10:38):
customer who didn't engage with site
search. And so there's almost immediate
opportunity for them to latch on to that.
And the reality is a lot of these
manufacturers and distributors, they're
experts at whatever they're manufacturing
and selling and so forth. They're not
experts at e-commerce. And they need a
partner because they don't know what they
(01:10:59):
don't know a lot of times. They don't
realize how critical site search is to
having a clean product catalog. The
catalog might look fine to a human eye.
But when you take a look at it from a
machine's perspective, it won't work. And
you're going to get those zero result
search pages, which are the kiss of
death. That's like handing your customer
(01:11:20):
over to your competitor because people
will bounce off of that site. And the
shopping habits are evolving very
quickly. And there's very little
tolerance for a zero result search to
come back. So you have to have clean
data. You have to have the right search.
It's got to be intelligent search. And
you also have to recognize that
especially the younger generation is
(01:11:40):
going to have
intent-based search and expectations
that they can just describe what they
want to do, not necessarily what they
want-- the specifics of what they want to
buy and get that back, even in the B2B
and manufacturing realm.
Yeah. I think it's important for-- and
usually we might be working with a VP of
(01:12:01):
marketing-- or I mean a VP of sales. We
might be working with a director of
marketing. Sometimes we're even engaged
with the CEO themselves of a 100-person
organization. And in the old days of
doing these types of website projects for
B2B and for distributors, they'll be
(01:12:21):
like, well, I have a built-in customer
base. They have to use this system
anyway. And so all we're going to do is
kind of take the out of the box theme,
put our logo in it, and call it a day and
deploy it. And I think that there's a
level of sophistication that people are
leaving money on the table by not kind of
thinking about how all of this fits
together and by making it easier and
(01:12:44):
helping your existing customers find your
products easier and maybe find other
products that they could leverage that it
will help increase your sales.
Yeah. That cross-sell, kitting, or
bundling, recommendations--
those are critical. Even if you've had a
(01:13:04):
customer, they've been there for 20
years. They know exactly what they want.
You still need to be able to present them
those other products that they might need
in addition to that, just so that they
don't have to come back the next day and
buy it, that they can take care all at
once. But it will increase your AOV.
It'll give you a tighter relationship
with those customers. And you can't take
(01:13:26):
your customers for granted. You've got to
give them a good experience.
Well, you have
customers that might have--
to push unique skew or model numbers,
even specific for one of their larger
trading partners, may say, hey, we need
to buy this based upon what's in our
identifier and our ERP on your system for
(01:13:49):
our people to find this. And so there's
all kinds of big business challenges to
go through. And good systems like Hawk
help you solve those challenges.
So who is-- when you think about the
world of commerce, what does your ideal
merchant look like? Is there
(01:14:10):
a size? Is there an industry?
What generally makes a
good fit for Hawk Search?
Our sweet spot, where we've really added
a ton of value, is probably the
mid-market distributor. So this is a
distributor that is $250 million to $4
(01:14:32):
billion in revenue. And they've got a
product catalog with between 2,000 and
10,000 products, each product having a
large number of attributes.
A lot of times, one of their pain points
is the quality of their own data. A lot
of times, they're also challenged with
(01:14:54):
helping their customers, like you
described, use their own product catalog
skew numbers and other things and be able
to make transformations for that. A lot
of times, they also have challenges with
respect to units of measure and doing
conversions. And we help those customers.
And we've seen a lot of demand in that
(01:15:15):
particular market. We do talk about-- and
we like to promote some of the large
businesses that we have to show where we
can scale. So as an example,
wepowerhp.com, the .com site. We do over
$1 million per hour in sales on that
site-- ink, laptops, houses, keyboards,
(01:15:39):
all sorts of stuff. So from a scaling
perspective, that's super helpful because
even a mid-market distributor, for
example, may have huge volume. They might
not have the HP volume, but they need to
know that they're partnering with a
company that can handle that level. We
also care a lot about providing--
(01:16:03):
I'll call it instantaneous gratification.
We want to be easy to implement very
quickly and to provide results very
quickly. So we like to see customers that
are able to be nimble as well and launch
in a matter of days and to be able to
show results right away and measurable
(01:16:24):
results, have great analytics so that
they can see exactly what the conversion
rates were before and after on Hawk, what
the order value was before and after and
so on and so forth. So that's another
important part that our
customers really value us for.
That's great. That's great. Now, in terms
of that's one use case, do you-- can your
(01:16:47):
system do good for a $25 or $50 or $100
million dollar e-tailer as well?
Yeah. We have a lot of those. And the
pricing works for those guys as well.
We see companies as small as $5 or $10
million dollars using Hawk Surge. We
actually see more demand at the larger
(01:17:09):
price, which is why that is top of mind
for me, or I mean the larger companies
than that. But I think that it's very
important that you create-- and this is
sort of the heart of where SAS really
took off, takes off, is to be able to
create a product that is easy to start
using right out of the gate. And we took
(01:17:31):
a lot of effort in making sure that we
could launch for our customers without
them having to kick off some giant and
expensive professional services project
or to rip all of their existing
infrastructure apart in order to install
something new. And that's where the
smaller companies on Shopify or whatever
are happy to use Hawk.
(01:17:52):
And so what's a general implementation
time for somebody that's thinking about
this, like that says, hey, I've got a
problem. I'm building a new site.
When should they be thinking about this?
And how long does it usually take to get
from day one to launch?
So we released a capability that we call
RapidUI. So the RapidUI feature uses a
(01:18:14):
framework called Handlebars. And it
allows companies to launch in just a
couple of days. So you've got the first
step is connecting your own product
catalog into Hawk Search. There are
multiple ways to do that, including
bare-bones CSV files with an FTP site if
you want to do it that way or Zapier or
(01:18:34):
direct connection to your PIM. All those
are available. And then the second half,
in addition to connecting your data, is
presenting your data. And that's where
the RapidUI comes in. And by just
injecting a couple of points of
JavaScript, you can have a landing page.
And you can have a product results page
and a search results page and so forth,
(01:18:56):
really right out of the box.
What platforms are you-- you mentioned
Shopify. So what platforms do you
generally work with? I know you work with
BigCommerce. I know
you work with Shopify.
BigCommerce is a great one. We think very
highly of that platform. Quite frankly,
we've seen very successful customers in
(01:19:17):
there. We just released our Catalyst
implementation, which is their latest
user interface. You're familiar with
Catalyst. And so we do
great business with BigCommerce.
Optimizely is another partner that,
especially in the B2B distributor space
that we've seen a lot of business with.
(01:19:41):
And we've got a partnership with them
that allows us to just point and click.
And boom, they can, right inside of their
regular, optimizely configured commerce
implementation, manage
Hawk Search directly.
Salesforce and Magento are also platforms
(01:20:01):
that we're seeing more and more of,
especially the Salesforce one. I'm a
little bit surprised, but we just updated
our AppExchange release with Salesforce.
So right there inside of the AppExchange,
you can get Hawk Search.
All right. How about Shopware? I know
they're a newer player
domestically here as well.
Yeah. They're newer player in the US, but
(01:20:23):
we're seeing a lot of demand for
Shopware. So they're doing a good job at
marketing. And I think that it's a small
team in the US, but they're showing up.
So yes, we're partners with them.
That's great. That's great. So how do
generally-- what should someone be
thinking about when they're considering
(01:20:44):
Site Search? What are some of the factors
when you're talking to a prospect that
says, hey, I'm going to talk to the CEO
of this thing and kick
the tires a little bit.
How are you reassuring them? What are you
telling them to think about and look for
as they're evaluating
Site Search products?
(01:21:06):
Yep. One of the important things that
they need to watch for is an ability to
measure results. You can't improve what
you can't measure. And you're not going
to know exactly how to configure Site
Search on day one. You're going to have
new capabilities, and you're going to
want to do experiments. We have that in
Hawk Search. But when I think about Site
(01:21:29):
Search, it's not about just returning a
result based on the search. It's about
merchandising. You need to stack the
deck. You've got certain products that
you need to promote at a higher level.
Your search itself should learn over time
for different personas and also change
the ranking in which results come. But
(01:21:50):
all of those capabilities to boost
products, to be able to promote different
products and spotlight products and so
forth, need to be measured and tested. So
definitely make sure that you have great
measurements coming out of your platform
so that you can confirm what you're
doing. And then also make sure that
you've got the resources internally to
(01:22:11):
experiment. You can't just plug Site
Search in and just think that you're
done. It is a platform upon which you can
change consumer behavior, promote
specific products, increase your average
order value, increase your conversion.
But it's not a black box that does it all
by itself. And if you see a Site Search
(01:22:32):
that just has an AI button and if I click
the AI button, I can just go hands-off.
That will not work. You need to be
involved with the configuration,
especially AI capabilities, and measure
the results of your commerce.
That's one of the things that I really
(01:22:52):
also talk to people about pretty
regularly is they'll set this thing up
when they do their replatforming and they
won't go look at it for two or three
years until they replatform again. And
it's just not like this thing is living
and breathing. And I'm not saying that it
needs to be somebody's full-time job, but
somebody needs to do care and feeding of
this and go in and look at the results
(01:23:14):
and upgrade the configurations and be
thinking about synonyms and direct
matches and all the things that go along
with kind of helping people to get where
they want to get to. If your website has
gift cards and people are searching for
gift certificates, how are you kind of
(01:23:35):
getting them to the place where they need
to go to with this? And that may not be
as important for B2B folks, but people
are definitely looking for things. Are
they looking for maybe last year's model
number? Are you creating new models every
12 to 24 months of upgraded products and
(01:23:55):
eliminating old products? Well, what
happens when people search for those
products and they don't find them in your
catalog anymore? There's a whole lot of
things that need to be kind of care and
fed with this for it to really, really
make money for you. But it doesn't need a
full-time person, but it definitely needs
a few hours a month of somebody looking
(01:24:16):
at their numbers and thinking about it
and how to make it work together.
And it's also important to,
I think most companies need a partner,
and you don't have to sign up for $20,000
a month or $10,000 a month or whatever.
But when you are partnering with an
expert that sees hundreds of different
(01:24:38):
commerce sites and has an intuition for
what's worked in some places and what
hasn't worked at others and what you
really should measure at this particular
point in time and focus upon, that advice
pays for itself in a huge way. You don't
need to increase your conversion by very
much to generate a lot of revenue. You
(01:25:00):
don't need to increase your average order
value by very much to generate a lot of
revenue. And you miss out if you just try
to go completely alone most of the time.
I mean, some companies have like a really
hardcore e-commerce team that has worked
at other places and seen a lot of
different things. But most of them are
their expertise is elsewhere and you
(01:25:22):
don't need a full timer. You just need a
sounding board and a partner who will
help you get from here to there.
What's your favorite new feature that
you've dropped in the
last 12 months or so?
Conversational search is definitely on
that list. So conversational search
really changes the way that our customers
(01:25:44):
or our customers' customers interact with
their site. It turns the search process
from being a one way, here's what I'm
looking for, to really a prescriptive
engagement where you start your search,
especially for a very large catalog. And
then in addition to getting results,
start getting feedback and clarifying
(01:26:04):
questions and it becomes a dialogue that
really helps customers, especially in the
B2B space with huge catalogs, find
exactly not just the one particular
product that they need, but the
peripheral products that they didn't
think about and bring those into the
chopping card as well.
I mean, that seems amazing, but it also
seems, sounds extremely
(01:26:25):
complicated to implement.
For our customers, it's not complicated,
for all the guys I got in the back room
that had to write it for us.
These large language models that we've
embraced, that AI is really seen mature,
(01:26:47):
are the foundation for the capabilities
like conversational search, where we
ingest not just the product catalog, but
also all of the peripheral marketing
material and can even automatically
digest competitor's marketing information
or the manufacturer's marketing
information. And all of that content ends
(01:27:09):
up informing the behind the scenes AI, so
that it can begin that dialogue, ask the
clarifying questions, understand the
actual real world use cases for the
products that it's recommending, so that
it can then go and recognize if you have
other products. And one of the great
things is that it does it all within the
(01:27:31):
guardrails of your particular catalog, so
that it doesn't end up making
recommendations for stuff you don't have,
enforcing that customer to then go to
another site, one of your competitors to
get what they might be looking for.
I'm really curious about that and want to
dig into that a little bit
more, because in terms of most,
(01:27:55):
like previously, the old world of site
search was, "Hey, here's a file, here's
either the XML sitemap file of all the
pages on our site, for maybe a content or
a community search." And then commerce
search was always just kind of based upon
a text file of product data and whatever,
it can only index what was in that
(01:28:17):
folder. So are you saying that HawkSearch
can start to identify other assets and
ingest that information
as well? Yes, exactly.
And the assets might not be like a
(01:28:37):
relational database, they could be PDF
files that include images and schematics.
And it's reading all of those
capabilities, training its behind the
scenes large language model, and then
using that to respond to you. A great
example would be imagine that you've got
(01:28:59):
a plumbing supply company, and there's
all sorts of documents as to how many
feet of pipe you need to do certain types
of residential rooms for bathrooms,
versus a living room, for example, and
being able to then provide HawkSearch the
floor plan and have it identify all of
(01:29:21):
the pipes and fittings and so forth that
are necessary for that. That is the type
of thing that we can do now with a large
language model. So we ingest all of those
specs, all of those best practices and
floor plans and so on and so forth, and
make real intelligent recommendations,
the kind of recommendations that you
(01:29:41):
would get from a salesperson
that is also an industry expert.
Interesting, very, very interesting. I
think that's going to be game changing
for folks, because we've often struggled
of how to combine, if somebody has a
large content based site, maybe a large
blog site or a large knowledge base, of
how to combine that into search results
(01:30:03):
with the commerce part
in terms of the catalog.
And so being able to kind of fit all of
that together has just been a struggle
over the years in order to kind of to get
there, because most of the product
searches just like focus specifically on
a specific catalog with the information
provided inside of it. So that's going to
(01:30:24):
be great to see the
evolution of that as we go.
What's kind of next? Where do you see
yourself evolving Hawk
over the next 12 to 24 months?
Agintic AI is going to become a
(01:30:45):
ubiquitous part of our shopping
experience. You're going to be able to
launch agents that perform certain types
of shopping and activities. On the
merchandiser side, you're going to have
agents that are monitoring behavior and
automatically updating your product
catalog and so forth. So we're really
(01:31:06):
embracing a Gentec right now and have
created what we call SmartBridge, which
is an interface that allows third party
agents. You might use Agent Force from
Salesforce, for example, to create an
agent that acts like part of your
merchandising team. And now that actor,
(01:31:27):
that Agent Force from Salesforce.com
actor can through a SmartBridge be able
to boost and bury and configure and do
all the types of things inside of Hawk
Search that a human might do through the
interface. In addition to that, we've got
our own agent builder that's coming out
and allow you to create agents inside of
(01:31:50):
Hawk Search itself. But I think that
you're going to see lots of platforms
that allow drag and drop creation of
agents, simple things like, hey Siri,
turn on the lights in my front yard. And
it does that for you, right? I just built
(01:32:10):
an agent when I told Siri those commands,
the same types of capabilities people are
going to expect to configure their
commerce website to do monitoring and to
even do shopping for them. So we're
embracing agents in a
big way at Bridgeline.
That's cool. That's really, really cool.
Which accessory? Kids'
(01:32:30):
bedrooms. Uh-oh. Mary!
They're everywhere. These agents.
The agents are
everywhere. Yeah, be careful.
Be careful. Oh my. Oh my. Hey Siri,
what's up? Yeah, so, you know, that's
fantastic. It's great to see kind of the
evolution. Now, like, do any of your
(01:32:52):
products within the Bridgeline kind of
portfolio, do they work together or are
they kind of all
stand alone with how they,
with their focus?
A fundamental part of our go-to-market
strategy is to have integration between
our different products. So from a
business perspective, software companies,
(01:33:14):
one of the largest challenges for any
company, but for software companies, is
the customer acquisition cost is called
CAC, which is really the amount of sales
and marketing spend necessary to sign a
new license. And what we're doing is a
fundamental part of that business
practice, but also because it adds value
(01:33:35):
to our customers, is we've got a common
interface for all of our products that is
able to recognize how you could further
improve your revenue, your online
revenue, by either adding additional
capabilities for traffic, capabilities
for conversion, or capabilities for
(01:33:56):
increasing your order value, and is able
to make recommendations at the
appropriate time for other Bridgeline
products that could help you in the area
that you're, where the low-hanging fruit
is. So all of our products have this
common interface and work together to
solve that one single problem. And we
(01:34:17):
grow both through our own innovation,
where we released several products last
year, eight products last year that we
developed in our own labs, but we also
grow through acquisition. And when we
acquire companies, we acquire them
because we think that they've got great
products, but we also acquire them
because we think that they have great
customer bases who would drive value from
(01:34:39):
our other products. So when we do an
acquisition, we put that same interface,
we call it the e-commerce 360 interface,
on top of those products and allow all of
their customers to then benefit from
being able to see where the rest of the
Bridgeline product line would be able to
(01:35:00):
further improve their online revenue, and
then also train that interface so that
all of the existing Bridgeline customers
are informed when the new product that we
have might help them.
I think, you know, and I find that so
intriguing because, you know, when I look
at your portfolio, I feel like, you know,
many of the products that you own or
(01:35:20):
maybe didn't get the right attention when
they were with their previous owners or
with their founders, and when you've been
able to help them, you know, help take
that on and kind of move them
significantly forward in terms of the
sophistication of what they do and just
even from a business development
standpoint of view, you know, overall.
(01:35:40):
Yeah. Well, you know, the Martech space
is overpopulated, frankly.
There's tons of products. A lot of them
are great. But in the SaaS world, you
need to inject capital upfront to win
customers because the payback period from
the initial customer win, the standard in
Martech is 20 months. So imagine that
(01:36:03):
you're a couple of smart entrepreneurs
who have created a great product. What's
the chances that your credit card is
thick enough to handle 20 months before
you actually break even on just your
sales and marketing expenses? Forget the
R&D. That's like another six months or
more. So a lot of these great products
(01:36:24):
just will not make it stand alone. And
there needs to be consolidation. And we
think that's an opportunity where we
speak to companies very frequently to
bring them together, have a strategy like
our e-commerce 360 strategy. And get
those products in front of the customers
that can really drive value from them. So
(01:36:44):
that's a core part of our business plan.
If you had one tip for a brand, for a
manufacturer, for a distributor, if you
had one tip for growth, you know, in
2025, what would you tell
folks to be thinking about?
You know,
(01:37:08):
your customers cannot
buy what they cannot find.
And yes, that comes back to Hawk Search,
but it also comes back to traffic. It
comes back to the shopping experience
overall, making sure that they find
things. So you should try to take a hard
look at yourself and put yourself in your
customer's shoes and just see if you're
(01:37:30):
the first thing that shows up when you're
looking for one of your products. Maybe
it's one of your competitors. Maybe you
don't show up at all and nobody does. But
take that objective look, make sure your
customers can find you and then find the
product that they want to buy from you.
Fantastic, fantastic
advice. Fantastic advice.
(01:37:52):
So if people want to find out more
information about Hawk Search or about
you, do they assume they go to
hawksearch.com or are
you active on LinkedIn?
Well, I am, thanks, Jeff. So I'm active
on LinkedIn. I'm not sure my LinkedIn ID
is, but Arikon at Bijlian, Arikon at Hawk
Search, I'm sure will find me. But
(01:38:13):
hawksearch.com is the
best way to learn about.
Bridgeline is the parent company, but
we're all in on Hawk Search. I mean, that
is really driving things. Keep on
thinking about, frankly, renaming the
company and all of our products are
sitting under the Hawk umbrella. So check
(01:38:33):
out hawksearch.com. You'll see lots of
really great information about what's
happening in artificial intelligence, how
to grow your online store. And of course,
where Hawk Search can help you get there.
That's awesome. That's awesome. If you if
you want more information, don't hesitate
to reach out to us here at Groove. We're
happy to help introduce you if you'd like
an introduction to the folks at Hawk.
We're happy to evaluate what you're
(01:38:55):
currently working with site search wise
and kind of give you some tips on whether
we think it's doing well or not. So
please feel free to reach out to us.
Again, as I mentioned, please, you know,
subscribe to our podcast and leave a
comment or review wherever you find it.
And we're video first. So head on over to
our YouTube channel and give us a view
(01:39:17):
there. Thank you again, Ari. I really,
really appreciate your time today. Very,
very insightful conversation just about
AI and site search and all those things.
So thank you very much for for joining us
today. Well, thank you. It's really been
my pleasure. Awesome, awesome. And to
everybody else, we'll see
you on the next episode.
(Music Playing) Ladies and gentlemen, welcome to the
E-commerce Masters Podcast. I'm your
host, Ethan Giffen, CEO of Groove
Commerce. And I'm excited to be here
today with a very, very awesome guest,
Ari Kahn, president and CEO of Bridgeline
Digital. They've got some really cool
things happening and that we want to talk
about in terms of search and AI that I
(01:00:24):
think you're just going to love. Before
we jump in to meet with Ari, first, I
want to remind you that we're a video
first podcast. So please, please, please,
for the best way to watch is to head on
over to our YouTube channel at Groove
Commerce and give us a like, give us a
share, leave us a comment or subscribe.
And we love, we love getting feedback
(01:00:44):
from you. And without further ado,
welcome to the podcast,
Ari. How are you today?
I'm great, Ethan. Thanks for
having me. Excited to be here.
That's awesome. That's awesome. Well, I'm
really honored to have you here. I know
we've had some great conversations over
the years. And it's cool to finally get
you here live on the podcast to talk
(01:01:06):
about search, AI, and all kinds of things
that are happening in our industry today.
So tell us a little bit about yourself.
How did you get into the e-commerce
industry is what most people
want to know about these days.
Yeah.
Well, my background is computer science.
I've got a PhD in artificial
(01:01:27):
intelligence. I build guidance systems
for the Tomahawk when I was in graduate
school. But after blowing stuff up for a
while, I decided that maybe being
constructive was the right path for life.
In 96, I started my first software
company, Fatwire. And Fatwire was one of
(01:01:48):
the first content management companies.
In fact, content management wasn't even a
term at the time. And we got lucky. Built
IBM's website, Motorola's website, just a
couple of kids in a garage practically.
Built that business
up. Sold it to Oracle.
And one of the things that I recognized
(01:02:11):
after selling that company and reflecting
upon it for a while was
at the end of the day,
CEOs and companies have two problems--
revenue and everything else.
If you can solve that revenue problem,
you can find a customer. And online
revenue is where it was at. And
e-commerce still had a long way to go. So
(01:02:34):
I came into Bridgeline. Actually, not the
founder. I became the largest shareholder
and ultimately the president and the CEO
with really a vision of helping companies
grow their online revenue by breaking
revenue into its three constituent
components. The first one for online
(01:02:57):
revenue is traffic. How many
people come to your website?
The second one is conversion. What
percentage of those visitors buy
something? And the third one is order
size. How much does each one spend? You
multiply those three components together,
that is your revenue. And MarTech,
marketing technology, is a saturated,
(01:03:19):
overpopulated space with 10,000 products
that most of them are trying to solve one
of those three problems. But none of them
really are bringing it all together and
looking at the broader end goal of how do
you bring these things in to solve that
revenue problem? So we started buying
(01:03:39):
some companies, companies that would fall
into the traffic bucket. We bought a
company called Wurank, for example.
Companies that fall into conversion. Hawk
Search being one of those. Companies that
fall into the order size. Hawk Search
actually solves that, Orchestra and
(01:04:01):
Unbound do as well as Celebros. And
bringing all of these products together
to be able to solve that ultimate problem
for customers. Then Hawk
Search became a runaway success.
And it started winning customers left and
right, driving tons of
value for those customers.
(01:04:22):
And all of a sudden, large language
models and all this great stuff with AI
matured overnight, fell right into the
sweet spot with our customers' need to
help their customers find the products
that they want. It started integrating
all of that into Hawk Search. And last
year alone, we released eight new
(01:04:42):
products. All large language model
AI-based products and are driving tons of
business for those customers, especially
in the B2B space, where there are so many
lagers that are able to leapfrog past all
of the feeling around in the dark stuff
that the B2C companies have kind of
(01:05:03):
figured out and get right to the point of
doing business now online.
That's great. That's great. Yeah, I mean,
it's been interesting to see the
portfolio evolve of all the different
products over the years and just see Hawk
just really start to grab hold and start
to push that forward. I think people--
in terms of--
(01:05:25):
one of the things that I think AI is
creating, especially with marketers, is
just a real understanding of how search
can impact every aspect of what you're
doing. And I think people thought about
SEO as bigger, different, maybe more
technical. But site search is one of the
(01:05:46):
key pieces that can have a direct impact
on conversion rate, on average AOV, all
of those things, right outside of the
box. Like, I mean, and--
yeah, so go ahead. I'm sorry.
Absolutely. Site search is your
salesperson. Your online store
(01:06:06):
salesperson is the search dialogue. More
than 70% of all online purchases go
through search. And search is not just a
one-way conduit anymore where people just
enter a keyword that describes what they
want and hope to get the right result
back. Search is actually interactive. So
(01:06:27):
with Hawk Search, we ask clarifying
questions. We provide summaries of
results. We anticipate what somebody is
typing and immediately start providing
results to them without making them spell
the whole thing out. We transform their
request, which might have typos or
(01:06:48):
improper language, into the right type of
search results. Search has to be very
smart. It has to be interactive. It has
to be an intelligent salesperson, or you
will lose business. Search is the conduit
between your product
catalog and your customer.
Well, you know, it's interesting because
(01:07:08):
machine learning has been around for
decades, right? Like, people have used
leverage machine learning within their
products within search. But it's just
literally the evolution of this long-form
language being able to be put into a
search box, understand its intent,
understand misspellings. It used to be,
well, maybe you could misspell a single
(01:07:29):
word, and you would be able to understand
that, or maybe have a synonym or some
other type of trigger there. But it's
that evolution of native language, long
queries, and being able to understand
intent and provide back recommended
results from that has been very, very
(01:07:50):
game-changing overall.
Yes, understanding that intent is a
fundamental change. It's a paradigm shift
in the way that people interact with the
site. So prior to the large language
model coming in to site search, you
search with keywords. And what that
means, fundamentally, is that you know
the product that you want to buy.
(01:08:12):
And OK, you might be pecking around at
the right keyword and stuff, but you know
what you want. The difference is now, you
don't need to know what product you want
to buy. You need to know the problem you
want to solve. I want to go hiking in the
Appalachian Trail. OK, great. Now the
site, rei.com, can come back with, you
know, here are some hiking shoes and
(01:08:33):
hiking poles. Oh, I didn't even know I
needed those things. Or ask clarifying
questions. When are you going hiking? Are
you going hiking in January? Are you
going hiking in July? All right, now
you're interacting with this, but you're
interacting by telling it what your goal
is, not what you want.
(01:08:54):
Well, you know, I talk to directors of
marketing. I talk to VPs of sales,
especially as we're kind of speaking to
manufacturers and distributors that are
maybe not as mature in terms of their
understanding of commerce and all the
available technology around that. And
we're helping them look and try to solve
(01:09:14):
problems. And they still have significant
amounts of zero return queries that come
out of their site search, right? And so
there's generally no silver bullet in
this business to increase somebody's
sales immediately. And I say there's two
(01:09:34):
times. One is if there's a bug in the
checkout. The second is if they don't
have their site search properly set up,
or if they don't have a proper site
search to begin with. And we'll find
people with five or six out of 10 of
their top queries still with zero results
or very minimal results coming back
there. And that's a significant amount of
(01:09:54):
revenue that is walking away at that
point if people can't find what they're
looking for. And it's not necessarily
their fault. They're not as educated.
They may not be as mature in their
understanding of that. And it's fixable.
But when we can come in and implement
that site search and determine what are
all of the components of that
(01:10:17):
to set that up and configure it properly,
it can have an immediate impact on
revenue. Because those people are three
to five times more likely to buy. Those
that engage with site search are three to
five x more likely to buy. I think that
their A of E is generally 25% to 50%
higher at times than a non-- than a
(01:10:38):
customer who didn't engage with site
search. And so there's almost immediate
opportunity for them to latch on to that.
And the reality is a lot of these
manufacturers and distributors, they're
experts at whatever they're manufacturing
and selling and so forth. They're not
experts at e-commerce. And they need a
partner because they don't know what they
(01:10:59):
don't know a lot of times. They don't
realize how critical site search is to
having a clean product catalog. The
catalog might look fine to a human eye.
But when you take a look at it from a
machine's perspective, it won't work. And
you're going to get those zero result
search pages, which are the kiss of
death. That's like handing your customer
(01:11:20):
over to your competitor because people
will bounce off of that site. And the
shopping habits are evolving very
quickly. And there's very little
tolerance for a zero result search to
come back. So you have to have clean
data. You have to have the right search.
It's got to be intelligent search. And
you also have to recognize that
especially the younger generation is
(01:11:40):
going to have
intent-based search and expectations
that they can just describe what they
want to do, not necessarily what they
want-- the specifics of what they want to
buy and get that back, even in the B2B
and manufacturing realm.
Yeah. I think it's important for-- and
usually we might be working with a VP of
(01:12:01):
marketing-- or I mean a VP of sales. We
might be working with a director of
marketing. Sometimes we're even engaged
with the CEO themselves of a 100-person
organization. And in the old days of
doing these types of website projects for
B2B and for distributors, they'll be
(01:12:21):
like, well, I have a built-in customer
base. They have to use this system
anyway. And so all we're going to do is
kind of take the out of the box theme,
put our logo in it, and call it a day and
deploy it. And I think that there's a
level of sophistication that people are
leaving money on the table by not kind of
thinking about how all of this fits
together and by making it easier and
(01:12:44):
helping your existing customers find your
products easier and maybe find other
products that they could leverage that it
will help increase your sales.
Yeah. That cross-sell, kitting, or
bundling, recommendations--
those are critical. Even if you've had a
(01:13:04):
customer, they've been there for 20
years. They know exactly what they want.
You still need to be able to present them
those other products that they might need
in addition to that, just so that they
don't have to come back the next day and
buy it, that they can take care all at
once. But it will increase your AOV.
It'll give you a tighter relationship
with those customers. And you can't take
(01:13:26):
your customers for granted. You've got to
give them a good experience.
Well, you have
customers that might have--
to push unique skew or model numbers,
even specific for one of their larger
trading partners, may say, hey, we need
to buy this based upon what's in our
identifier and our ERP on your system for
(01:13:49):
our people to find this. And so there's
all kinds of big business challenges to
go through. And good systems like Hawk
help you solve those challenges.
So who is-- when you think about the
world of commerce, what does your ideal
merchant look like? Is there
(01:14:10):
a size? Is there an industry?
What generally makes a
good fit for Hawk Search?
Our sweet spot, where we've really added
a ton of value, is probably the
mid-market distributor. So this is a
distributor that is $250 million to $4
(01:14:32):
billion in revenue. And they've got a
product catalog with between 2,000 and
10,000 products, each product having a
large number of attributes.
A lot of times, one of their pain points
is the quality of their own data. A lot
of times, they're also challenged with
(01:14:54):
helping their customers, like you
described, use their own product catalog
skew numbers and other things and be able
to make transformations for that. A lot
of times, they also have challenges with
respect to units of measure and doing
conversions. And we help those customers.
And we've seen a lot of demand in that
(01:15:15):
particular market. We do talk about-- and
we like to promote some of the large
businesses that we have to show where we
can scale. So as an example,
wepowerhp.com, the .com site. We do over
$1 million per hour in sales on that
site-- ink, laptops, houses, keyboards,
(01:15:39):
all sorts of stuff. So from a scaling
perspective, that's super helpful because
even a mid-market distributor, for
example, may have huge volume. They might
not have the HP volume, but they need to
know that they're partnering with a
company that can handle that level. We
also care a lot about providing--
(01:16:03):
I'll call it instantaneous gratification.
We want to be easy to implement very
quickly and to provide results very
quickly. So we like to see customers that
are able to be nimble as well and launch
in a matter of days and to be able to
show results right away and measurable
(01:16:24):
results, have great analytics so that
they can see exactly what the conversion
rates were before and after on Hawk, what
the order value was before and after and
so on and so forth. So that's another
important part that our
customers really value us for.
That's great. That's great. Now, in terms
of that's one use case, do you-- can your
(01:16:47):
system do good for a $25 or $50 or $100
million dollar e-tailer as well?
Yeah. We have a lot of those. And the
pricing works for those guys as well.
We see companies as small as $5 or $10
million dollars using Hawk Surge. We
actually see more demand at the larger
(01:17:09):
price, which is why that is top of mind
for me, or I mean the larger companies
than that. But I think that it's very
important that you create-- and this is
sort of the heart of where SAS really
took off, takes off, is to be able to
create a product that is easy to start
using right out of the gate. And we took
(01:17:31):
a lot of effort in making sure that we
could launch for our customers without
them having to kick off some giant and
expensive professional services project
or to rip all of their existing
infrastructure apart in order to install
something new. And that's where the
smaller companies on Shopify or whatever
are happy to use Hawk.
(01:17:52):
And so what's a general implementation
time for somebody that's thinking about
this, like that says, hey, I've got a
problem. I'm building a new site.
When should they be thinking about this?
And how long does it usually take to get
from day one to launch?
So we released a capability that we call
RapidUI. So the RapidUI feature uses a
(01:18:14):
framework called Handlebars. And it
allows companies to launch in just a
couple of days. So you've got the first
step is connecting your own product
catalog into Hawk Search. There are
multiple ways to do that, including
bare-bones CSV files with an FTP site if
you want to do it that way or Zapier or
(01:18:34):
direct connection to your PIM. All those
are available. And then the second half,
in addition to connecting your data, is
presenting your data. And that's where
the RapidUI comes in. And by just
injecting a couple of points of
JavaScript, you can have a landing page.
And you can have a product results page
and a search results page and so forth,
(01:18:56):
really right out of the box.
What platforms are you-- you mentioned
Shopify. So what platforms do you
generally work with? I know you work with
BigCommerce. I know
you work with Shopify.
BigCommerce is a great one. We think very
highly of that platform. Quite frankly,
we've seen very successful customers in
(01:19:17):
there. We just released our Catalyst
implementation, which is their latest
user interface. You're familiar with
Catalyst. And so we do
great business with BigCommerce.
Optimizely is another partner that,
especially in the B2B distributor space
that we've seen a lot of business with.
(01:19:41):
And we've got a partnership with them
that allows us to just point and click.
And boom, they can, right inside of their
regular, optimizely configured commerce
implementation, manage
Hawk Search directly.
Salesforce and Magento are also platforms
(01:20:01):
that we're seeing more and more of,
especially the Salesforce one. I'm a
little bit surprised, but we just updated
our AppExchange release with Salesforce.
So right there inside of the AppExchange,
you can get Hawk Search.
All right. How about Shopware? I know
they're a newer player
domestically here as well.
Yeah. They're newer player in the US, but
(01:20:23):
we're seeing a lot of demand for
Shopware. So they're doing a good job at
marketing. And I think that it's a small
team in the US, but they're showing up.
So yes, we're partners with them.
That's great. That's great. So how do
generally-- what should someone be
thinking about when they're considering
(01:20:44):
Site Search? What are some of the factors
when you're talking to a prospect that
says, hey, I'm going to talk to the CEO
of this thing and kick
the tires a little bit.
How are you reassuring them? What are you
telling them to think about and look for
as they're evaluating
Site Search products?
(01:21:06):
Yep. One of the important things that
they need to watch for is an ability to
measure results. You can't improve what
you can't measure. And you're not going
to know exactly how to configure Site
Search on day one. You're going to have
new capabilities, and you're going to
want to do experiments. We have that in
Hawk Search. But when I think about Site
(01:21:29):
Search, it's not about just returning a
result based on the search. It's about
merchandising. You need to stack the
deck. You've got certain products that
you need to promote at a higher level.
Your search itself should learn over time
for different personas and also change
the ranking in which results come. But
(01:21:50):
all of those capabilities to boost
products, to be able to promote different
products and spotlight products and so
forth, need to be measured and tested. So
definitely make sure that you have great
measurements coming out of your platform
so that you can confirm what you're
doing. And then also make sure that
you've got the resources internally to
(01:22:11):
experiment. You can't just plug Site
Search in and just think that you're
done. It is a platform upon which you can
change consumer behavior, promote
specific products, increase your average
order value, increase your conversion.
But it's not a black box that does it all
by itself. And if you see a Site Search
(01:22:32):
that just has an AI button and if I click
the AI button, I can just go hands-off.
That will not work. You need to be
involved with the configuration,
especially AI capabilities, and measure
the results of your commerce.
That's one of the things that I really
(01:22:52):
also talk to people about pretty
regularly is they'll set this thing up
when they do their replatforming and they
won't go look at it for two or three
years until they replatform again. And
it's just not like this thing is living
and breathing. And I'm not saying that it
needs to be somebody's full-time job, but
somebody needs to do care and feeding of
this and go in and look at the results
(01:23:14):
and upgrade the configurations and be
thinking about synonyms and direct
matches and all the things that go along
with kind of helping people to get where
they want to get to. If your website has
gift cards and people are searching for
gift certificates, how are you kind of
(01:23:35):
getting them to the place where they need
to go to with this? And that may not be
as important for B2B folks, but people
are definitely looking for things. Are
they looking for maybe last year's model
number? Are you creating new models every
12 to 24 months of upgraded products and
(01:23:55):
eliminating old products? Well, what
happens when people search for those
products and they don't find them in your
catalog anymore? There's a whole lot of
things that need to be kind of care and
fed with this for it to really, really
make money for you. But it doesn't need a
full-time person, but it definitely needs
a few hours a month of somebody looking
(01:24:16):
at their numbers and thinking about it
and how to make it work together.
And it's also important to,
I think most companies need a partner,
and you don't have to sign up for $20,000
a month or $10,000 a month or whatever.
But when you are partnering with an
expert that sees hundreds of different
(01:24:38):
commerce sites and has an intuition for
what's worked in some places and what
hasn't worked at others and what you
really should measure at this particular
point in time and focus upon, that advice
pays for itself in a huge way. You don't
need to increase your conversion by very
much to generate a lot of revenue. You
(01:25:00):
don't need to increase your average order
value by very much to generate a lot of
revenue. And you miss out if you just try
to go completely alone most of the time.
I mean, some companies have like a really
hardcore e-commerce team that has worked
at other places and seen a lot of
different things. But most of them are
their expertise is elsewhere and you
(01:25:22):
don't need a full timer. You just need a
sounding board and a partner who will
help you get from here to there.
What's your favorite new feature that
you've dropped in the
last 12 months or so?
Conversational search is definitely on
that list. So conversational search
really changes the way that our customers
(01:25:44):
or our customers' customers interact with
their site. It turns the search process
from being a one way, here's what I'm
looking for, to really a prescriptive
engagement where you start your search,
especially for a very large catalog. And
then in addition to getting results,
start getting feedback and clarifying
(01:26:04):
questions and it becomes a dialogue that
really helps customers, especially in the
B2B space with huge catalogs, find
exactly not just the one particular
product that they need, but the
peripheral products that they didn't
think about and bring those into the
chopping card as well.
I mean, that seems amazing, but it also
seems, sounds extremely
(01:26:25):
complicated to implement.
For our customers, it's not complicated,
for all the guys I got in the back room
that had to write it for us.
These large language models that we've
embraced, that AI is really seen mature,
(01:26:47):
are the foundation for the capabilities
like conversational search, where we
ingest not just the product catalog, but
also all of the peripheral marketing
material and can even automatically
digest competitor's marketing information
or the manufacturer's marketing
information. And all of that content ends
(01:27:09):
up informing the behind the scenes AI, so
that it can begin that dialogue, ask the
clarifying questions, understand the
actual real world use cases for the
products that it's recommending, so that
it can then go and recognize if you have
other products. And one of the great
things is that it does it all within the
(01:27:31):
guardrails of your particular catalog, so
that it doesn't end up making
recommendations for stuff you don't have,
enforcing that customer to then go to
another site, one of your competitors to
get what they might be looking for.
I'm really curious about that and want to
dig into that a little bit
more, because in terms of most,
(01:27:55):
like previously, the old world of site
search was, "Hey, here's a file, here's
either the XML sitemap file of all the
pages on our site, for maybe a content or
a community search." And then commerce
search was always just kind of based upon
a text file of product data and whatever,
it can only index what was in that
(01:28:17):
folder. So are you saying that HawkSearch
can start to identify other assets and
ingest that information
as well? Yes, exactly.
And the assets might not be like a
(01:28:37):
relational database, they could be PDF
files that include images and schematics.
And it's reading all of those
capabilities, training its behind the
scenes large language model, and then
using that to respond to you. A great
example would be imagine that you've got
(01:28:59):
a plumbing supply company, and there's
all sorts of documents as to how many
feet of pipe you need to do certain types
of residential rooms for bathrooms,
versus a living room, for example, and
being able to then provide HawkSearch the
floor plan and have it identify all of
(01:29:21):
the pipes and fittings and so forth that
are necessary for that. That is the type
of thing that we can do now with a large
language model. So we ingest all of those
specs, all of those best practices and
floor plans and so on and so forth, and
make real intelligent recommendations,
the kind of recommendations that you
(01:29:41):
would get from a salesperson
that is also an industry expert.
Interesting, very, very interesting. I
think that's going to be game changing
for folks, because we've often struggled
of how to combine, if somebody has a
large content based site, maybe a large
blog site or a large knowledge base, of
how to combine that into search results
(01:30:03):
with the commerce part
in terms of the catalog.
And so being able to kind of fit all of
that together has just been a struggle
over the years in order to kind of to get
there, because most of the product
searches just like focus specifically on
a specific catalog with the information
provided inside of it. So that's going to
(01:30:24):
be great to see the
evolution of that as we go.
What's kind of next? Where do you see
yourself evolving Hawk
over the next 12 to 24 months?
Agintic AI is going to become a
(01:30:45):
ubiquitous part of our shopping
experience. You're going to be able to
launch agents that perform certain types
of shopping and activities. On the
merchandiser side, you're going to have
agents that are monitoring behavior and
automatically updating your product
catalog and so forth. So we're really
(01:31:06):
embracing a Gentec right now and have
created what we call SmartBridge, which
is an interface that allows third party
agents. You might use Agent Force from
Salesforce, for example, to create an
agent that acts like part of your
merchandising team. And now that actor,
(01:31:27):
that Agent Force from Salesforce.com
actor can through a SmartBridge be able
to boost and bury and configure and do
all the types of things inside of Hawk
Search that a human might do through the
interface. In addition to that, we've got
our own agent builder that's coming out
and allow you to create agents inside of
(01:31:50):
Hawk Search itself. But I think that
you're going to see lots of platforms
that allow drag and drop creation of
agents, simple things like, hey Siri,
turn on the lights in my front yard. And
it does that for you, right? I just built
(01:32:10):
an agent when I told Siri those commands,
the same types of capabilities people are
going to expect to configure their
commerce website to do monitoring and to
even do shopping for them. So we're
embracing agents in a
big way at Bridgeline.
That's cool. That's really, really cool.
Which accessory? Kids'
(01:32:30):
bedrooms. Uh-oh. Mary!
They're everywhere. These agents.
The agents are
everywhere. Yeah, be careful.
Be careful. Oh my. Oh my. Hey Siri,
what's up? Yeah, so, you know, that's
fantastic. It's great to see kind of the
evolution. Now, like, do any of your
(01:32:52):
products within the Bridgeline kind of
portfolio, do they work together or are
they kind of all
stand alone with how they,
with their focus?
A fundamental part of our go-to-market
strategy is to have integration between
our different products. So from a
business perspective, software companies,
(01:33:14):
one of the largest challenges for any
company, but for software companies, is
the customer acquisition cost is called
CAC, which is really the amount of sales
and marketing spend necessary to sign a
new license. And what we're doing is a
fundamental part of that business
practice, but also because it adds value
(01:33:35):
to our customers, is we've got a common
interface for all of our products that is
able to recognize how you could further
improve your revenue, your online
revenue, by either adding additional
capabilities for traffic, capabilities
for conversion, or capabilities for
(01:33:56):
increasing your order value, and is able
to make recommendations at the
appropriate time for other Bridgeline
products that could help you in the area
that you're, where the low-hanging fruit
is. So all of our products have this
common interface and work together to
solve that one single problem. And we
(01:34:17):
grow both through our own innovation,
where we released several products last
year, eight products last year that we
developed in our own labs, but we also
grow through acquisition. And when we
acquire companies, we acquire them
because we think that they've got great
products, but we also acquire them
because we think that they have great
customer bases who would drive value from
(01:34:39):
our other products. So when we do an
acquisition, we put that same interface,
we call it the e-commerce 360 interface,
on top of those products and allow all of
their customers to then benefit from
being able to see where the rest of the
Bridgeline product line would be able to
(01:35:00):
further improve their online revenue, and
then also train that interface so that
all of the existing Bridgeline customers
are informed when the new product that we
have might help them.
I think, you know, and I find that so
intriguing because, you know, when I look
at your portfolio, I feel like, you know,
many of the products that you own or
(01:35:20):
maybe didn't get the right attention when
they were with their previous owners or
with their founders, and when you've been
able to help them, you know, help take
that on and kind of move them
significantly forward in terms of the
sophistication of what they do and just
even from a business development
standpoint of view, you know, overall.
(01:35:40):
Yeah. Well, you know, the Martech space
is overpopulated, frankly.
There's tons of products. A lot of them
are great. But in the SaaS world, you
need to inject capital upfront to win
customers because the payback period from
the initial customer win, the standard in
Martech is 20 months. So imagine that
(01:36:03):
you're a couple of smart entrepreneurs
who have created a great product. What's
the chances that your credit card is
thick enough to handle 20 months before
you actually break even on just your
sales and marketing expenses? Forget the
R&D. That's like another six months or
more. So a lot of these great products
(01:36:24):
just will not make it stand alone. And
there needs to be consolidation. And we
think that's an opportunity where we
speak to companies very frequently to
bring them together, have a strategy like
our e-commerce 360 strategy. And get
those products in front of the customers
that can really drive value from them. So
(01:36:44):
that's a core part of our business plan.
If you had one tip for a brand, for a
manufacturer, for a distributor, if you
had one tip for growth, you know, in
2025, what would you tell
folks to be thinking about?
You know,
(01:37:08):
your customers cannot
buy what they cannot find.
And yes, that comes back to Hawk Search,
but it also comes back to traffic. It
comes back to the shopping experience
overall, making sure that they find
things. So you should try to take a hard
look at yourself and put yourself in your
customer's shoes and just see if you're
(01:37:30):
the first thing that shows up when you're
looking for one of your products. Maybe
it's one of your competitors. Maybe you
don't show up at all and nobody does. But
take that objective look, make sure your
customers can find you and then find the
product that they want to buy from you.
Fantastic, fantastic
advice. Fantastic advice.
(01:37:52):
So if people want to find out more
information about Hawk Search or about
you, do they assume they go to
hawksearch.com or are
you active on LinkedIn?
Well, I am, thanks, Jeff. So I'm active
on LinkedIn. I'm not sure my LinkedIn ID
is, but Arikon at Bijlian, Arikon at Hawk
Search, I'm sure will find me. But
(01:38:13):
hawksearch.com is the
best way to learn about.
Bridgeline is the parent company, but
we're all in on Hawk Search. I mean, that
is really driving things. Keep on
thinking about, frankly, renaming the
company and all of our products are
sitting under the Hawk umbrella. So check
(01:38:33):
out hawksearch.com. You'll see lots of
really great information about what's
happening in artificial intelligence, how
to grow your online store. And of course,
where Hawk Search can help you get there.
That's awesome. That's awesome. If you if
you want more information, don't hesitate
to reach out to us here at Groove. We're
happy to help introduce you if you'd like
an introduction to the folks at Hawk.
We're happy to evaluate what you're
(01:38:55):
currently working with site search wise
and kind of give you some tips on whether
we think it's doing well or not. So
please feel free to reach out to us.
Again, as I mentioned, please, you know,
subscribe to our podcast and leave a
comment or review wherever you find it.
And we're video first. So head on over to
our YouTube channel and give us a view
(01:39:17):
there. Thank you again, Ari. I really,
really appreciate your time today. Very,
very insightful conversation just about
AI and site search and all those things.
So thank you very much for for joining us
today. Well, thank you. It's really been
my pleasure. Awesome, awesome. And to
everybody else, we'll see
you on the next episode.
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
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.