This is a recent recording from my appearance on Jordan West's podcast. There’s been a lot of buzz lately about the incrementally of YouTube after Haus Analytics revealed a groundbreaking report from 190 incrementally tests involving YouTube (link: https://www.haus.io/blog/do-youtube-ads-perform-lessons-from-190-incrementality-tests).
On this episode I share eye-opening case studies demonstrating how YouTube drives significant incremental growth - even when traditional attribution models fail to capture it. If you've been skeptical about YouTube's effectiveness or struggled to measure its impact, this conversation reveals why it might be the missing piece in your marketing strategy and how to approach it correctly.
Key Takeaways:
-Beyond Last-Click Attribution: Discover why obsessing over ROAS metrics can be misleading, and how YouTube drives significant business impact that often isn't captured in traditional attribution models but shows up dramatically in incrementality testing.
-The Real-World Proof: Brett shares compelling case studies, including a brand that saw Amazon sales cut in half when YouTube ads were paused, and another that achieved a 12-25% lift in Walmart store sales through targeted YouTube campaigns.
-The Anatomy of High-Converting YouTube Ads: Learn the essential elements of effective YouTube ads, including how to craft the perfect hook in the first 5 seconds, the importance of standard and "over-the-top" product demonstrations, and why 45-90 second videos often outperform shorter cuts.
-Production Value Matters: Unlike TikTok, YouTube requires slightly higher production quality (though not Super Bowl commercial level) - Brett explains the right balance and why combining polished footage with authentic UGC can be particularly powerful.
-Testing & Measurement Strategies: Get practical advice on how to properly measure YouTube's impact through geo tests, holdouts, search lift studies, and why tracking your overall MER (Media Efficiency Ratio) might be the simplest yet most effective approach.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
This is like TV advertisingfor our generation,
but way better than TVadvertising could ever be.
Well, hello and welcome to another editionof the E-Commerce Evolution podcast.
(00:24):
I'm your host, BrettCurry, CEO of OMG Commerce,
and today is a unique episode.
This is actually a recording that I did.
I was a guest on my buddy Jordan West's
podcast and we dove into YouTube. Now,
Jordan West is a pro. He isa master of TikTok shops.
(00:46):
He's built multiple brands,
and his podcast is the secretsto scaling your E-Commerce brand.
And we talked all about YouTube. Now,
the reason we dove intothis episode right now,
this is coming off the heels ofreally a groundbreaking research
piece or research project thatwas released by House Analytics,
HAUS.
(01:08):
Olivia Corey and team there have donesome brilliant work all around the
idea and the practice of measuringincrementality. We're seeing the true
impact of channels. And so they tested,
they did 190 incrementalitytests that involved YouTube with
74 brands, and theresults were mind blowing.
(01:29):
She was on a podcast with Andrew FerrisAlbe on that podcast soon as well.
But the idea here is that YouTube severelyunderreport or Google Ads severely
underreport, the actual impact theYouTube ads has to the tune of about
70% underreporting. Oranother way to look at that.
The reverse is it's about 3.42 x more
(01:53):
incremental than what you see there.
So if you're seeing a one rowas an example in platform,
the actual impact is probablya 3.42 x ROAS or 3.42 roas,
which is amazing. And so in this podcast,
Jordan asked me about ourexperiences as an agency.
We're known as one of the topYouTube ad agencies for D two C
brands. So tell lots of stories, lotsof examples on the creative side,
(02:15):
on the measurement side ways we'veused YouTube to grow retail sales,
omnichannel sales, I think it'sgoing to be informative, instructive,
and hopefully fun. So with that,
please enjoy the interviewthat I did with Jordan West
YouTube ads.
(02:36):
Hey guys, Jordan West back here Today.
I am so excited aboutthis podcast episode.
I have one of the experts in thespace in the YouTube space that is
Brett Curry on the podcast for I believea second time, maybe even third time.
Brett, welcome back to the show.
Yeah, Jordan Min awesome to be here. Itis for sure, at least the second time.
(02:57):
And maybe the third, I'm not sure,but let's make this the best one yet.
That'll be our goal.
I was so happy when you reached showedto me, I'm like, yes, I want to.
I was really hoping we were goingto talk about YouTube. I just said,
Brett Ray showed. I'm like, Iwill have him on the podcast.
I have no doubt that there'sgoing to be tons of value here.
So really looking forward to this. Brett,before we begin and really dive in,
(03:17):
tell for people whodunno anything about you,
tell us a little bit about whoyou are and what you do. Totally.
So I'm the CEO co-founder of OMG Commerce.
We are a performancedigital marketing agency.
We focus primarily on D two c e-commerceand omnichannel retail brands,
specialty with YouTube adsas we'll talk about today.
But also all thingsGoogle. We lean into meta.
(03:38):
We're full service on Amazon andwe do retention marketing as well.
So email and SMS,
but I'm just passionate aboutmarketing that works and marketing that
builds a brand and alsodrives bottom line impact.
So got team of about 40. We've been doingthis since 2010, won some cool awards,
been featured with some Google casestudies and some other fun stuff.
(03:58):
But at my core, I'm amarketing and business junkie,
and so thrilled to betalking with you today.
Awesome, awesome. I'm really lookingforward to this conversation.
I think one of the interesting thingsthat you and I were touching on before we
actually pressed record here wasthis idea around trying to measure
incrementality of some of these channelsthat are really hard to measure.
(04:20):
So for me, everyone knows I've gonesuper hard on TikTok this year.
I think the reason why TikTok shop wasso interesting to me is that we always
knew TikTok was driving business results.
Yes. But.
No one could ever tell you thatthere was no measure, couldn't.
Measure it.
Right? Because it just doesn'twork like that on TikTok.
It's not a direct response channel. Inthe same way now TikTok shop is right.
(04:42):
Shop has changed all of that,
and we have this nice closedecosystem that you can measure,
but I talk about this halo offof TikTok shop the entire time.
But interestingly,
YouTube is such a similar channel tome where it's like how many times do I,
am I watching something on YouTube?I see an ad and I'm like, yeah,
I should go buy that product.I totally forgot about that.
(05:03):
I never in my life have ever clicked froma YouTube ad over into buy a product,
never once because I'm watchingsomething, but it's there with me.
It's the TV of 2025.
Exactly.
Yeah. And.
It's so interesting.
Walk me through this.What are your thoughts?
Why are you still in the YouTube game.
And.
How do you continue to.
Scale? Partially becauseI'm a glu for punishment.
(05:24):
I don't like to do anythingthe easy way. Apparently.
I like to get beat up on a dailybasis. So leaning into YouTube,
it's a bit of a battle.
A lot of people have rid off YouTubeas a viable channel for growing their
business.
A lot of people opt to just pour alltheir dollars back into meta and hey,
I'm a believer in meta ads. We spendon meta, even as an agency, of course,
(05:46):
not just for our clients,but for omg, but for.
Yourselves, of course.
I love. But YouTube is so unique,and I love the way you framed it,
and I fully agree with this statement.
This is TV advertising for our generation,
but way better than TVadvertising could ever be.
And I'll kind of set this up a little bit.
So I got my start when I was incollege in the early two thousands.
(06:07):
I worked at a radio station and theygot involved in some TV and stuff,
but I just remember talking to localretailers, local furniture stores,
local car dealerships, local restaurants,
and they would tell me things like whenwe advertise on radio, maybe it works,
maybe it doesn't, I don'tknow. But when I run TV ads,
people come in and people mention it,and they couldn't directly attribute,
(06:30):
but they could feel it and they couldsee it. When the TV ads were on,
people were coming in, right?That's incrementality. You shut TV off.
That business goes away. Andso that's what YouTube is now,
and we've all just gotten addictedto last click or multitouch
attribution models wherewe want to see the click.
(06:51):
We want to see the evidence thatthis ad led to this sale or this
result. Ultimately, I thinkwe were all coming to that.
We've kind of come full circle. Itwas like, oh, offline is so difficult.
You can't measure it. Online ismagical. You can measure everything.
To always we're measuring online isreally shortsighted and maybe not very
helpful. We want to just as anexample, lean in fully to roas. Well,
(07:15):
if you want to maximize roas, just runbranded search. That's the maximum.
ROAS course. Of course. And thentry scaling in. Yeah, right.
This is the hilarious thing. It is funny.
I've been trying tolean in on how, for me,
I owned a bunch of brands over the yearsand trying to lean in on how I can help
e-comm brand owners,
(07:36):
especially the smaller onesthat we don't work with anymore.
I'm trying to figure out, I'm like,
how do I help you guys thatare listening to this podcast?
A lot of you guys arelistening to this podcast,
and profitability is an interesting thing.
We've been obsessed with thesemarketing metrics that just don't matter
What we see in branded search,
what we see in most peoplerunning Facebook ads on bottom of funnel. I'm like,
(07:59):
good. You think thatlooks good? Who cares?
It doesn't do anything for your business.
It's not going to make youprofitable at the end of the day.
This is something I obsessed with.
It off up, probably no negative impactto your business. Shut that off.
You'll probably get all thesales. You're currently kidding.
Totally, totally. Exactly. YouTube again,
I was telling you about a funny littlesmall case study and from a colleague of
(08:20):
yours that's running this adaccountant, it's us spending a ton,
I think it's 20 or 30 KA month right now,
but I will tell you it as wescale it up, right on platform,
it's showing about a one return onad spend. So normally you'd say,
let's turn that off, right? Yeah. Well,when we turn that off, everything dies.
There's no business. Sothat's number one. Number two,
(08:42):
the MER is about 3.5,
so I'm not dumb. I knowsomething's happening there.
So that's my sort of look atall of this of like, wait,
we've got to get out of this granularitythat we think that we know how a
customer journey works. Come on.
(09:03):
We'll never know how acustomer journey exactly works.
Customers don't know how their customerjourney works. They don't know they're,
they're seeing things andreacting and then reacting again.
And they probably couldn't evenstitch together their buying journey.
So what makes you think you'll nailthat with 100% accuracy? You just won't.
So tell me what you guys do.
(09:23):
You've been all in on YouTube sinceI've known you. You're the YouTube guy,
Brett.
Yeah. I mean, this iswhat we're known for.
A lot of big brands come to us becausewe've cracked the code, so to speak,
on how to make YouTube adswork. And the good news is,
while it's not measurablein the same way that meta
ads are, or in the same waythat search or shopping ads are,
(09:47):
you can measure it. You've justgot to work a little bit harder.
You got to kind of triangulatethe truth a little bit,
but you nailed it earlier,Jordan, when you said,
I've never clicked on a YouTube ad, right?YouTube is the TV of our generation.
Nobody clicks on their tv. So youjust use that as your comparison.
And actually this is interesting.Over 50% of views now,
and I've seen numbers show as muchas 60% of all views on YouTube,
(10:11):
just the YouTube app are on TV screens.
It's the fastest growing platform orfastest growing screen for YouTube
consumption. Yes,
there's the option to send a phone oryou can scan a QR code on those YouTube
ads that appear on a smartTV or a connected tv.
But that's so rare that usuallyyou're doing something on YouTube,
(10:32):
you're probably notgoing to do those things.
And so it's just people don'tinteract with YouTube the way they do,
even Facebook or TikTok. And so you have,
I want to give you that at acore level. Yeah, go ahead.
I just want to give everybodya fun anecdote here.
So my kids watch a lot of YouTube. They'vegot the few creators that they love.
My daughters love Mariah Elizabeth,she does all these crafts and stuff,
(10:56):
and then obviously Mr. Beast,right? Everyone, Mr. Beast.
Mr. Beast. Beast. Yep. He's the best.
And there's these massively longYouTube ads in between, right?
There's like 62nd.
My kids recently came to me and theysaid, dad, I hope in the next election,
now remember I'm Canadian and weare now going, right, of course.
Just like everyone's making the swings.
(11:18):
And so they come up tome and they're like, dad,
I really hope in the next election thatyou're going to be voting for Pierre
Polly because his taxplan seems really good.
Remember this is 11 and 9-year-old.
I've heard a lot of things aboutwhat he said, and I'm like,
don't ever tell me YouTube adsdon't work because exactly. You.
Have to be watching them,dude, that is so good.
(11:40):
And I remember as a kid, for me,
and this still was stuck in my brain andprobably meant that I was destined for
a career in advertising.
I remember trying to convince myparents to buy Ginsu knives because I
saw these great commercials and I'mlike, guys, you don't understand.
You can cut through a can of Pepsiand then you can cut a tomato without
(12:01):
sharpening. And I'm just regurgitatingthe ad, but it's such a good example.
So yeah, your kids are now aspokesperson for political qualities.
Exactly. Exactly.
Yeah, that's such a good story. Anything.
So what's funny is that this isyears ago, I think this is 2018,
maybe 2017 or 2018,somewhere around there.
(12:22):
And we just had this littletiny agency at the time.
Now we're kind of similarto your guys' size.
I was actually just lookingaround 40 right now.
And.
I think there was three of us at the time,
and my accountant decided to run formayor, and I thought to myself, well,
what's the best way thatI can get him seen by
(12:42):
everybody in the city that we live in?Abbotsford, it's called. It's like,
how can I get him seen? And I was like,I think I'm just going to go to YouTube.
And so I've made thesereally long videos of him.
Remember he was acandidate that nobody knew.
He ended up getting secondplace and getting about,
I think it was somewhere around like30% of the vote. And I was like, whoa.
And even more, we have this meme accounton Instagram called Asford Memes,
(13:06):
and one of the posts on there was abouthis YouTube videos being everywhere.
And I was like, crazy. Yes, this works.
I wish I would've justtripled down on that.
No doubt. No doubt. Yeah.
Last kind of little anecdote that'srelated to this part of what we're talking
about. We have an appliance store client,
someone that I've known a long time there.
They've got locations here inSpringfield in Missouri where I'm based,
but all over Oklahoma and stuff.
(13:28):
So we've been talking to themfor a long time about, Hey,
give us a little bit of your TVbudget, because they spent a lot on tv,
sell appliances,
give us a little bit for YouTube becausewe were running all their Google search
ads and stuff like that. And sothey finally gave us a budget,
not a ton of budget,
but some for a few key marketsfor one of their appliance chains.
And they just came back to us and theysaid, guys in Q1 so far of this year,
(13:51):
these stores are up 35% over last year.
One of their other chains isflat, basically. And they're like.
Interesting. So you gotto have a case study.
A case study. Yeah. So I don't know ifthey want anything released publicly. So
I've got a few case studies I can getinto in detail with you, please. Yeah.
(14:12):
Yeah, 35% over other stores.
And the only thing they changedwas taking some TV budget,
putting it onto YouTube, andvoila, they got more gross.
I mean, this is the thing that we'vetalked, and we've talked about this.
It's funny because I'm on TikTok landover here and you're over in YouTube land,
and these are the two things that nobodyon Twitter really ever talks about
(14:33):
because they can't measure them,
and yet we all know thatthey drive both of these,
drive this massive halo. Right? There is,
I'm sure we'll talk invery generalizations here,
but I was talking to a guy the other daywho happens to do some advertising for
one of the biggest supplementcompanies on Amazon,
and they're the biggestsupplement company right now on,
(14:56):
or one of them out on TikTok shop.
Their Amazon business has doubledthis year with no more Amazon spend.
All I'm saying is we've got,
if you guys are going to getanything from this episode,
just remember that in the meme curve,
the person that is obsessedwith Triple Whale is
(15:16):
not the one at this end ofthe curve, right? Yep, yep.
If you're obsessively lookingat your numbers like that,
that's not how you run a business.
Totally.
That's not how you do it.
Yeah, yeah. So true. So let megive you a couple of examples,
a couple of additional examples,
and then we may want to talkthrough too at some point.
When does it make sense tostart testing YouTube ads?
When are you maybe not ready to teston YouTube ads? Because I will say,
(15:37):
even though I am a believer in YouTubeads, I talk about it all the time.
I speak on stages.
I get invited to YouTube offices a coupletimes a year to teach brands how to do
this. It's not for everybody,
and it's not the place tostart if you're a young brand,
probably so happy to talk through that.
But one case study that Ithink is really interesting,
so this was a brand we workedwith a number of years ago.
(16:00):
It's a haircare product forwomen with thinning hair,
and this actually works sowell that now the founder,
he sold the business andnow he's a partner in OMG,
which is kind of a cool fun first.
That's so fun.
Yeah. Yeah, exactly. So with Kiran,
they were doing a lotwith direct response tv,
doing a lot with Metaand a few other channels,
but they could not get YouTubeto work. They tried it,
(16:20):
tried and failed a numberof times. So we came in,
we remapped the campaign strategy,
we tweaked their creatives becausethe creative were powerful,
but they weren't quite built forYouTube, so we redid the creatives.
Are we going to get into that?
Because I think that's an interestingthing to get into is what makes a YouTube
creative amazing. Sure.
Yeah. So we can kind of dig into ourformula guidelines there on creative,
(16:43):
but we tweaked the creative,we redid the campaigns.
We focused on who we're going to targeton YouTube because targeting is so
important on YouTube,
we went from zero to a million dollarsin spend on YouTube and under 90 days
while hitting their target CPA.So this was direct response,
applicable sales. Wow. Theyhad a high CPA ceiling,
(17:05):
they had a landing page that convertedreally well. They had a great product.
They had a high take rate on subscriptionsso they could pay a lot for new
customers. So it all worked, butthat was just purely D two C. Now,
what we found then is as we started doingthat for months and months and months,
we were asked to help them with theirAmazon business, and they were like,
Amazon's really taken off.Amazon is going like gangbusters.
(17:26):
And so we started helpingthem grow on Amazon.
We do full Amazonchannel management. Well,
we hit a patch where we had to pauseYouTube ads for a bit due to some internal
tracking and a couple of the thingsyou had to pause it for a bit.
Oh, my favorite kind of ABtests where it's like, yeah.
Exactly.
I was forced to do this.
The credit card bounced and I didn'tsee it for seven days. The best tests.
Yeah, yeah, yeah, exactly.It wasn't a credit card best,
(17:48):
but they had to pause fora bit right when they did,
because we had fullvisibility into Amazon,
we were managing Amazonbranded search cut in half,
cut in half when we had to pause YouTube.
Then we got YouTube back and they wererock and rolling again before long,
but we were like, holycrap, that's pretty crazy.
So basically we came to the conclusionthat it was about a one-to-one ratio for
(18:10):
every conversion. We saw D two C, wewere getting at least one on Amazon,
and one of the data scientists onthe team was like, it might be two.
It might be two to one for everyone you're getting D two C,
you might be getting two on Amazon.
And so it's one of these things wherewere they in retail as well that were they
in retail as well? Theywere not in retail.
Now I've got another case that I talkabout that there was fully retail in
(18:31):
Walmart stores. That's incrediblyfascinating. But this brand, no,
they were multi-channel.
So they were D two C plus Amazonplus a few other marketplaces.
They were not in stores at that time.
Okay, okay. Insane.
And also I think that we allbrand people who have been
in this game for a while are thosethings you do that you can't measure.
(18:56):
That.
Is the right thing to do and actuallydrives the top of funnel business,
but then you're convinced bysomebody, maybe it's a new CMO,
maybe it's a new somebodywho comes into the space,
maybe it's just a viral tweet on Twitteror on whatever we call that thing now
that just shouldn't have goneviral because it was wrong.
(19:19):
You must measure everything. I onlyscale things that are measurable.
And then you're like, oh, okay,
well I'm going to turn this thingoff and then your business tanks.
And I think it's important.
We're certainly not making the case tomeasure less and we're not making the
case to pay attention to your numbers.
It's just that not everythingcan be measured the same way.
I heard this, some of my team made thisanalogy one time. They were like, Hey,
(19:43):
think about a light switch.
You flip a switch and the resultyou don't see on the light switch,
the result is over herewith the light. I was like,
this is a pretty good analogy wherethat's a really good analogy, Brett,
like YouTube, you're running YouTube,you're doing this thing right here.
The results are over here. You seethe results over here, they're there.
You've just got to look for them.And measuring them is harder.
(20:06):
So I don't think you should measureless. I think you should measure more.
I just think you should understand thatROAS is a bad metric overall if that's
all you're looking atand the whole picture,
understanding incrementalityand real growth and real impact.
It takes some work. It takessome work to measure it.
So you got to look at everything.
I can also show you a casestudy that might be helpful.
(20:26):
It's from Arctic coolers, so coolersand drinkware, a Yeti competitors.
Happy to dive into that or I think youmay have had a question right there.
I do have a question I have to put inthe middle here, Brad, I'm so sorry.
But how are tools like Prescientand mms, do you trust them?
Do you think that they aretelling the full story?
Do we need to move from anMTA model into an mm M model?
Just what do you think about that?
(20:47):
It's a great question.
So I think there's acouple things at play here.
I still like MTA tools. Ithink they serve a purpose.
We have clients who use North Beamclients who use triple whale. I like them.
I think we can gain some insights there.
They got different modelsyou can kind of play with.
And so it can give someinsights. Certainly,
I do tend to think thatan incrementality test
(21:10):
plus an MMM type of approach isprobably a little bit better.
I would lean that way. But I thinkyou can also do things like, hey,
if you can track in platformperformance and you can do things like
incrementality tests on yourown, which you can do that.
You can set up geo tests, geoholdouts and things like that.
Totally just do holdouts, right?It's that simple. Just do a geo.
(21:32):
Let's say New York and Californiaare bringing you basically the same.
That's right. California and Texas,
just because they're similar and similarpopulations. If they're doing that,
you can just hold out Texas and thenyou can know measure the impact.
There's the.
Impact. And so there's also, there'ssome tools that Google can help you with.
They'll do things like search liftstudies, they do conversion lift studies,
(21:54):
which basically is a way for them to dotheir own kind of holdouts and tell you
the incrementality of a particularcampaign type. Quite useful.
So I think that can kind ofget you on the right track.
And I do think one of the bestnumbers, the best metrics,
which you really don't need anyfancy tool for is mer, right?
Look at your total mediaefficiency ratio and watch how that
(22:16):
fluctuates as you lean intoone channel over another. Hey,
when I really lean intoYouTube for a couple of months,
my MER improves and my top line grows.
That sometimes simple is all you need,
but you've got to be looking inthe right places. I love MMM.
I've not used prescient, but I'veheard good things about them.
But I do think the world is probablyshifting more to MMM and incrementality
(22:39):
tests and less towards MTA.
Yeah, yeah.
MTAI think was something that webrought in when we needed it, right?
And especially post iOS 14.5.
Solve some real problems therefor a bit. Yeah, for sure.
Absolutely. So Brett,
I want to get into your case studyand then after your case study,
I want to hear your, so guys,
stick around because Brett is going tobreak down what makes a perfect YouTube
(23:01):
ad.
Yeah, yeah, yeah. So excited about that.
I guess about a year ago I was speakingat the YouTube offices in Los Angeles
teaching a group of brands how to growon YouTube. Met the folks from Arctic,
which I was not that familiar withthat brand prior to that point,
but Arctic is, they sell wheeledcoolers like indestructible,
the best cooler I've ever had.They're direct competitor to Yeti.
(23:23):
I think they're better and they're lessexpensive. They also sell drinkware.
And I'm a big Tumblr fan. Idrink a lot of coffee. Jordan,
are you a coffee drinkers? Well, Jordan.
I drink a cup or two a day. Myson got me this cup here. I dunno,
like a liter basically.
It's kind of cool. Yeah, love coffee.
I go to bed singing aboutright as some of them big fan.
(23:47):
But Arctic has drinkware thatis ceramic lined. One pet peeve,
and this is going to showthat I'm a little bit of a snob when it comes to drink
beverages. If you drinkout of a metal Tumblr,
it smells and tastes alittle bit like steel, right?
Almost like you have blood in yourmouth, a little bit of that, right?
The ceramic liningcompletely takes that away.
It tastes like you're drinking out ofyour favorite ceramic mug or whatever.
(24:08):
It's brilliant. Anyway, metArctic. They were like, Hey,
we want to get YouTube to work forus, and here is our hypothesis.
Can you help us prove it? We want to seeif YouTube will drive in-store sales.
They were just put into all Walmartstores are almost all Walmart stores
nationwide. They wanted to know ifwe ran some targeted YouTube tests,
can we see a measurable lift inWalmart sales? And so we said,
(24:32):
we love this idea. Theyhad great creatives.
We coached them a little bit andhelped them with their creatives,
which we'll get into more in aminute. But here's what we did.
We basically did some studies withthe help of Google to find what
cities had really high category demand,
meaning people looking forcoolers and wield coolers,
(24:52):
but then had varying ingredientsof demand for Arctic.
And this was all based onGoogle's search behavior.
So we mapped out and wesaid, okay, let's do a test.
Let's choose 19 markets thatwe think are poised for growth,
got good store count, goodinventory levels, things like that.
Let's match those with 19 control marketsthat are basically the same size and
same everything.
(25:12):
And we're going to spend YouTube dollarsin the test markets and we're going to
hold out and the control markets,we're going to do that for five. And.
There was Walmarts in all. Walmarts.
In all of them. Yeah,exactly. So then we could do,
so it's a mashed pair holdout type study.
So we lean in hard and we used all ofour resources to build out the right
campaigns and maximize views,
(25:33):
and we did lean inheavily to connected tv.
So that was a great placement forthese ads. And we measured on YouTube.
On YouTube. All YouTube's connectedlike YouTube's TV placements. Exactly.
This is now.
YouTube tv. A lot of people ask aboutthat. I'm a subscriber to YouTube tv.
It's a good cable replacement.That's a separate inventory though.
It's powerful, but it's actually apretty small percentage of YouTube views.
(25:56):
And if you buy YouTube ads in theplatform, like in the Google Ads platform,
you're going to be on traditional YouTube.
But here's one really crazyside note that ties into this.
More people streamYouTube on connected TVs.
So the good old fashioned YouTube appon connected TVs that stream any other
platform. So YouTube, I.
Do not doubt that even for a second.
(26:17):
Bigger than Netflix, bigger thanHulu, bigger than Disney plus,
it's bigger than Hulu. DisneyPlus and a few others combined.
It's bigger than Netflix,but not that dramatically.
So we leaned in heavily to connectto TVs. There's so much inventory,
there's so much. The inventory is insane.
So many people aren't advertising herebecause they don't get it and they don't
understand it. And so CPMs are prettylow compared to men and other places.
(26:40):
A lot of opportunity here.
So we did this five week test andwe measured it all along the way.
We did see Lyft and D two C, so we wereable to measure those and track those.
All the ads said, go to Walmart,go buy this at your local Walmart.
You could click to find a local storethrough Google tracked people that visited
stores. All of this was all about local,but we did see some D two C sales,
(27:00):
right? Some people were justlike, of course I'll buy it.
Online course. No, I want to buy it now.
Yeah, I want to buy it now.I want to buy it online.
We saw Lyft on Amazon because alot of people were like, yeah,
I want to buy it now, but I don'twant to buy it from your store.
I want to buy it on Amazon becausethat's my favorite. So we saw that.
But here's what we sawinside a Walmart stores.
We had three different groups of markets.
The worst performing groupsaw a 12% retail store
(27:22):
lift in comparison of the testmarkets to the control markets,
test markets 12% highdegree of confidence lift
over their control markets were verysimilar in size and nature and all that
stuff. The best group was 25% lift
over the control market. Sobasically they were like,
(27:45):
holy crap, this is great.Let's do more of this.
It actually lets us run a whole bunch of,
or getting more of their business.Google was like, Hey,
we got an agency excellence award for it.
We really leaned into thismore in the process too.
We also did what's called the searchlift study where Google can look at, hey,
people that saw your ad versuspeople that didn't see your ad,
(28:05):
what kind of impact does that haveon people searching for your brand?
So we saw a 241% lift in branded searches
from this ad campaign, andthis is where it's like, okay,
you've got to kind of look at theholistic picture. And I will say,
this is where YouTube reallyshines from omnichannel retailers.
They can shine for multichannel retailersas well, but the results were stagger.
(28:29):
Incredible. We don't havemuch time here. Brett,
I want you to walk methrough before you go,
walk me through what theperfect YouTube ad looks like.
Yeah,
there are a few elements and I've gota guide that we can mention that's free
that walks through 17 ofour favorite YouTube ads.
It breaks down each oneand shows. Great. We'll.
Get that down in the show notes. And ifyou guys are watching this on YouTube,
(28:50):
it'll be down in the description.
Below. Awesome. So the firstthing, no surprise here probably,
but you need a hook first. Fiveseconds is critical. People are,
if it's a skippable ad,which is most of what we run,
you got five seconds toreally hook somebody.
Some people are watching hovering withtheir finger of the mount over the skip
button ready to click. So you gotfive seconds to convince them, Hey,
(29:12):
wait a minute, check thisout a little bit longer.
So hook is important and what we say isyou need to make sure you're hooking the
right person. So don't come up withjust some random hook explosion,
something funny guy in a gorillasuit, something just wild.
You want to interrupt the rightperson, right? So your ideal target,
but then you want to upbecause you only pay based on.
(29:33):
An actual view through. Yeah, exactly.Yeah. So you don't want the wrong,
you basically want to be like,Hey, everyone else don't watch.
Yeah,
you want the non-ideal shopper to skipbecause you only are paying for good
views.
And so you want someone to choose towatch your ad if they're in your market.
So something to grab their intention,thought provoking question,
a bold statement showingsomething right in their face.
(29:55):
But it needs to berelated to what you saw.
So it needs to be related to the problemyou're solving or the relief you're
bringing or the enjoymentthat your product is offering.
So it's got to be relevant,a relevant interruption,
a relevant hook to the right person.So the hook is critical. Next,
if you're D two C,
essentially most of ourclients are retail in some way.
It's got to be a product demonstration.
(30:16):
You need to now show me theproduct in action. And Jordan,
I like a typical demonstration.
So what's the proper use case of thisone? A good example is a flex seal,
like the spray on thing toseal. If you have a leak,
leak in your gutter orleak in a pot or whatever,
you spray flex seal on your obsolete.
So I like a standarduse case demonstration.
(30:38):
This is how you seal yourgutters type of thing.
But then I like an overthe top demonstration.
If you can throw that in theretoo. It's kind of like the,
I dunno if you like the.
Pepsi can thing you talked about before.
Yeah, exactly. So it's like the knife,
the Ginsu knife commercial wherethey cut through the can or the old,
old super glue commercialwhere they in one cut,
put super glue in the bottom of a guy'sshoe and stuck him to the ceiling,
(31:00):
held it, and then he stayed there anyway.
So show a dramaticproduct demonstration too,
because you want them to see thestandard demonstration and say,
I could do that, right? But thenthey're also kind of skeptical.
I like talk about Missouri,
where I live is we're the show mestate and our state animal is a mule.
We're stubborn. We believe nothing. Yougot to show us before you believe us.
(31:21):
Picture that your audienceis from Missouri, right?
You got to show them andthey're going to be skeptical.
Standard product demo and thenover the top product demo.
Then I like some social proof,
so show me something thatproves that other people like
me love this product.So this can be reviews.
So when we were doing stuff for nativedeodorant and we helped them really scale
(31:42):
on YouTube, we said like, Hey, therewas only over 50,005 star reviews.
How could 50,000 people be wrong typeof thing. So some kind of social proof.
Then I like some kind of risk reversalor some kind of offer where it's like,
hey, do this, try this, experiencethis. We'll take away the risk.
And then a firm call to action.And so that call to action is, Hey,
(32:04):
click here to learn more.Go to Amazon and buy it.
Buy it from your favorite retailer.Something like that. So how.
Professional does the filmingneed to be on YouTube? Again,
I talk about TikTok all the time,right? It's the opposite, right?
You don't want professional filming,it doesn't work. It's the antithesis.
What about YouTube?
(32:24):
Yeah, it's kind of the same with metatoo. Sometimes it's the unpolished,
the shot with the phone kind of rawcandid stuff. It works on YouTube.
You do need a little more polished.
Think again about the frame that thisis like TV and a lot of people are
probably watching it on tv.A little more polished is important.
Now we found that we've a lot of successwith an ad that is professionally
(32:45):
produced. You don't have to spendtens of thousands of dollars on it,
but you need good lighting and a goodcamera and good audio and things like
that.
But sometimes when you mix goodproduction value with some raw stuff mixed
in, so good high production value,
then throw in some UGC eitherin the middle or at the end.
That combination extremelypowerful. Now we've also,
(33:07):
we scaled this auto brand onYouTube a couple of years ago,
and basically all they had wasUGC, but it was really good UGC.
And so we did UGC mashups,
which just a whole bunch ofcustomers showing their car,
talking about why they lovethe product, things like that.
But then we added some editing and somegraphics and some transition that really
made it look polished andthat thing scaled like crazy. So it's definitely going
(33:28):
to be higher productionvalue than TikTok or Meta,
but it does not have to be million dollartype of stuff that you see on Super
Bowl commercials. You don't need that.
And then I'll also say if you're runningan ad on Instagram reels or on TikTok
or TikTok shops, whatever,
you can take that ad almost exactlyhow it is and run it on YouTube
(33:48):
shorts. I will say though,
I haven't seen as much results withYouTube shorts only as I have when you're
kind of doing all of YouTubeor you got YouTube shorts,
you got YouTube on desktopand mobile and connected TVs,
which takes all a little bitdifferent format for each of those.
That's when it really works.
We've seen less of an impactwhen it's just YouTube shorts.
Last question, becauseI am late for a meeting.
(34:09):
Yeah.
What length are we looking at here?Is there an ideal minimum length?
And I am saying minimumbecause I understand that there probably is a minimum.
There is.
I prefer 45 seconds to threeminutes as kind of an ideal
length that if you could shoot for 90seconds or 60 seconds, that's great.
Ultimately it's got to check thebox, it's got to hook somebody,
(34:29):
it's got to be compelling,
it's got to have the product demo andthe social proof and the offer and all
that. But usually if it's a littlebit longer, the more you tell,
the more you sell, right? Themore you show, the more dough.
I just made that on the spot.That's actually really lame.
We've that are like 62nd cuts,
outperform 32nd cuts when it's basicallythe same ad, just the 60 numbers,
A 32nd,
the 62nd does betterwith clicks and overall
(34:54):
lift in sales. So generally speaking,
we want to go kind of in that45 to minute and a half range,
but as much as threeminutes can also work.
Thank you. Thank you, thank you.
Where can people find out more aboutwhat you are up to and chat with OMG
commerce?
Absolutely. So omg commerce.com,
click the Let's talk button and that'show you end up getting with our team.
(35:16):
You can find my podcast thereby clicking on resources.
You can see the YouTubeguide and up there as well.
And then I'm active on LinkedIn,so hit me up on LinkedIn.
Would love to connect to you.
I do talk about all things D two C andretail and business and Google and Amazon
and all kinds of stuff. So would loveto connect on LinkedIn as well. Awesome.
And.
I'm looking forward to coming on your pod.
It's going to be, let's do it, man.
It's going to be a blast schedule talkTikTok shop. So I'm excited. Awesome.
(35:39):
Well thanks again for your timetoday, Brett. Thanks Jordan.
And as always, thank you for tuningin. We'd love to hear more from you.
So let us know what topics wouldyou like to cover on this pod.
And if you know somebody thatwould benefit from this episode,
please share it and leave us a reviewon iTunes. That would make our day.
(35:59):
And with that, until nexttime, thank you for listening.
This is like TV advertisingfor our generation,
but way better than TVadvertising could ever be.
Well, hello and welcome to another editionof the E-Commerce Evolution podcast.
(00:24):
I'm your host, BrettCurry, CEO of OMG Commerce,
and today is a unique episode.
This is actually a recording that I did.
I was a guest on my buddy Jordan West's
podcast and we dove into YouTube. Now,
Jordan West is a pro. He isa master of TikTok shops.
(00:46):
He's built multiple brands,
and his podcast is the secretsto scaling your E-Commerce brand.
And we talked all about YouTube. Now,
the reason we dove intothis episode right now,
this is coming off the heels ofreally a groundbreaking research
piece or research project thatwas released by House Analytics,
HAUS.
(01:08):
Olivia Corey and team there have donesome brilliant work all around the
idea and the practice of measuringincrementality. We're seeing the true
impact of channels. And so they tested,
they did 190 incrementalitytests that involved YouTube with
74 brands, and theresults were mind blowing.
(01:29):
She was on a podcast with Andrew FerrisAlbe on that podcast soon as well.
But the idea here is that YouTube severelyunderreport or Google Ads severely
underreport, the actual impact theYouTube ads has to the tune of about
70% underreporting. Oranother way to look at that.
The reverse is it's about 3.42 x more
(01:53):
incremental than what you see there.
So if you're seeing a one rowas an example in platform,
the actual impact is probablya 3.42 x ROAS or 3.42 roas,
which is amazing. And so in this podcast,
Jordan asked me about ourexperiences as an agency.
We're known as one of the topYouTube ad agencies for D two C
brands. So tell lots of stories, lotsof examples on the creative side,
(02:15):
on the measurement side ways we'veused YouTube to grow retail sales,
omnichannel sales, I think it'sgoing to be informative, instructive,
and hopefully fun. So with that,
please enjoy the interviewthat I did with Jordan West
YouTube ads.
(02:36):
Hey guys, Jordan West back here Today.
I am so excited aboutthis podcast episode.
I have one of the experts in thespace in the YouTube space that is
Brett Curry on the podcast for I believea second time, maybe even third time.
Brett, welcome back to the show.
Yeah, Jordan Min awesome to be here. Itis for sure, at least the second time.
(02:57):
And maybe the third, I'm not sure,but let's make this the best one yet.
That'll be our goal.
I was so happy when you reached showedto me, I'm like, yes, I want to.
I was really hoping we were goingto talk about YouTube. I just said,
Brett Ray showed. I'm like, Iwill have him on the podcast.
I have no doubt that there'sgoing to be tons of value here.
So really looking forward to this. Brett,before we begin and really dive in,
(03:17):
tell for people whodunno anything about you,
tell us a little bit about whoyou are and what you do. Totally.
So I'm the CEO co-founder of OMG Commerce.
We are a performancedigital marketing agency.
We focus primarily on D two c e-commerceand omnichannel retail brands,
specialty with YouTube adsas we'll talk about today.
But also all thingsGoogle. We lean into meta.
(03:38):
We're full service on Amazon andwe do retention marketing as well.
So email and SMS,
but I'm just passionate aboutmarketing that works and marketing that
builds a brand and alsodrives bottom line impact.
So got team of about 40. We've been doingthis since 2010, won some cool awards,
been featured with some Google casestudies and some other fun stuff.
(03:58):
But at my core, I'm amarketing and business junkie,
and so thrilled to betalking with you today.
Awesome, awesome. I'm really lookingforward to this conversation.
I think one of the interesting thingsthat you and I were touching on before we
actually pressed record here wasthis idea around trying to measure
incrementality of some of these channelsthat are really hard to measure.
(04:20):
So for me, everyone knows I've gonesuper hard on TikTok this year.
I think the reason why TikTok shop wasso interesting to me is that we always
knew TikTok was driving business results.
Yes. But.
No one could ever tell you thatthere was no measure, couldn't.
Measure it.
Right? Because it just doesn'twork like that on TikTok.
It's not a direct response channel. Inthe same way now TikTok shop is right.
(04:42):
Shop has changed all of that,
and we have this nice closedecosystem that you can measure,
but I talk about this halo offof TikTok shop the entire time.
But interestingly,
YouTube is such a similar channel tome where it's like how many times do I,
am I watching something on YouTube?I see an ad and I'm like, yeah,
I should go buy that product.I totally forgot about that.
(05:03):
I never in my life have ever clicked froma YouTube ad over into buy a product,
never once because I'm watchingsomething, but it's there with me.
It's the TV of 2025.
Exactly.
Yeah. And.
It's so interesting.
Walk me through this.What are your thoughts?
Why are you still in the YouTube game.
And.
How do you continue to.
Scale? Partially becauseI'm a glu for punishment.
(05:24):
I don't like to do anythingthe easy way. Apparently.
I like to get beat up on a dailybasis. So leaning into YouTube,
it's a bit of a battle.
A lot of people have rid off YouTubeas a viable channel for growing their
business.
A lot of people opt to just pour alltheir dollars back into meta and hey,
I'm a believer in meta ads. We spendon meta, even as an agency, of course,
(05:46):
not just for our clients,but for omg, but for.
Yourselves, of course.
I love. But YouTube is so unique,and I love the way you framed it,
and I fully agree with this statement.
This is TV advertising for our generation,
but way better than TVadvertising could ever be.
And I'll kind of set this up a little bit.
So I got my start when I was incollege in the early two thousands.
(06:07):
I worked at a radio station and theygot involved in some TV and stuff,
but I just remember talking to localretailers, local furniture stores,
local car dealerships, local restaurants,
and they would tell me things like whenwe advertise on radio, maybe it works,
maybe it doesn't, I don'tknow. But when I run TV ads,
people come in and people mention it,and they couldn't directly attribute,
(06:30):
but they could feel it and they couldsee it. When the TV ads were on,
people were coming in, right?That's incrementality. You shut TV off.
That business goes away. Andso that's what YouTube is now,
and we've all just gotten addictedto last click or multitouch
attribution models wherewe want to see the click.
(06:51):
We want to see the evidence thatthis ad led to this sale or this
result. Ultimately, I thinkwe were all coming to that.
We've kind of come full circle. Itwas like, oh, offline is so difficult.
You can't measure it. Online ismagical. You can measure everything.
To always we're measuring online isreally shortsighted and maybe not very
helpful. We want to just as anexample, lean in fully to roas. Well,
(07:15):
if you want to maximize roas, just runbranded search. That's the maximum.
ROAS course. Of course. And thentry scaling in. Yeah, right.
This is the hilarious thing. It is funny.
I've been trying tolean in on how, for me,
I owned a bunch of brands over the yearsand trying to lean in on how I can help
e-comm brand owners,
(07:36):
especially the smaller onesthat we don't work with anymore.
I'm trying to figure out, I'm like,
how do I help you guys thatare listening to this podcast?
A lot of you guys arelistening to this podcast,
and profitability is an interesting thing.
We've been obsessed with thesemarketing metrics that just don't matter
What we see in branded search,
what we see in most peoplerunning Facebook ads on bottom of funnel. I'm like,
(07:59):
good. You think thatlooks good? Who cares?
It doesn't do anything for your business.
It's not going to make youprofitable at the end of the day.
This is something I obsessed with.
It off up, probably no negative impactto your business. Shut that off.
You'll probably get all thesales. You're currently kidding.
Totally, totally. Exactly. YouTube again,
I was telling you about a funny littlesmall case study and from a colleague of
(08:20):
yours that's running this adaccountant, it's us spending a ton,
I think it's 20 or 30 KA month right now,
but I will tell you it as wescale it up, right on platform,
it's showing about a one return onad spend. So normally you'd say,
let's turn that off, right? Yeah. Well,when we turn that off, everything dies.
There's no business. Sothat's number one. Number two,
(08:42):
the MER is about 3.5,
so I'm not dumb. I knowsomething's happening there.
So that's my sort of look atall of this of like, wait,
we've got to get out of this granularitythat we think that we know how a
customer journey works. Come on.
(09:03):
We'll never know how acustomer journey exactly works.
Customers don't know how their customerjourney works. They don't know they're,
they're seeing things andreacting and then reacting again.
And they probably couldn't evenstitch together their buying journey.
So what makes you think you'll nailthat with 100% accuracy? You just won't.
So tell me what you guys do.
(09:23):
You've been all in on YouTube sinceI've known you. You're the YouTube guy,
Brett.
Yeah. I mean, this iswhat we're known for.
A lot of big brands come to us becausewe've cracked the code, so to speak,
on how to make YouTube adswork. And the good news is,
while it's not measurablein the same way that meta
ads are, or in the same waythat search or shopping ads are,
(09:47):
you can measure it. You've justgot to work a little bit harder.
You got to kind of triangulatethe truth a little bit,
but you nailed it earlier,Jordan, when you said,
I've never clicked on a YouTube ad, right?YouTube is the TV of our generation.
Nobody clicks on their tv. So youjust use that as your comparison.
And actually this is interesting.Over 50% of views now,
and I've seen numbers show as muchas 60% of all views on YouTube,
(10:11):
just the YouTube app are on TV screens.
It's the fastest growing platform orfastest growing screen for YouTube
consumption. Yes,
there's the option to send a phone oryou can scan a QR code on those YouTube
ads that appear on a smartTV or a connected tv.
But that's so rare that usuallyyou're doing something on YouTube,
(10:32):
you're probably notgoing to do those things.
And so it's just people don'tinteract with YouTube the way they do,
even Facebook or TikTok. And so you have,
I want to give you that at acore level. Yeah, go ahead.
I just want to give everybodya fun anecdote here.
So my kids watch a lot of YouTube. They'vegot the few creators that they love.
My daughters love Mariah Elizabeth,she does all these crafts and stuff,
(10:56):
and then obviously Mr. Beast,right? Everyone, Mr. Beast.
Mr. Beast. Beast. Yep. He's the best.
And there's these massively longYouTube ads in between, right?
There's like 62nd.
My kids recently came to me and theysaid, dad, I hope in the next election,
now remember I'm Canadian and weare now going, right, of course.
Just like everyone's making the swings.
(11:18):
And so they come up tome and they're like, dad,
I really hope in the next election thatyou're going to be voting for Pierre
Polly because his taxplan seems really good.
Remember this is 11 and 9-year-old.
I've heard a lot of things aboutwhat he said, and I'm like,
don't ever tell me YouTube adsdon't work because exactly. You.
Have to be watching them,dude, that is so good.
(11:40):
And I remember as a kid, for me,
and this still was stuck in my brain andprobably meant that I was destined for
a career in advertising.
I remember trying to convince myparents to buy Ginsu knives because I
saw these great commercials and I'mlike, guys, you don't understand.
You can cut through a can of Pepsiand then you can cut a tomato without
(12:01):
sharpening. And I'm just regurgitatingthe ad, but it's such a good example.
So yeah, your kids are now aspokesperson for political qualities.
Exactly. Exactly.
Yeah, that's such a good story. Anything.
So what's funny is that this isyears ago, I think this is 2018,
maybe 2017 or 2018,somewhere around there.
(12:22):
And we just had this littletiny agency at the time.
Now we're kind of similarto your guys' size.
I was actually just lookingaround 40 right now.
And.
I think there was three of us at the time,
and my accountant decided to run formayor, and I thought to myself, well,
what's the best way thatI can get him seen by
(12:42):
everybody in the city that we live in?Abbotsford, it's called. It's like,
how can I get him seen? And I was like,I think I'm just going to go to YouTube.
And so I've made thesereally long videos of him.
Remember he was acandidate that nobody knew.
He ended up getting secondplace and getting about,
I think it was somewhere around like30% of the vote. And I was like, whoa.
And even more, we have this meme accounton Instagram called Asford Memes,
(13:06):
and one of the posts on there was abouthis YouTube videos being everywhere.
And I was like, crazy. Yes, this works.
I wish I would've justtripled down on that.
No doubt. No doubt. Yeah.
Last kind of little anecdote that'srelated to this part of what we're talking
about. We have an appliance store client,
someone that I've known a long time there.
They've got locations here inSpringfield in Missouri where I'm based,
but all over Oklahoma and stuff.
(13:28):
So we've been talking to themfor a long time about, Hey,
give us a little bit of your TVbudget, because they spent a lot on tv,
sell appliances,
give us a little bit for YouTube becausewe were running all their Google search
ads and stuff like that. And sothey finally gave us a budget,
not a ton of budget,
but some for a few key marketsfor one of their appliance chains.
And they just came back to us and theysaid, guys in Q1 so far of this year,
(13:51):
these stores are up 35% over last year.
One of their other chains isflat, basically. And they're like.
Interesting. So you gotto have a case study.
A case study. Yeah. So I don't know ifthey want anything released publicly. So
I've got a few case studies I can getinto in detail with you, please. Yeah.
(14:12):
Yeah, 35% over other stores.
And the only thing they changedwas taking some TV budget,
putting it onto YouTube, andvoila, they got more gross.
I mean, this is the thing that we'vetalked, and we've talked about this.
It's funny because I'm on TikTok landover here and you're over in YouTube land,
and these are the two things that nobodyon Twitter really ever talks about
(14:33):
because they can't measure them,
and yet we all know thatthey drive both of these,
drive this massive halo. Right? There is,
I'm sure we'll talk invery generalizations here,
but I was talking to a guy the other daywho happens to do some advertising for
one of the biggest supplementcompanies on Amazon,
and they're the biggestsupplement company right now on,
(14:56):
or one of them out on TikTok shop.
Their Amazon business has doubledthis year with no more Amazon spend.
All I'm saying is we've got,
if you guys are going to getanything from this episode,
just remember that in the meme curve,
the person that is obsessedwith Triple Whale is
(15:16):
not the one at this end ofthe curve, right? Yep, yep.
If you're obsessively lookingat your numbers like that,
that's not how you run a business.
Totally.
That's not how you do it.
Yeah, yeah. So true. So let megive you a couple of examples,
a couple of additional examples,
and then we may want to talkthrough too at some point.
When does it make sense tostart testing YouTube ads?
When are you maybe not ready to teston YouTube ads? Because I will say,
(15:37):
even though I am a believer in YouTubeads, I talk about it all the time.
I speak on stages.
I get invited to YouTube offices a coupletimes a year to teach brands how to do
this. It's not for everybody,
and it's not the place tostart if you're a young brand,
probably so happy to talk through that.
But one case study that Ithink is really interesting,
so this was a brand we workedwith a number of years ago.
(16:00):
It's a haircare product forwomen with thinning hair,
and this actually works sowell that now the founder,
he sold the business andnow he's a partner in OMG,
which is kind of a cool fun first.
That's so fun.
Yeah. Yeah, exactly. So with Kiran,
they were doing a lotwith direct response tv,
doing a lot with Metaand a few other channels,
but they could not get YouTubeto work. They tried it,
(16:20):
tried and failed a numberof times. So we came in,
we remapped the campaign strategy,
we tweaked their creatives becausethe creative were powerful,
but they weren't quite built forYouTube, so we redid the creatives.
Are we going to get into that?
Because I think that's an interestingthing to get into is what makes a YouTube
creative amazing. Sure.
Yeah. So we can kind of dig into ourformula guidelines there on creative,
(16:43):
but we tweaked the creative,we redid the campaigns.
We focused on who we're going to targeton YouTube because targeting is so
important on YouTube,
we went from zero to a million dollarsin spend on YouTube and under 90 days
while hitting their target CPA.So this was direct response,
applicable sales. Wow. Theyhad a high CPA ceiling,
(17:05):
they had a landing page that convertedreally well. They had a great product.
They had a high take rate on subscriptionsso they could pay a lot for new
customers. So it all worked, butthat was just purely D two C. Now,
what we found then is as we started doingthat for months and months and months,
we were asked to help them with theirAmazon business, and they were like,
Amazon's really taken off.Amazon is going like gangbusters.
(17:26):
And so we started helpingthem grow on Amazon.
We do full Amazonchannel management. Well,
we hit a patch where we had to pauseYouTube ads for a bit due to some internal
tracking and a couple of the thingsyou had to pause it for a bit.
Oh, my favorite kind of ABtests where it's like, yeah.
Exactly.
I was forced to do this.
The credit card bounced and I didn'tsee it for seven days. The best tests.
Yeah, yeah, yeah, exactly.It wasn't a credit card best,
(17:48):
but they had to pause fora bit right when they did,
because we had fullvisibility into Amazon,
we were managing Amazonbranded search cut in half,
cut in half when we had to pause YouTube.
Then we got YouTube back and they wererock and rolling again before long,
but we were like, holycrap, that's pretty crazy.
So basically we came to the conclusionthat it was about a one-to-one ratio for
(18:10):
every conversion. We saw D two C, wewere getting at least one on Amazon,
and one of the data scientists onthe team was like, it might be two.
It might be two to one for everyone you're getting D two C,
you might be getting two on Amazon.
And so it's one of these things wherewere they in retail as well that were they
in retail as well? Theywere not in retail.
Now I've got another case that I talkabout that there was fully retail in
(18:31):
Walmart stores. That's incrediblyfascinating. But this brand, no,
they were multi-channel.
So they were D two C plus Amazonplus a few other marketplaces.
They were not in stores at that time.
Okay, okay. Insane.
And also I think that we allbrand people who have been
in this game for a while are thosethings you do that you can't measure.
(18:56):
That.
Is the right thing to do and actuallydrives the top of funnel business,
but then you're convinced bysomebody, maybe it's a new CMO,
maybe it's a new somebodywho comes into the space,
maybe it's just a viral tweet on Twitteror on whatever we call that thing now
that just shouldn't have goneviral because it was wrong.
(19:19):
You must measure everything. I onlyscale things that are measurable.
And then you're like, oh, okay,
well I'm going to turn this thingoff and then your business tanks.
And I think it's important.
We're certainly not making the case tomeasure less and we're not making the
case to pay attention to your numbers.
It's just that not everythingcan be measured the same way.
I heard this, some of my team made thisanalogy one time. They were like, Hey,
(19:43):
think about a light switch.
You flip a switch and the resultyou don't see on the light switch,
the result is over herewith the light. I was like,
this is a pretty good analogy wherethat's a really good analogy, Brett,
like YouTube, you're running YouTube,you're doing this thing right here.
The results are over here. You seethe results over here, they're there.
You've just got to look for them.And measuring them is harder.
(20:06):
So I don't think you should measureless. I think you should measure more.
I just think you should understand thatROAS is a bad metric overall if that's
all you're looking atand the whole picture,
understanding incrementalityand real growth and real impact.
It takes some work. It takessome work to measure it.
So you got to look at everything.
I can also show you a casestudy that might be helpful.
(20:26):
It's from Arctic coolers, so coolersand drinkware, a Yeti competitors.
Happy to dive into that or I think youmay have had a question right there.
I do have a question I have to put inthe middle here, Brad, I'm so sorry.
But how are tools like Prescientand mms, do you trust them?
Do you think that they aretelling the full story?
Do we need to move from anMTA model into an mm M model?
Just what do you think about that?
(20:47):
It's a great question.
So I think there's acouple things at play here.
I still like MTA tools. Ithink they serve a purpose.
We have clients who use North Beamclients who use triple whale. I like them.
I think we can gain some insights there.
They got different modelsyou can kind of play with.
And so it can give someinsights. Certainly,
I do tend to think thatan incrementality test
(21:10):
plus an MMM type of approach isprobably a little bit better.
I would lean that way. But I thinkyou can also do things like, hey,
if you can track in platformperformance and you can do things like
incrementality tests on yourown, which you can do that.
You can set up geo tests, geoholdouts and things like that.
Totally just do holdouts, right?It's that simple. Just do a geo.
(21:32):
Let's say New York and Californiaare bringing you basically the same.
That's right. California and Texas,
just because they're similar and similarpopulations. If they're doing that,
you can just hold out Texas and thenyou can know measure the impact.
There's the.
Impact. And so there's also, there'ssome tools that Google can help you with.
They'll do things like search liftstudies, they do conversion lift studies,
(21:54):
which basically is a way for them to dotheir own kind of holdouts and tell you
the incrementality of a particularcampaign type. Quite useful.
So I think that can kind ofget you on the right track.
And I do think one of the bestnumbers, the best metrics,
which you really don't need anyfancy tool for is mer, right?
Look at your total mediaefficiency ratio and watch how that
(22:16):
fluctuates as you lean intoone channel over another. Hey,
when I really lean intoYouTube for a couple of months,
my MER improves and my top line grows.
That sometimes simple is all you need,
but you've got to be looking inthe right places. I love MMM.
I've not used prescient, but I'veheard good things about them.
But I do think the world is probablyshifting more to MMM and incrementality
(22:39):
tests and less towards MTA.
Yeah, yeah.
MTAI think was something that webrought in when we needed it, right?
And especially post iOS 14.5.
Solve some real problems therefor a bit. Yeah, for sure.
Absolutely. So Brett,
I want to get into your case studyand then after your case study,
I want to hear your, so guys,
stick around because Brett is going tobreak down what makes a perfect YouTube
(23:01):
ad.
Yeah, yeah, yeah. So excited about that.
I guess about a year ago I was speakingat the YouTube offices in Los Angeles
teaching a group of brands how to growon YouTube. Met the folks from Arctic,
which I was not that familiar withthat brand prior to that point,
but Arctic is, they sell wheeledcoolers like indestructible,
the best cooler I've ever had.They're direct competitor to Yeti.
(23:23):
I think they're better and they're lessexpensive. They also sell drinkware.
And I'm a big Tumblr fan. Idrink a lot of coffee. Jordan,
are you a coffee drinkers? Well, Jordan.
I drink a cup or two a day. Myson got me this cup here. I dunno,
like a liter basically.
It's kind of cool. Yeah, love coffee.
I go to bed singing aboutright as some of them big fan.
(23:47):
But Arctic has drinkware thatis ceramic lined. One pet peeve,
and this is going to showthat I'm a little bit of a snob when it comes to drink
beverages. If you drinkout of a metal Tumblr,
it smells and tastes alittle bit like steel, right?
Almost like you have blood in yourmouth, a little bit of that, right?
The ceramic liningcompletely takes that away.
It tastes like you're drinking out ofyour favorite ceramic mug or whatever.
(24:08):
It's brilliant. Anyway, metArctic. They were like, Hey,
we want to get YouTube to work forus, and here is our hypothesis.
Can you help us prove it? We want to seeif YouTube will drive in-store sales.
They were just put into all Walmartstores are almost all Walmart stores
nationwide. They wanted to know ifwe ran some targeted YouTube tests,
can we see a measurable lift inWalmart sales? And so we said,
(24:32):
we love this idea. Theyhad great creatives.
We coached them a little bit andhelped them with their creatives,
which we'll get into more in aminute. But here's what we did.
We basically did some studies withthe help of Google to find what
cities had really high category demand,
meaning people looking forcoolers and wield coolers,
(24:52):
but then had varying ingredientsof demand for Arctic.
And this was all based onGoogle's search behavior.
So we mapped out and wesaid, okay, let's do a test.
Let's choose 19 markets thatwe think are poised for growth,
got good store count, goodinventory levels, things like that.
Let's match those with 19 control marketsthat are basically the same size and
same everything.
(25:12):
And we're going to spend YouTube dollarsin the test markets and we're going to
hold out and the control markets,we're going to do that for five. And.
There was Walmarts in all. Walmarts.
In all of them. Yeah,exactly. So then we could do,
so it's a mashed pair holdout type study.
So we lean in hard and we used all ofour resources to build out the right
campaigns and maximize views,
(25:33):
and we did lean inheavily to connected tv.
So that was a great placement forthese ads. And we measured on YouTube.
On YouTube. All YouTube's connectedlike YouTube's TV placements. Exactly.
This is now.
YouTube tv. A lot of people ask aboutthat. I'm a subscriber to YouTube tv.
It's a good cable replacement.That's a separate inventory though.
It's powerful, but it's actually apretty small percentage of YouTube views.
(25:56):
And if you buy YouTube ads in theplatform, like in the Google Ads platform,
you're going to be on traditional YouTube.
But here's one really crazyside note that ties into this.
More people streamYouTube on connected TVs.
So the good old fashioned YouTube appon connected TVs that stream any other
platform. So YouTube, I.
Do not doubt that even for a second.
(26:17):
Bigger than Netflix, bigger thanHulu, bigger than Disney plus,
it's bigger than Hulu. DisneyPlus and a few others combined.
It's bigger than Netflix,but not that dramatically.
So we leaned in heavily to connectto TVs. There's so much inventory,
there's so much. The inventory is insane.
So many people aren't advertising herebecause they don't get it and they don't
understand it. And so CPMs are prettylow compared to men and other places.
(26:40):
A lot of opportunity here.
So we did this five week test andwe measured it all along the way.
We did see Lyft and D two C, so we wereable to measure those and track those.
All the ads said, go to Walmart,go buy this at your local Walmart.
You could click to find a local storethrough Google tracked people that visited
stores. All of this was all about local,but we did see some D two C sales,
(27:00):
right? Some people were justlike, of course I'll buy it.
Online course. No, I want to buy it now.
Yeah, I want to buy it now.I want to buy it online.
We saw Lyft on Amazon because alot of people were like, yeah,
I want to buy it now, but I don'twant to buy it from your store.
I want to buy it on Amazon becausethat's my favorite. So we saw that.
But here's what we sawinside a Walmart stores.
We had three different groups of markets.
The worst performing groupsaw a 12% retail store
(27:22):
lift in comparison of the testmarkets to the control markets,
test markets 12% highdegree of confidence lift
over their control markets were verysimilar in size and nature and all that
stuff. The best group was 25% lift
over the control market. Sobasically they were like,
(27:45):
holy crap, this is great.Let's do more of this.
It actually lets us run a whole bunch of,
or getting more of their business.Google was like, Hey,
we got an agency excellence award for it.
We really leaned into thismore in the process too.
We also did what's called the searchlift study where Google can look at, hey,
people that saw your ad versuspeople that didn't see your ad,
(28:05):
what kind of impact does that haveon people searching for your brand?
So we saw a 241% lift in branded searches
from this ad campaign, andthis is where it's like, okay,
you've got to kind of look at theholistic picture. And I will say,
this is where YouTube reallyshines from omnichannel retailers.
They can shine for multichannel retailersas well, but the results were stagger.
(28:29):
Incredible. We don't havemuch time here. Brett,
I want you to walk methrough before you go,
walk me through what theperfect YouTube ad looks like.
Yeah,
there are a few elements and I've gota guide that we can mention that's free
that walks through 17 ofour favorite YouTube ads.
It breaks down each oneand shows. Great. We'll.
Get that down in the show notes. And ifyou guys are watching this on YouTube,
(28:50):
it'll be down in the description.
Below. Awesome. So the firstthing, no surprise here probably,
but you need a hook first. Fiveseconds is critical. People are,
if it's a skippable ad,which is most of what we run,
you got five seconds toreally hook somebody.
Some people are watching hovering withtheir finger of the mount over the skip
button ready to click. So you gotfive seconds to convince them, Hey,
(29:12):
wait a minute, check thisout a little bit longer.
So hook is important and what we say isyou need to make sure you're hooking the
right person. So don't come up withjust some random hook explosion,
something funny guy in a gorillasuit, something just wild.
You want to interrupt the rightperson, right? So your ideal target,
but then you want to upbecause you only pay based on.
(29:33):
An actual view through. Yeah, exactly.Yeah. So you don't want the wrong,
you basically want to be like,Hey, everyone else don't watch.
Yeah,
you want the non-ideal shopper to skipbecause you only are paying for good
views.
And so you want someone to choose towatch your ad if they're in your market.
So something to grab their intention,thought provoking question,
a bold statement showingsomething right in their face.
(29:55):
But it needs to berelated to what you saw.
So it needs to be related to the problemyou're solving or the relief you're
bringing or the enjoymentthat your product is offering.
So it's got to be relevant,a relevant interruption,
a relevant hook to the right person.So the hook is critical. Next,
if you're D two C,
essentially most of ourclients are retail in some way.
It's got to be a product demonstration.
(30:16):
You need to now show me theproduct in action. And Jordan,
I like a typical demonstration.
So what's the proper use case of thisone? A good example is a flex seal,
like the spray on thing toseal. If you have a leak,
leak in your gutter orleak in a pot or whatever,
you spray flex seal on your obsolete.
So I like a standarduse case demonstration.
(30:38):
This is how you seal yourgutters type of thing.
But then I like an overthe top demonstration.
If you can throw that in theretoo. It's kind of like the,
I dunno if you like the.
Pepsi can thing you talked about before.
Yeah, exactly. So it's like the knife,
the Ginsu knife commercial wherethey cut through the can or the old,
old super glue commercialwhere they in one cut,
put super glue in the bottom of a guy'sshoe and stuck him to the ceiling,
(31:00):
held it, and then he stayed there anyway.
So show a dramaticproduct demonstration too,
because you want them to see thestandard demonstration and say,
I could do that, right? But thenthey're also kind of skeptical.
I like talk about Missouri,
where I live is we're the show mestate and our state animal is a mule.
We're stubborn. We believe nothing. Yougot to show us before you believe us.
(31:21):
Picture that your audienceis from Missouri, right?
You got to show them andthey're going to be skeptical.
Standard product demo and thenover the top product demo.
Then I like some social proof,
so show me something thatproves that other people like
me love this product.So this can be reviews.
So when we were doing stuff for nativedeodorant and we helped them really scale
(31:42):
on YouTube, we said like, Hey, therewas only over 50,005 star reviews.
How could 50,000 people be wrong typeof thing. So some kind of social proof.
Then I like some kind of risk reversalor some kind of offer where it's like,
hey, do this, try this, experiencethis. We'll take away the risk.
And then a firm call to action.And so that call to action is, Hey,
(32:04):
click here to learn more.Go to Amazon and buy it.
Buy it from your favorite retailer.Something like that. So how.
Professional does the filmingneed to be on YouTube? Again,
I talk about TikTok all the time,right? It's the opposite, right?
You don't want professional filming,it doesn't work. It's the antithesis.
What about YouTube?
(32:24):
Yeah, it's kind of the same with metatoo. Sometimes it's the unpolished,
the shot with the phone kind of rawcandid stuff. It works on YouTube.
You do need a little more polished.
Think again about the frame that thisis like TV and a lot of people are
probably watching it on tv.A little more polished is important.
Now we found that we've a lot of successwith an ad that is professionally
(32:45):
produced. You don't have to spendtens of thousands of dollars on it,
but you need good lighting and a goodcamera and good audio and things like
that.
But sometimes when you mix goodproduction value with some raw stuff mixed
in, so good high production value,
then throw in some UGC eitherin the middle or at the end.
That combination extremelypowerful. Now we've also,
(33:07):
we scaled this auto brand onYouTube a couple of years ago,
and basically all they had wasUGC, but it was really good UGC.
And so we did UGC mashups,
which just a whole bunch ofcustomers showing their car,
talking about why they lovethe product, things like that.
But then we added some editing and somegraphics and some transition that really
made it look polished andthat thing scaled like crazy. So it's definitely going
(33:28):
to be higher productionvalue than TikTok or Meta,
but it does not have to be million dollartype of stuff that you see on Super
Bowl commercials. You don't need that.
And then I'll also say if you're runningan ad on Instagram reels or on TikTok
or TikTok shops, whatever,
you can take that ad almost exactlyhow it is and run it on YouTube
(33:48):
shorts. I will say though,
I haven't seen as much results withYouTube shorts only as I have when you're
kind of doing all of YouTubeor you got YouTube shorts,
you got YouTube on desktopand mobile and connected TVs,
which takes all a little bitdifferent format for each of those.
That's when it really works.
We've seen less of an impactwhen it's just YouTube shorts.
Last question, becauseI am late for a meeting.
(34:09):
Yeah.
What length are we looking at here?Is there an ideal minimum length?
And I am saying minimumbecause I understand that there probably is a minimum.
There is.
I prefer 45 seconds to threeminutes as kind of an ideal
length that if you could shoot for 90seconds or 60 seconds, that's great.
Ultimately it's got to check thebox, it's got to hook somebody,
(34:29):
it's got to be compelling,
it's got to have the product demo andthe social proof and the offer and all
that. But usually if it's a littlebit longer, the more you tell,
the more you sell, right? Themore you show, the more dough.
I just made that on the spot.That's actually really lame.
We've that are like 62nd cuts,
outperform 32nd cuts when it's basicallythe same ad, just the 60 numbers,
A 32nd,
the 62nd does betterwith clicks and overall
(34:54):
lift in sales. So generally speaking,
we want to go kind of in that45 to minute and a half range,
but as much as threeminutes can also work.
Thank you. Thank you, thank you.
Where can people find out more aboutwhat you are up to and chat with OMG
commerce?
Absolutely. So omg commerce.com,
click the Let's talk button and that'show you end up getting with our team.
(35:16):
You can find my podcast thereby clicking on resources.
You can see the YouTubeguide and up there as well.
And then I'm active on LinkedIn,so hit me up on LinkedIn.
Would love to connect to you.
I do talk about all things D two C andretail and business and Google and Amazon
and all kinds of stuff. So would loveto connect on LinkedIn as well. Awesome.
And.
I'm looking forward to coming on your pod.
It's going to be, let's do it, man.
It's going to be a blast schedule talkTikTok shop. So I'm excited. Awesome.
(35:39):
Well thanks again for your timetoday, Brett. Thanks Jordan.
And as always, thank you for tuningin. We'd love to hear more from you.
So let us know what topics wouldyou like to cover on this pod.
And if you know somebody thatwould benefit from this episode,
please share it and leave us a reviewon iTunes. That would make our day.
(35:59):
And with that, until nexttime, thank you for listening.
Popular Podcasts
United States of Kennedy
United States of Kennedy is a podcast about our cultural fascination with the Kennedy dynasty. Every week, hosts Lyra Smith and George Civeris go into one aspect of the Kennedy story.
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.
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