How to survive in the cookieless era

Welcome to the E Commerce Growth Stories podcast. If you worry about your growth, if the business model isn't working anymore, if you're running out of ways to optimize your e commerce, or if you're looking to get the e commerce fundamentals right, then that's the show for you. Hello everyone, and welcome to a new episode of E Commerce Growth Stories. I'm here with Dennis Yu, which is back home and he can sleep in his bed after traveling. Hey, Dennis, how's it going?

Hey, good to see you again. Not as good as Bucharest, but this is still good.

All right, so today we're going to speak about the topic which is affecting all the e commerce landscape, which is the fact that the cookies are disappearing. The consequences of that, how to tackle this issue. And pretty much that's all about in this episode of E Commerce Growth Stories. So, Dennis, how we got here, that's the first question. Why do you think the cookies are not being served by the Internet giants?

Well, Facebook wants to keep the data and they can use the excuse of government regulation. Google wants to keep the data. They don't want other people doing retargeting and analytics against data that's on the website. And Google controls the Chromium browser, even though technically it's open source. So they told us two years ago that they were going to kill third party cookies, of course, to protect the privacy of the users. But this was inevitable. It wasn't just because the big tech wants to control the data. It's because GDPR starting in Europe was going to go across everywhere in the world so that the hiding of individual user behavior would happen so that data wouldn't always be passed back directly to the advertiser. And that does make sense. The idea of floc, federated location of cohorts and the idea of only allowing first party cookies makes a lot of sense. However, the way it's being implemented, as you can see from the various documents, things Google are saying, is that Google is going to turn all of their stuff, Google Analytics and the Google remarketing in the first party and everything else is third party. So it's a convenient way for Google to kill everybody else in the same way for Facebook to kill everyone else except for themselves. So now we have these different silos of data. And so us as advertisers, marketers, entrepreneurs, e commerce store owners, we need to consider what does that mean in terms of the ramification of the data, what are we going to do? And then how do we compete? Because when these changes happen, a Lot of our competitors are going to be competed out of business because their cost per click and their cost per acquisition is going to appear way higher and the old advertising techniques that are relied upon targeting won't work anymore.

Yeah, and that also means that it was that story, you know, with the two mice like who stole my cheese? That also means that we're all invited to change. I mean, when things are changing, we should be changing too, because that's the sign of intelligence. If you continue to do the same things, expecting for different results, then you might have, you will not might, you will surely have a problem in the, in the near future. I've been looking recently in the data from a few customers of ours at their Facebook spend, at their Google Ad spend, and for me it was amazing. It's like they don't have a business anymore. So one of them used to have 9.8x ROAS. Right. And now in the last month it was 1 point x 12. That was it. So that's amazing. The only thing that keeps the company alive is that they have a huge customer retention rate. And thankfully we're working together, we're gonna start doing something. But let's focus a bit on this particular example. I have a few solutions, usually have a few solutions. So let's presume that we are this company. What you would be doing if you would be in their shoes.

Well, first off, you can't rely upon roas anymore. We've seen most e commerce stores that we're working on lose 30 to 40% of the reported conversions. That doesn't mean the actual conversions are down 30, 40%. It just means the ones that are being reported are down because of tracking. And because of iOS 14 and 15, a lot of that data is not being passed back because people are saying, do not allow this app to track with the app transparency and tracking prompt, which is what that's called. When the cookie death starts to kick in, I expect the numbers are going to get far, far worse. So for those who are working on manual bidding to the cpa, they can't use the same CPA anymore because if the system, if you're profitable at a $50 CPA and now 40% of your conversions aren't being reported, you're going to have to bid up proportionately to compensate. So you're going to have to bid at $70. Right. Otherwise the algorithm can't find enough conversions and it'll just basically kill your campaigns. So part, the first part is the reported conversions are down and the second part is Some of the actual conversions will be down because of remarketing and because of date. Not just data that's not being reported on the conversions, but people you can't target anymore. And the more reliant you are upon retargeting and especially upon Facebook, the bigger the impact is going to be.

Yeah, one other important aspect Dennis, is that right now the attribution is messed up. So what do you do? And the trap is that you go to look like you work with your CFO and you look only at the bottom line and you can go to pour money into everything like SEO, social media, whatever, that's a route. But another route would be to look at the CLV to cac. So mainly you get all the spending you had. Or you could go retroactively. Yeah, you could look at your initial, your historical data so that you can see your CPA on various channels. And once you look at the CPA on various channels, you'll be able to extrapolate this and to look at some sort of proportion. Another way to work with attribution, to fix attribution would be to use post purchase surveys so that you can simply ask the customers where you've heard about us. And with this you'll be able to have some clues. If you would be doing these post purchase surveys one year from one year ago, that would be even better because you will be able to compare channel, I don't know, the social TikTok or whatever. So in this manner you'll be able to have some clues. What else? What do you suggest?

So like you said, we can't rely upon the close in roas and revenue and conversion data because a lot of it is going to like attribution doesn't work. So a lot of it's going to show up as unattributed or not tied directly to a campaign. So now we have to look at the overall picture. So overall spend on like a rolling 14 or 28 day basis because conversions that would have been reported today may show up two or three days later because of the change in the reporting or not show up at all. And you guys know that like Google with FLoC, right? The Federated location of cohorts, they are, instead of reporting back individual clicks and individual conversions, they're grouping up buckets of users. So to protect, right. To be able to anonymize some of these data by putting these people, by putting in the groups, which makes sense because us as advertisers, we don't, we don't need to know one particular user. We just need enough data to be able to optimize. So by anonymizing and putting people into groups, they're putting a delay. So the conversion, instead of being reported immediately, they're intentionally delaying a random amount of time from one to three days. So you can't look at ad level and campaign level data on a day by day refresh your campaign performance standpoint. You have to look at a weekly or 14 day or 28 day, like a bigger picture. And then you have to look at how much is coming through unattributed or how much is coming through email. Because often if you do social or you do display, then you'll still get the conversion, but it just won't show up. So then it'll show up in the last touch, which may be email, or it'll show up as type in or unattributed. Right. So we have to take the whole picture into account, like you said.

Yeah. So let's go back, Dennis, to this example. We have a business which is it had something like 6 to 1 customer lifetime value versus customer acquisition cost, cookies, whatever, iOS 15. Now in November it got something like 2.5 to 1. They are profitable only if they do 3 to 1. Yeah, they are breaking even at 3 to 1. What do you do in this under this condition? You look at the agency, you look at the ads, you look at where do you look to fix this business, to help this business survive and thrive?

Yeah. So the first thing that you don't do, which is what we see a lot of people been doing, is they just cut back spend because they say the ROS fell. So therefore we're going to cut the spend because to make break even we have to be at least a 3.2 or 3.5. We need to recognize that a lot of the data is not the same. So we have to engage in matching. If we can capture a lot of those conversions that we weren't seeing before because we're using the conversion API, because we're tying together email audiences, because we're being smarter about UTM parameters. So always want to use UTM parameters in our ads, always use UTM parameters in our emails, then you're going to start to find those conversions that were otherwise dropped by cookies, but you're able to match on an email address. So that will cause your reported overall campaign level roas to be higher than it was before because you're able to match back some of those pieces that you left next. And I've talked to some Folks from actually talked to Facebook this morning about this. There's a greater reliance upon winning creatives. We used to be able to do lots and lots of campaigns because we could micro target, but because we can't and a lot of that data is going away, there's a greater emphasis on a winner. So when you find a winning campaign, find everything you can to try to really focus on getting more conversions or more tweaking or more refinement against just a few key winners. I see. I'd love to see what you see as well, Valentin. But across the campaigns that we're looking at, there are fewer winners, but there's a greater investment in those particular winners to try to get more Right. So the campaigns look a lot more, a lot simpler because there's less data and there's more optimization on the algorithm.

Yeah. So my perception about the ads game is that if something goes wrong and if something is changing in the market, you should be looking at your winners. You should be looking closely at if you don't measure CLV to CAC ratio, you should be looking at that. You should be getting away from thinking that these conversions which are being reported are only those conversions that matter. So mainly your CPA is going up, but you should be looking broader. However, if you've made your homework, if you have the right data in terms of CLV to cac, if you've added more than the media cost so that you are sure what you need to break even and to be profitable, that then is the time to zoom in in terms of the campaign. So what I would do is I would fix the traffic, I would make sure that I'm tracking correctly. As you've said, I would be relying on the UTMs because this can't lie too much, let's say. And moreover, I would be asking myself if I have the right team by my side because the agencies that are not data driven, the agencies that used to just tweak the bids and change the creat, are not going to be the ones helping the business. The e commerce companies thrive in the next era. And another thing that I would be doing, I would be starting to look at the campaigns as experiments, right? So I'm doing this experiment, I'm using these utms, I'm changing the creatives, I'm looking at the zero party data. So I'm collecting data from Mike existing customers, I'm finding why they are buying my product, I'm finding their struggles. Right. I know that if I want to sell a product, I need to be way more relevant than I used to be. Right. Because before that, the algorithm was helping me to find the right customers. Now I have to tweak my messaging and my creatives and the offers as well, so that I'm way more relevant than I used to be. So that's my perception about how to move the needle.

Yeah. And there's the greater focus now on the creative with the targeting going away and us needing different kinds of creative. So instead of relying so heavily on images in a carousel, we can do dynamic creative testing where we have different kinds of combinations of headline versus A creative versus landing pages. And I was on a call with Facebook just a few hours ago talking about how they're shifting towards the optimization of creative, where they're even building in. I got to see some of the new things they're building where the, like, the tools that Canva has, like that kind of editing is being built into the tool because there's a greater and greater emphasis on on us inputting the right creative. And then the system is going to do a lot of the optimizing. So a lot of the fancy, like you said, changing bids or making lots and lots of ad groups and lots and lots of campaigns and lots and lots of targets are no longer effective. And the more people try to do that, the more ineffective it is because there's not enough data for the algorithm. Plus, the average cost we've seen on Facebook in the last year has gone up from maybe a $10 cpm to a 14 to $20 cpm. And for retargeting and small audiences, we see $30, $40 cpms, which I had not seen for several years. Right. Because the cost of traffic is so high, we get penalized increasingly if we attempt to subtarget. So the refinement of the content's the biggest change that Facebook's recommending. And we see that as well. Right. A fewer winners that we like. We just need a couple winners that are ideally 50 plus conversions per ad set per week. Right. It's not about 50 different things. It's about one winner that's driving 50 conversions. It's exited the learning phase. And that way we can try to get more and then we can work on scaling second piece, which we alluded to in the beginning. But I want to talk about a little bit more detail is the digital plumbing. So digital plumbing is getting all your data back together. So if we're passing through the conversion events, because we have up to eight conversion events that we can pass back through Facebook or maybe we're doing like a lead ad to collect an email address and then trying to sell through that, depending on how we're doing marketing, or maybe we're optimizing the lifetime value and we're passing through not just the initial purchase, but a marker for where they buy again or something. Then we're feeding a better signal into the algorithm so we can start to bid based on value instead of just bidding on the lowest cost per conversion. Because maybe in our store we have some products that are $10 and some products that are a couple hundred dollars. So if we just want to get as many conversions as we can for $20, the system may end up optimizing towards $10 purchases instead of the $200 purchases. So by looking at the CLV, we have to feed that data back into Facebook because if we're only going for the single purchase cart checkout conversion, we're going to miss all these other pieces. And if we can tie in UTM parameters, if we can run GA4 in addition to GA Universal, we're going to get more data. If we're tying in these conversion events as being triggered as part of Google Tag Manager, we're triggering the Facebook Pixel and Google Pixel and TikTok Pixel and Twitter Pixel and all of these at the same time. Then we're feeding more data into the system to allow it to optimize towards the conversion events at the maximum value that we say we can bid against. And I encourage advertisers to bid slightly above what their breakeven ROAS is or LTV over cac, whatever the breakeven is, but a little higher. Knowing that you're really getting an extra 30% more than what the system is reporting, which takes a leap of faith. But if you're tracking over time, you will see that. And we found this to be true accidentally. We accidentally ran a test where we had some campaigns turn off and we were able to calculate what the difference was with the campaigns on and off. And the falloff was much greater than the incremental sales that Facebook told us was happening.

That's great. That's a happy accident. So to synthesize, what you're telling is that you say that if you want to improve your E commerce marketing efficiency, let's say you need to get away from the idea of testing 25 different AD groups because you are not going to have enough time to validate and you're going to get penalized as this. The cost is higher right now and that means you have to get out.

Of the learning phase. That's the issue. The first question before we look at any kind of optimization is what percent of your spend is outside the learning phase? If you have more than 50% that's stuck on the learning phase, then either you're not spending enough money to be serious or you're bidding too low or you have too many campaigns going on. Until you fix that, it doesn't matter what else we're talking about with tracking.

Right.

The key is you need to have at least one winner. One, you know, one campaign that's at least 50 conversions per ad set. And then try to do everything you can to focus that and kill everything else.

Yeah. One campaign which is having more than 50 conversions per week, right?

Yes.

All right. So mainly if I want to start to change things around, if I'm being hit by this cookie deprecation, what I have to do is to go back to find why the customers are buying, to do research, to go in depth, to understand what's going on with the products and why people are really buying. Not what they say, but what is the main reason for buying that. Then we should be doing some great creatives and very relevant creatives. Not go to do 50 creatives. Let's say do two or three creatives. And then with these two free campaigns, the focus is to get more than 50 conversions to get past the learning phase so that your E commerce is going to relevant. Right. So mainly after learning, you will be in a way more better position to tweak and to change and to go gradually to change the creatives as you have a winner. Right. So is that organic spread of this virus? Right?

Yeah. So if we're bidding based on a true LTV and we're getting the 50 plus conversions, we're not hearing you anymore, Dennis. Oh, no, you can't. I'm looks like it's coming through. I see it's working on my end. Can you hear me?

I don't know what's happening.

I am not on mute. I hear you and I see my microphone levels in zoom. Looks good.

I don't know what's happening with the recording. Let's see. Can you say something?

Yeah.

All right. Okay. All right. So going further, what else? What can we do further? Let's say we've done these two free creatives is not working. What can we do if it's not working?

Well, if you have one. Well, first off, if you have one that's working, focus on why that is, and it's going to be some angle and something that's causing something about you're addressing a need or addressing the feature or addressing the price or it's funny in a particular angle angle about that creative that is helping Facebook do the sub targeting. Right. So, so remember, all this targeting is going away. So it's the behavior inside, the behavior on the creative that is telling the algorithm what to do. So then because we're doing surveys, because we understand the customer, all the stuff about, you know, know your customer, we're going to do other creatives that are thematically around concepts that we know absolutely are the reason why people buy.

Yeah.

You know, maybe it's something around a particular product, maybe it's a particular feature, maybe it's the way it's being used and demonstrated physically. One of my friends has a used golf ball business. Right. He grabs golf balls out of the lakes and whatever, and he sells them online. And he found the thing that worked the best was showing stories of an, oh, a new order came in and then showing the process of being packed and being shipped out. And he found that that theme worked well. So he made more and more of that one theme. So the campaigns, or I guess the ad sets really are being organized by the creative theme. And because we're not targeting. Right. Yeah. And so we're not just making 50 random videos or different copies or, you know, using AI tools. Right. Headlines, which is still fine. You could do that. It's knowing in advance what is going to work. Because you've done the market research to say, you know, because you've, you've talked for five minutes to every single customer and said, why did you buy? Was it that inspired you to buy? Oh, it was because this one, this one particular phrase, or the two or three things that people are saying, you're going to create creatives around that. And then it's a matter of finding the right way, the right treatment, the right visual, the right background music, the right person, but not the theme itself. Right. So it's the treatment of the creative and not the theme. And most people, when they're doing testing, they're mixing everything together at the same time. So you don't know what it is that creates the winning creative.

Yeah. One other aspect, Dennis, which I want to touch here, is that when you are trying to do a creative, once you've gathered the data about the customer research and you know the reason why customers are buying, you don't need to over invest into the production because that's going to eat a lot of your budget. And if you're Going to bet on a thing which is going to be viral and if you wait for, I don't know, five weeks to do the greatest production, you're going to fall into the same trap that I fall, I've fallen as well. When we got on the TV station back on 2010, we've invested something like 100,000 into a TV campaign. We've invested like 40,000 to make that video with actors and whatever. And the results were mediocre to say the least. So my suggestion, if you want to move the needle once you've got the insights about the consumer behavior and why they are buying, is to test out creatives which make them relate to when they are going to need your product or when they are going to be struggle struggling. Because the idea is that we need to make them relate. So it's not like when you need this, it's when I need to floss my teeth or when I need to X when I, when I'm in this situation. Because we are actually marketing for future moments when people are going to be needing our product and without our products they are going to be feeling incomplete, right? So with this product, I'm going to feel complete because I'm going to be in this future situation. But when we are, when we are sending this type of ads, this type of creative, this type of messages, they are scrolling on their phones, they are on traffic, they are not in the situation when they are going to be needing the product. So the idea is that we are always marketing for a future struggling moment that they've been aware that they are going to feel.

But those campaigns appear to have low ROI because you're educating or because you're marketing to the future event. But they actually work well if you do it this way, and that is you have short vertical videos on your cell phone that are 15 seconds. We are teaching people how to do something and that something is related to the product solving the problem, talking about the problem which your product solves. And then you remarket to the people who watched, you know a through play or watched 75% of the video, whatever that amount is. Because anyone who is watching the video about why your teeth are yellow and how it causes the girl to reject you when you want to go on a date, then they're going to buy your teeth whitening or the story about how you really want your kid to do well in sports and be able to score goals on the soccer team. And you tell a few of these stories of the mom that really is interfering with the coach because she really wants her son to score better goals or something. And then you can promote our coaching products to be able to hit the ball better or to be able to like our socks. We have these face socks with people's faces and we show stories of how to say thank you and how I want to honor a client of mine or whatnot. And so naturally, anyone who watches that video for 15 seconds clearly is going to be interested in purchasing my product. Now here's a test that we ran. Valentin.

Yep.

We did two things. One was we drove cold traffic straight to purchase. Right. Which makes sense. Like you want to optimize to a sale. We, I think our target CPA was $30. Right. So as many sales as we could and did.

Okay.

And then we tried a two campaign strategy. One is telling the story, you know, about, oh, how the mom loves the kid or about some kind of story that relates to the product that anyone who would watch that story, the theme of it is such that we know that they would then want to buy the product. So then we got these video views cheap for like half a penny. And then we remarketed to the people who saw that initial video, the quality score factors and the base CPM and the cost per click and the click through rate, all of them was so much better in that audience by warming them up that the full cost of the two step campaign was a lot less than the cold straight all the way to a sale kind of campaign. And then we found that in the, in the second campaign because initially we ran against just retargeting audiences because we figure, oh well, if they see the first story, then we're going to show them the next part where they're, you know, maybe like you said, they're driving the car or they're walking around, they're not ready to buy right now. But you know, we just caught them in the supermarket. We just interrupted them just for a moment, but they didn't have enough time to buy. But they saw it was interesting, you know.

Yeah.

So we thought, okay, we're going to remarket to that. And so remarketing does work, but the base cost, the traffic is so high because the smaller the audience, the more Facebook is going to charge. Right. Then we found that that second campaign for conversion, we could just target everybody and we didn't have to do retargeting. And because of the way Facebook optimizes to the conversion event, it is automatically building the retargeting into the audiences. And that is something I didn't see a couple of years ago. So both sets of camp the mid funnel, tell the story, engage, entertain them, just to get them to watch for a certain amount of time. And the conversion campaign, both are untargeted.

That's great.

The net cost of both of these together per conversion was a lot less. You know, maybe it's like $22 instead of $30.

That's fantastic. And also you have a lot of other people in the pipeline because you are, let's say training them to want the product in the future. So it's like a compound effect that you are, you are assisting and I think few e commerce companies and few e commerce leaders are aware of that. They need to act like movie directors, they need to build scenarios, they need to orchestrate different experiences throughout the entire customer journey. And we have the initial customer journey. When you are a nobody, right, you are, you are a new company which is entering the market. You are the new guy which came to the party where we already were. And in order to be there to be present, you need to identify how you need to present yourself to the market. So what is the first touch point of the product, how you are coming to the market. You need to educate about your product and then as you've said, you need to do step one, step two, step three. And what happens if this happens? Right. So what if the audience is not engaging? What if the audience got to my website? How do I remarket that? And I've seen that in YouTube for instance. Now you can build this type of scenarios like this. I've seen video one and then I've seen video two and then if I'm seeing video two and video three, I'm going to go to video five. So mainly you are actually building these type of scenarios.

Yeah, building these different sequences like a. Choose your own adventure, like a funnel of many steps. But what we've noticed is that we can not everyone is going to always go step one to two to three to four. Some people will jump step one to three or you know, one to four to five or two or any sort of combination of these different pieces. So we find that it's not necessary to use those kinds of retargeting audiences or we can put them in there as one ad set just to let it run and see what. Because the conversion rate is higher, but the CPM is also higher. So the net CPA might not be as good or maybe it's good, but the volume is low. But you're going to find that most of those conversion, those bottom of funnel Conversions are actually going to come against large audiences. So Facebook knows how many people consumed any kind of engaging how do you do this pre selling kind of content and the conversion? We don't know that, but we know Facebook knows that and they can figure out what those combinations are. If we are tracking events like the view through on video content, for example, that's actually informational, or a page view where we send them to a blog post on 10 Ways on How you can pay less taxes and you're selling the tax software, something like that, then that's another way of a stronger signal. But we have to build one instead of just conversion. I see this all the time. I look at so many of these carts that everyone's optimizing just to the cac. But if you can optimize to one step above, like collecting the email or as simple as like watching a particular video for 15 seconds, then you're optimizing your creative for. And so it's really, if you have 10% of people that watch to 15 seconds versus 30% of people watching the 15 seconds, that's a huge difference. So that, that tells you where to optimize on your creative.

Yeah, and I think that's amazing because it's the same as in CRO when, when you want to improve the macro conversion, which is the same you are actually looking for the micro conversions, how many people are getting from the landing page to the product page and from the product page to the checkout. And that's why you need to look at all these steps in the customer journeys or all these touch points are what matters. Because if you don't look granular on the micro conversions, you are missing out the micro conversions. And it's an average, right? So if you look at the conversion rate, if you look at the ctr, if you look at any kind of parameter is going to be an average. And the trap of the average is that it's not allowing you to see the full story. And one important aspect which I'm seeing here, Dennis, is that we can focus on this type of micro conversions because also we are having enough sample data to validate if Creative 1 or Creative 2 or Creative 3 is the best in pushing people like in a domino effect, right? Is this moving the first part or the second part? The second step better? And in this manner you're getting going to get passed through this learning phase because it's all about learning how to act, how to message and how to market the product. If you, if you've learned that's it then you're, you're, you, you can't unlearn, right? It's like, like in life. That's why you need to be into a continuous learning mode.

Most people, when they think about customer journey, they're thinking about stages of retargeting in a pure web format. But the thing that I see a lot of e commerce people missing is email. And if we capture the email along the way and we know what the emails work. So my friend Colin Wayne Ehrman runs Redline Steel. He's the largest manufacturer of home decor items. So like a flag and metal or your initials for your home that you hang like these steel cut items right in this big factory, 110,000 square feet. And initially he was doing okay, he's doing a few million dollars a month in sales. But then he implemented this thing called $2 Tuesdays. So on Tuesdays he had these $2 items which the cost of his items cost almost nothing, I can probably tell you to print them. And so people would get this $2 item but they had to sign up for the email and of course they had to pay for shipping. So he actually wasn't losing any money by selling these items at $2, even though the list price of these other items might be 30, $40. But you can sell the first one at $2. But the key is then, because it's $2 Tuesday, everyone's in anticipation of oh, every Tuesday, what's the new item? Oh, this time it's a Christmas item. This time it's a wreath. This time it's something you hang over the toilet. This time it's this thing in the kitchen. Right. So he has this excitement and he's creating touchpoint frequency. Most e commerce players, they have so few touch points that they just have a sale. And therefore there's no extension to CLV because there aren enough touch points prior to that initial sale and there's no ongoing touch points. So if we can just get more touch points using content that's relevant. For example, he's segmenting by people who are veterans or people who are moms and he's putting different, different content down those particular pathways. But the key is he's Facebook Ads is his number one acquisition tool, but he's running, he's found that he can switch his optimization to lead. So he knows if he acquires that lead at $2 or less because of the CLV, it's very profitable. Before he was bidding just to CAC or to cart checkout, which is a third of whatever the conversion is from add to cart to conversion. But then he decided, hey, there's so much more data. I can feed the system by running to the lead, by collecting the email address. So then if we go lead, then eventually to cart checkout and then to conversion, we're giving the date. Like how for every one actual initial purchase, we have three people that added to cart and then we have.

15.

Like 10 to 15 people that gave us an email address. So we're getting 15 times more data by optimizing, by having the bulk of our spend optimized to the lead. So I'd ask all you guys, do you have some event that's one stage up in the funnel or two stages up that's around email collection that gives you the permission to send people new examples every week There's a client which we've openly shared the analytics and the Facebook ads and the Google Ads. So you guys can see how we've applied. It's called Seeds of Life. Right. They sell trees, trees that you plant for a loved one or flowers where you're honoring somebody or saying thank you or your grandfather died. You know, all these different situations to buy trees and flowers. And we found that the frequency of our communication was super low because basically people would buy one thing and then we didn't continue that. We would send like a thank you email. But we didn't say things to like, hey, share this with a friend to get 20% off. We didn't say, hey, here's the newest things in stock or if you like this, you also like that. Or did you know that this upcoming holiday, here's a special story behind this holiday. So all these different ways to be able to keep the conversation going even before they purchased so we can continue to be aware. So then when they think, oh, I need to do something special for my wife. Oh, yeah, yeah, yeah. I've been, I've been watching the sequence and my friends at Seeds of Life have been sharing these cool new products. I'm going to buy this gift basket. I've been looking at the. Because every week they show me what the latest gift basket is. So I don't see E Commerce players optimizing to email. Yeah, right, because they're optimizing straight to the checkout on Shopify. What? Look. What, look at what happens when you optimize to email. You get 15 times more touch points. You double or triple your CLV. You can bid a lot more because you have more touch points. You have more campaigns that can exit the learning phase and you can scale up and you're also not as reliant upon the cookie data because you have the email address and when people click on the email, you have the UTM parameters. So you can track them between the web and your email.

Yeah. Moreover, you can also ask them things. I mean, when you collect their email address, you can also ask them what's important to them. Let's say with Seeds of Life, you can ask them things related to the product or things related to the person that are. That you want to honor with that. And once you have this data, your follow up is going to be way more effective because it's going to be informed and relevant to the person. What I also want to touch here, Dennis, is a thing regarding to the triggers. So we are habitual, human beings are habitual and we want to do things, but we are not aware about the things that we are doing. So we are not always aware because the focus can't be on all the things that we are doing. So that's how we are brushing our teeth. Mindless, completely thinking about other things. So what you've stated that every Tuesday you have this $2 thing. Those things can be replicated because it's about when. So what I think from all the questions, I think when is the most understated question. When do you need the product? In what situation? When you are going to be needing the product. And what is the trigger? So if you can connect your product to a certain trigger. So every time you do this, every time you, every time it's when you wake up in the morning, when you go to the gym, when you, when you go out with your friends, when you are going to close next time. So if you can, can connect the product with the thing that it's already there. Right. A habit which is already there every time you close your phone. Think about, I don't know, our books, the fact that you, you haven't read a book. And if you can install this, this is going to be like a seed in the mind of your consumers. And that's when they are going to remind about your product.

You want to hear a wild example? Valentin, of course hit me yesterday. I was in Atlanta and I spoke at a conference with a bunch of lawyers that are actually very sophisticated. One of these lawyers spends $24 million a year on ads and he makes $160 million and it's personal injury. So when someone gets in a car accident, you call this lawyer and the lawyer basically sues the insurance company to get as much money as they can, right?

Yeah.

So what they're trying to do is find when people get in a car accident, but before they were running ads on Google related to car accident or a crash. But it's very expensive, $100 a click because all the other attorneys are bidding on the same keywords. Very expensive on Google. So some of these attorneys, I looked at their campaigns because they opened it up and let me take a look. And they are using social media, using the campaign strategy that we just talked about, to be able to get in front of people on a high frequency. They found that their ideal purchase frequency, or not purchase frequency, ideal frequency for education, was about 4.3 per campaign per month. So they're showing messages like you said, when you go to the park, when you're in the car, when you go to the grocery store, when the police harass you, when you get a speeding ticket, when you get married, when you go to the gas station, they have created triggered content about that. And this is maybe not so surprising, but Facebook is able to figure out when to show that content. There's no targeting. We're not targeting when people take their dog to the park. Facebook knows. So these attorneys are literally making content like, hey, when. When you take your dog to the park and another dog bites you, bites you or bites them, you know, here's what the rules are. If you get bitten by a dog, you need to go to the hospital right away to get the examination so we can get the medical thing. You need to then call me as the attorney so I can make sure that they have insurance and we can get your proper medical care. And then you're like explaining the whole thing. But instead of buying the Google keyword in the park, being bitten by someone else's dog or dog bite attorney, they're making content knowing that the social network is going to target around that situation that occurred. I thought that was genius. But then a lot of people say, yeah, I open up my phone, I talk to a friend or a friend tells me about a product and I see an ad for it, even though I've never been there before. It was listing on my phone. Don't you think that the system is smart now?

Yeah, for sure. This works. I recall a strategy that we had at my former E Commerce about these moments. Yes. Not when, not who, but when. And we've been into a difficult position. Like my some competitor got into the market, the CPAs were off the roof. We've made the calculations, we've realized how much we could pay, and we realized, hey, why don't we use our existing customers to leverage the network effect and what we should be doing, We've made some brainstorming and we realized that we know their birthday and we should be sending a birthday cake so that they can feel the trigger, the reciprocity. Right. So we, we sent 3,000 birthday anniversary cakes and the effect was that we were doing this post purchase survey. How did you. Have you heard about us? And it got from 16% to 60% and our sales got to 40% higher by leveraging these 3,000ambassadors. And next time we looked at this cohort, the retention rate for this cohort was 91%. So next year, because they were renewing, we were selling car insurance. They were renewing their car insurance every year. So from 58% to 91%, that was the customer retention impact plus the network effect. So when you are using when, because next time, when it was their anniversary, they were renewing their insurance as well. So that's why you need to use these kind of things and to make the, the people aware that you exist because that also triggers their emotional state differently. Right. So they will associate your company with their anniversary, but it shouldn't be only anniversary. You need to make this pair. Right. What they are associating, what is the moment that they can associate my product to?

Yeah. And then having enough frequency so that people will remember, because people, they see your message one time or they see it as an ad, they're not going to remember. So we have to hit them four or five times with any one of these particular messages. And then if they stay and watch that video for 15 seconds, that's a good sign that they're interested in that. And not only does that allow Facebook to automatically retarget against that person without us even needing to build a retargeting pool, but it helps Facebook find other people that are just like that person that watched that video for 15 seconds plus. So lookalike targeting is now built into broad targeting.

Yeah, I think another, another strategy and going, going a bit up. Dennis. Like we've, we've been talking about how to tackle the problem of cookies by doing better ads. But it's not only the acquisition part, it's also the conversion part and it's also the retention part. Because these are the three pillars of customer value applic optimization. So if you are not doing well in terms of roas, if you are not doing well because of the cookies, maybe it's time to look at the other components of the whole customer journey. Maybe you should be leveraging the conversions. Maybe you should be leveraging the customer retention as well. Because looking at the whole customer journey and identifying how can I improve the customer lifetime value now that the customer acquisition cost is too high? And only by asking yourself this question, you're going to come up with a lot of solutions while the people which are being trapped into the old model, they don't look at the whole customer journey and they ask themselves how can I leverage acquisition better? And they go to hire a new PR agency, they are looking for other channels, they are investing in TV or whatever they are doing to improve the customer acquisition. But I think now is the moment to switch the paradigm and to understand that it's about the whole customer journey is not about acquisition anymore. And if you are going to get stay trapped in the old model, you're going to become obsolete in a few months from now because the market is changing quickly. And I've been saying this, Dennis, so 44% of all the customer the E commerce companies that are now live have been launched in the last 12 months alone according to Bitwidth. So the competition is huge. The inventory of the ads is the same. So it's not like Facebook is having 44% more inventory in the last 12 months. And that's if, if the demand is higher and the offer is the same, then that means the cost is higher. And if the cost is higher, you need to change your strategy, you need to make it more effective and you need to focus on how much you are getting from any every customer so that you can justify such a high cpa.

Yeah. And the shift that you mentioned to CLV which is based on acquisition conversion retention means that all of us in 2022 need to have two key people in house. One is a creative person that ideally can work with video that can run this process where we're collecting the feedback from these customers at different points and recycling the UGC back into our ads. So those our campaigns are driven by that UGC like we have with the sock company. Number two is we need a smart data person. And this data person understands the different touch points. They understand the email sequencing, they understand where the implied triggers are so we can build more content. So the data person, the content person I believe must live in house. And I think a lot of agencies that do E commerce are going to be flushed down the toilet because they just won't, they won't know how to do that for the client.

Yeah, another important aspect, Dennis, I think is going to happen in the next year or the next years as well, is that in the future I don't think that the agencies which are narrow minded and highly specialized in a single channel are going to be able to make it. And that's because you need to need more, you have to focus on more. You can't nail attribution like you used to do anymore. You can't do things if you are not looking at email, if you're not looking at the other PPC channels as well. So the agencies of the future are going to be data driven and are going to take into account customer lifetime value as a metric, even though it was ignored for decades. They will be forced to transition into a model where they can understand data. They can understand things like payback period. Right. If the, the client of yours, if you're an agency and if your client is not making money because of the fact that he's not educated enough to understand the balance between customer lifetime value and customer acquisition cost and the payback period, you have to understand that because you are in the business of providing value to your customers, if the value is not going to be there, you're going to be out of business. And this payback period, for instance, it's so important. If you're in E commerce, your payback period should be less than six months. So what you're paying to acquire customers must come back to the company faster so that you can invest faster, so that you can grow your company. Because Even though your CLV2 CAC is fantastic, but if you get your money back in one and a half years because you're selling, I don't know, laptops and the purchase cycle is not that fast, then you're not going to make it, your client is not going to make it, and you as an agency, you're going to be out of business as well.

Yeah, I agree. The specialists are not going to make it because they're also being competed down at Fiverr. And you have these people that are working for very cheap offshore where they're doing a particular task like video editing or they're doing social media posting and all these, there's a lot of tools that will do a lot of these sorts of things. So the people that are doing things that do not have a direct impact on clv, not because we theoretically think that creating video will increase CLV or we theoretically believe that switching to this new tool just because will increase clv, but by actually adjusting our segments, adjusting our copy, adjusting our data, or switching budget from Facebook to TikTok because we see Arbitrage with a traffic cost is, instead of $12 is $3 per thousand impressions or whatever it might be. That kind of balancing and arbitrage is what we need to do instead of being channel or silo specific. So the agencies of the future are focusing on customer Personas and being able to advance people through the clv, through a customer journey, which means they have to go across all the other channels. And the paradigm before was you had people that just did SEO, PPC or landing pages or video editing and so they got really good in one particular narrow tool, but they didn't understand how it all fits together. So now the horizontal markers that go across multiple things to understand the customer override the depth of being in any one tool. Because if you're really good vertically, then you don't, you can't go across horizontally. If you're really good horizontally, you can't go vertically. But we find that vertically or horizontally is where everything's moving because the algorithms are going to handle all that kind of complexity. The tools, the automation, the AI that automatically does targeting, that figures, engagement, that determines who should see the next message in the newsfeed, that's all handled by the social networks. So we don't need to have this kind of tool, deep expertise anymore.

Yeah. And I think the opportunity is quite amazing in E commerce right now because this is work in progress. I've been looking at some statistics from Benedict Evans. He shared that in the last two years, the spike in a month in UK, the online retail was 40% of the whole retail for two months in a row. And the fact that this world is being, I don't know, transforming is affecting positively the E commerce players. And another thing which is actually important is that the very small and the small players in the cpg, for instance, are being positively affected by that. So if you're doing CPG or if you're doing marketing or if you have clients in the CPG, you are in the right place because now 75% of all the sales in the US for CPG brands, the online sales have been made by very small and small companies. And that's, that's a good sign because competition is always better for, for, for the client.

Yeah, it's great for all of us. And I love all the statistics that you have from Benedict Evans. And you know, 78.3% of statistics, Valentin, are made up.

That's great. But we don't know if that statistic is made up or not. This one with the 78.3. So there's a 22% chance that this is. No, it's a 78% chance that this is made up as well. So if you go down the rabbit hole. All right. So, Dennis, thanks a lot for this, today's episode. Any last thoughts for this one?

This is a great opportunity for all of us to be making more money. So we need to jump on it while everyone else is panicking. You guys see the online sector is going through the roof. All the tech companies, all my friends that are doing E commerce, they're making so much money because everyone has moved into digital. And while you're not going to hear a lot of these stories because it's going to sound bad if the economy, if there's problems or whatnot, and our E commerce friends are making a lot of money, they're not going to want to go boast about this because it sounds bad if other people are suffering. But I would like all of you guys to be making more money. And by applying the techniques that Valentin and me and other experts are sharing, I would love nothing more than than to hear what you guys are doing and what's working for you. So we can feature you in one of these episodes.

Excellent. Great. Then thanks everyone for watching. And we are going to see each other in two weeks from now. Till then, all the best from Bucharest and happy holidays, everyone.

Happy holidays.