4 Proven Automation Plays That Consistently Grow Ecommerce Profits

Episode Description

In this Automation Layering Masterclass, Matthieu Tran-Van shares four tested automation plays that consistently grow ecommerce accounts.

Episode Takeaways

Matthieu walks us through four automation strategies that have reliably fueled e-commerce growth for his clients, going beyond basic smart bidding to what he calls “profit-driven automation.” These aren’t just theoretical—they’re real tactics tested in live Google Ads accounts that you can confidently scale.

The conversation shows that automation doesn’t have to be intimidating or feel like a “black box.” When used thoughtfully alongside human insight, it becomes a powerful tool that takes the busy work off your plate while consistently delivering profitable results.

1. The philosophy of automation layering

When it comes to advertising, automation isn’t about replacing people. Rather, it’s about helping them do their best work. Instead of spending hours on repetitive tasks, marketers can use automation to ensure consistent execution while focusing their energy on strategy, creativity, and new ideas.

Fred explained that automation works best when layered on top of human insights. Practitioners should test, explore, and discover what works, then let automation carry out those winning approaches at scale. This way, good ideas don’t get lost to day-to-day busywork.

Matthieu Tran-Van echoed this perspective, stressing that automation doesn’t mean losing control or being lazy.

“Automation does not replace the strategy itself, but it can truly amplify it by executing it with perfection on a daily basis. It frees you from babysitting anything in your account so you can truly focus on the real strategy. And I think that in an AI-driven age, strategy—our strategy and creativity is our competitive advantage as humans, so we should focus our energy, efforts, and brain juice, you know, on this.” explains Matthieu.

2. Evolution of match types in modern PPC

Search advertising has changed a lot over the years, and so have match types. What used to be a clear choice between broad, phrase, and exact is now much simpler: phrase match is slowly fading into the background, leaving broad and exact match as the two types that really matter.

Broad match is powerful because it gives you huge reach, but it’s also risky since it can easily lead to wasted spend. Matthieu Tran-Van pointed out that its success often depends on how much you’re spending. In his analysis of more than 30 accounts, he found that bigger budgets tend to get better results with broad match.

“I have actually studied more than 30 accounts in my MCC across different industries and of course all running on smart bidding and what you see is that broad match basically is not working the same way for all advertisers. Long story short, the more you spend the better it is. Given the reach that you can have with broad match, to me now the rule is simple.” shared Matthieu

Fred added that this shift is making older campaign structures, like the alpha-beta model, relevant again. By pairing broad and exact match, advertisers can experiment widely, figure out what performs best, and then double down on proven winners.

3. The purgatory rule for keyword management

Matthieu shared a practical system he uses, something he calls the “Purgatory Rule.” The idea is to let automation handle the heavy lifting of keyword management by constantly testing, refining, and shifting keywords between match types based on performance.

“The idea and the key here is to let automation do the heavy lifting. So you can start broad, collect some signals and then let automation automatically shift keywords into exact match for instance only when they prove you know return on that spend.

Now if a keyword that you have in broad match overperforms then automatically it’s going to be graduated in exact match because you want to take specific control on this very important keyword for you. If the keyword is underperforming, then I’m going to automatically test it again, but with a different match type. And I’m going first to test it in phrase. And if it’s still under performing, I’m going to test it in exact match. And again, if I see no performance, I’m going to pause it.” explains Matthieu.

This process is paired with another rule at the search term level. Highly profitable search terms that show strong conversions and a return on ad spend above the campaign average are also promoted into exact match. Together, these two rules create a layered system that steadily filters out underperformers while locking in control over top-performing terms.

4. DSA campaigns built from business data

Dynamic Search Ads (DSAs) can be a real game-changer when built the right way. Matthieu explained that you don’t need a product feed from Google Merchant Center to create them. Any business spreadsheet can work. As long as the sheet is refreshed daily, Optmyzr’s Campaign Automator can take that data and instantly generate highly targeted campaigns.

“The Campaign Automator can give you thousands of very tailored ads without the headache I would say. And you know it’s actually simpler than using the ad customizer that you have in the Google Ads UI.” shares Matthieu

The results speak for themselves. In many cases, DSAs built with Campaign Automator not only produced highly relevant ads but also outperformed traditional non-branded search campaigns. Advertisers saw stronger conversion rates, better conversion-by-impression, and a higher return on ad spend.

5. Profit-based bidding vs ROAS optimization

Advertisers often focus on Return on Ad Spend (ROAS), but Matthieu argued that ROAS doesn’t always tell the full story because revenue doesn’t necessarily equal profit. True optimization comes from looking at profit directly, especially the balance between revenue per click and the incremental cost per click.

“When your revenue per click or the RPC is higher than the incremental cost of that click, of course you are earning money. You’re making profit. But when the revenue per click is lower you are losing money. So the sweet spot is exactly where your your revenue per click equals the incremental cost of a click. That’s where your profit peak basically. The goal is not to maximize return on spend at all cost. It’s to maximize profits.” explains Matthieu.

To manage this in practice, he uses a matrix framework that guides bidding decisions. For example, if profit trends are slipping and the revenue-to-cost ratio is declining, bids should be cut significantly. Conversely, when the ratio looks healthy, bids can be maintained or increased.

Because incremental cost per click isn’t available as a Google Ads metric, Matthieu uses ratios as a proxy. Ratios don’t give an exact number, but they clearly show whether performance is trending up or down. This makes them a powerful directional guide for smarter bidding. He shared that by applying these profit-based rules, some accounts doubled their profits without sacrificing efficiency.

Fred added an important nuance: not all businesses optimize for profit alone—some may prioritize growth or revenue.

“Profit is is good but there are some businesses that don’t necessarily care about profit. They might be in growth mode. So they might want to optimize revenue. But ultimately you want to understand what is your actual business goal and then understand how that mathematically corresponds to how Google Ads works.” Fred says.

6. Importance of meaningful metrics

Not all metrics tell the full story of campaign performance. Both Fred and Matthieu stressed the importance of looking beyond surface-level numbers like conversion rate. While a higher conversion rate sounds good, it doesn’t necessarily mean the campaign is driving enough total conversions to impact the business.

Fred highlighted the value of tracking conversions per impression—a metric you have to calculate yourself, since it isn’t provided in Google Ads.

“What I like is that you’ve included the the metric of conversions per impression which is not actually a Google metric but you had to calculate that. And it’s really important because you might say, ‘Hey, look, my conversion rate is more than double and my ROAS is more than double. And it’s like, that’s great, but maybe you sold two instead of a 100. So like obviously that’s not good for business, right?”

If you bring it down to conversions per impression, that is really a metric that I think will eventually connect to profitability because yes, in exchange for boosting your conversion rate, you may be buying fewer clicks, getting fewer impressions along the way, but what ultimately matters is how many conversions they still drive." explains Fred.

Matthieu agreed, noting that conversions per impression encapsulates the full value chain. It ties performance back to actual profitability and helps advertisers understand the real effectiveness of their campaigns.

7. Full business data integration and automation

Matthieu shared a case study that showed how complex business realities can be turned into a fully automated advertising strategy. The example was a retailer selling swimming pool equipment—a highly seasonal business where performance depended not only on weather but also on margins that varied widely across thousands of SKUs.

To handle that complexity, Matthieu and his team gathered multiple layers of business data: competitor pricing, product margins, and inventory priorities. They translated all of it into custom labels in Merchant Center, which allowed them to build a campaign structure aligned directly to business goals. Every product could be categorized by competitiveness, margin, and stock level. Something that Matthieu called the “perfect trifecta” of right assortment, right budget, and right bid strategy.

On top of this, they layered in smart automation. Campaigns were designed to activate dynamically based on conditions like weather and seasonality. For instance, bestsellers would automatically be promoted in sunny cities where demand was highest, with geo-targeting adjusted in real time. All of this ran without manual intervention. Humans only had to set the strategy, and automation executed it consistently.

The results were dramatic: revenue uplift ranged from 100% to nearly 300% in some months, achieved without adding extra manual work. The case study underscored how integrating full business data with layered automation can transform even a commodity retailer into a strong performer.

8. Key success principles for automation

For many advertisers, Google Ads automation feels powerful but also a little intimidating. Matthieu explained that the key is to ground automation in principles that are proven, repeatable, and rooted in real account data.

“The most impactful automations are actually tied directly to some kind of business output, profits, margin, or demand signals. But it’s not about click and impressions. If you can tie your automation to something that is truly meaningful for the business, you have a high likelihood to actually win.” explains Matthieu

Matthieu emphasized that tools like Optmyzr aren’t “magic wands,” but orchestration layers. They take human strategy and make it scalable, consistent, and profitable at a much larger level.

The real takeaway is that automation is most powerful when paired with human creativity and strategic thinking. By testing, refining, and tying automation to the right business outputs, advertisers can scale not just their accounts—but also their profits with confidence.

Tools & Resources

Campaign Automator : User Guide

7 PPC Automations That Save You From Burnout

Episode Transcript

Frederick Vallaeys: Hello and welcome to this episode of Automation Layering Masterclass. My name is Fred Vallaeys. I’m your host. I’m also CEO and co-founder at Optmyzr, a PPC management tool. For today’s episode, we have the enormous pleasure of having a returning guest. Matthieu Tran-Van is back. He has done one of our most favorite episodes of Automation Layering Masterclass, and he’s got some new material to share on how he’s using automation layering to drive growth for some of the companies that he works with. Matthieu is a leading expert, definitely also on the rise in the field. He’s an ex-Googler, so he’s got really good stuff to share. I’m very excited about this episode. And with that, let’s get rolling with this episode of Automation Layering Masterclass. Matt, welcome back to the show.

Matthieu Tran-Van: Hello, Fred. Nice to meet you again and thank you so much for having me again for a second episode on the Automation Layering Masterclass.

Frederick Vallaeys: Yes. Well, you always share good stuff and you’ve been busy, right? So for the folks who didn’t see the last episode, tell them a little bit about yourself and your book that you’ve written, your top 100 PPC influencer, your days at Google.

Matthieu Tran-Van: Yeah, sure. So I’m Matthieu Tran-Van, an independent Google Ads expert from France. You can hear it with my beautiful accent, right? Today I manage around 26 million euros in ad spend. Before that, I spent 10 years at Google helping the world’s top 40 advertisers with more than $350 million budget—big chunk of money, right? And yeah, I’m also the author of the Google Ads Strategist Handbook. And in 2025, as you said, I was ranked also among the top 100 PPC experts worldwide. And since the person interviewing me today is ranked number one, I would say that I’m in excellent company, right?

Frederick Vallaeys: I think between the two of us, we’re going to do some amazing things here today. It’s really all on you, right? So Automation Layering Masterclass, for folks who haven’t watched these before, it’s about taking a look at the principle of using all this amazing automation from the ad platforms but using your human brilliance to take it to the next level and then layering in your own strategies and your own automations to control those. So yes, you’re very kind, Matthieu, but you’re the one who’s going to blow our minds here today. I’ve had a preview of the slides, so why don’t we get rolling with those? We’ll put them up on the screen. There’s a couple more points. Maybe we have an intro slide right there.

Matthieu Tran-Van: Yeah, sure. So today, for the second episode, I thought that maybe we could talk together about what I call profit-driven automation. More specifically, I thought that maybe I could share four proven automation plays that consistently have driven e-commerce growth for my clients. And to me, the key message I would say is that when I speak with e-commerce advertisers, it seems that Google Ads automation—they understand that Google Ads automation can be incredibly powerful, but at the same time, for many of them, it feels a little bit intimidating and even sometimes a little bit scary, right? So my goal today is to show that automation, especially with tools like Optmyzr, does not have to be a black box. And I wanted to precisely share those four proven e-commerce plays, right?

So I suggest that maybe we can look at those four automation strategies that, again, in my opinion, you can confidently scale because they are tested, repeatable, and more importantly, rooted in real Google Ads accounts. That’s the important thing to me. And to make this even more relevant, I have actually structured my presentation around the framework that you know very well, Fred, because it comes from Unlevel the Playing Field, the book that you have written. I’m a big fan of it. And so the framework goes from, I would say, the simplest to the most advanced tactics.

Maybe we can start with strategic targeting, where we are going to discuss about keywords more specifically. Then moving on to strategic messaging, where I’m going to highlight, again, one of my favorite types of campaigns—DSA. Strategic bidding and speaking on how you can include in your bidding automation some profit-driven aspects thanks to Optmyzr. And finally, one of the most exciting cases, in my opinion, is strategic automation, where it’s really the next level of automation. We are going to see how we turn entire business data feeds into a full ad strategy from end to end. So basically, that’s the journey I propose you follow today. Does that sound good for you, Fred?

Frederick Vallaeys: Yeah, sounds amazing. Let’s start with—and even the basic stuff here is going to be relatively advanced. This is not an intro-level class, so even that’s going to be very good. But I love the framework, obviously. So yeah, let’s see what you got for strategic targeting.

Matthieu Tran-Van: Okay. So let’s start with targeting. And here, my idea was to specifically share with you an automation tactic that can actually make broad match behave, right? Because that’s one of the challenges for PPC practitioners. The thing is that when you think about modern search, to be honest, now the match types have truly evolved since we started PPC, Fred. Now match types have nothing in common, right? And today, honestly, I think that phrase match also is a little bit fading in importance. And there are really just two match types that truly matter now—it’s broad and exact match, right?

But the challenge that you also have with broad match is that it brings you massive reach but also a lot of wasted spend, to be honest. So the idea and the key here is to let automation do the heavy lifting. So you can start broad, collect some signals, and then let automation automatically shift keywords into exact match, for instance, only when they prove return on that spend. And so that’s the tactic so you can get the best of both worlds—meaning reach plus control, right? Because the exact match superpower is that it always has the top priority in that serving, so you can have more control.

Frederick Vallaeys: Yeah, this is fascinating because there’s the alpha-beta structure—it’s got various other names. And I think it’s lost a little bit of favor over the years. But now with this evolution of match types, as you’re saying, and really it going down to broad and exact are the two—and phrase, yeah, it still exists, but maybe it’s not quite as relevant anymore once you have exact and broad like that—alpha-beta methodology and sort of culling what works might actually be super relevant. So I think this is really interesting.

Matthieu Tran-Van: Yeah. So here we are not going to discuss about the structure itself because alpha-beta, I’m also a big fan of it. I have a lot of accounts that actually follow this structure. But here it’s more about the process that I execute thanks to Optmyzr on autopilot that, again, tries to prune a little bit those keywords. Because, you know, to illustrate what I was sharing earlier regarding broad match and the amount of wasted spend that you can have, just to illustrate this phenomenon, I have actually studied more than 30 accounts in my MCC across different industries. And of course, all running on smart bidding. And what you see is that broad match basically is not working the same way for all advertisers. Long story short, the more you spend, the better it is, of course, right?

However, given the reach that you can have with broad match, to me now the rule is simple. You should, every time you can, start broad and then automate what I’m going to call here keyword pruning. And to do that, I use a heuristic that I call the purgatory rule. So I don’t know if it’s a good marketing term, but that’s how I call this. And it consists of automatically looking at the keyword performance and then making some decisions according to that. So if a keyword that you have in broad match overperforms, then automatically it’s going to be graduated to exact match because you want to take specific control on this very important keyword for you.

Frederick Vallaeys: Let me pause for a second. So just to make sure everyone fully understands, you’re saying you’re looking at the keyword as an entirety or you’re looking at the search terms that derive from that keyword?

Matthieu Tran-Van: Yeah, very good question. So here I’m looking at the keyword itself. Okay, so that’s the keyword, not the search term. Okay, so I look at the keyword entity. And if a keyword overperforms, I’m going to add it in exact match. Even if, yes, you’re right, a broad match keyword is going to give us dozens or hundreds sometimes of different search terms. But I want to automatically add this keyword that I have in broad match in exact match. Okay? And at the opposite, if the keyword is underperforming, then I’m going to automatically test it again but with a different match type. And I’m going first to test it in phrase. And if it’s still underperforming, I’m going to test it in exact match. And again, if I see no performance, I’m going to pause it.

So my idea here, or the process, consists in narrowing down the semantic field little by little, just to see if by doing so you can have a specific keyword coming from being non-profitable to profitable. Basically, that’s the point.

Frederick Vallaeys: Yeah. Okay. So that makes a lot of sense, and I can totally see how this can be automated. It’s a very clearly defined rule. But what happens to all the search terms that would have come off of the broad match that now potentially stop showing up as you move it into an exact match? How do you still capture those?

Matthieu Tran-Van: That’s a very good question. I did not design the slide specifically for this, but that’s my secret sauce here. Because, and you have found it, is that this rule is always paired with another rule where all my search terms that are highly profitable with a certain volume of conversion and a return on ad spend that is way above the campaign average, I’m going to add them also in exact match. Okay? So I have two rules—one which is doing this keyword pruning thing that I’m describing right now on the screen, and I have another rule that is going to work at the search term level where I can have one broad match keyword that triggers, I don’t know, 10 different search terms. Out of those 10 different search terms, I see two search terms that are highly profitable—boom, I’m going to add them in exact match also.

Frederick Vallaeys: That makes total sense. And that’s the whole concept of layering, right? It’s not just one thing. It’s multiple things that you stack on top of each other that work together to give you the best results. So brilliant.

Matthieu Tran-Van: Exactly. Exactly.

Frederick Vallaeys: All right. That makes total sense.

Matthieu Tran-Van: Yeah. And we cannot hide anything to you, Fred. Actually, you already anticipated the signal that I had.

Frederick Vallaeys: I promise I didn’t poke around your Optmyzr account to figure this out.

Matthieu Tran-Van: Okay. Okay. No problem. And so the thing is that if we come back to this purgatory rule that I was sharing, the good news is that you can translate this super easily into a tool like Optmyzr, of course, because the goal for you as a PPC practitioner is not to babysit every keyword manually, of course. And so you need to build a system that is going to automatically execute this heuristic. And by doing so, it’s going also to automatically then reward the winners and add them in exact match and eliminate the losers without any kind of human intervention.

And the real added value of Optmyzr here versus the scripts and so on is that, first of all, it’s simpler to implement, but also you have a very high and granular control on the different thresholds that we want to apply on the behavior of the rule itself. And at the same time, you maximize your productivity and efficiency because, again, this goes without any human intervention, right? So that’s just magic.

Frederick Vallaeys: Total sense. Let me ask you a question here. Maybe I’m not going to like the answer, but I’m going to ask it anyway. So using the rule engine from Optmyzr, you’re sticking these keywords—you’re basically keeping them in the same structural place where you had them before, but you’re modifying the match type. Do you ever think about moving it to a different campaign with a different budget once it graduates to exact?

Matthieu Tran-Van: Yes. So basically, the first step is that I’m going to add those keywords into the same ad group where they have been triggered. Okay. So campaign one, ad group A. I have one broad match keyword in that ad group A. It’s profitable. I’m going to add the same keyword in exact match in ad group A. But once a month, what I’m going to do—and here I’m going to restructure or reshape a little bit my structure—so basically, I’m going to look at all the ad groups where I have more or less more than 20 keywords. And I’m going to build, again, the alpha-beta strategy, right? A top keyword campaign where I’m going to feed all those keywords in exact match into my top keyword campaign.

And this top keyword campaign is going to have a dedicated budget, which of course is never constrained. Okay? Super open budget on this. And very often a specific portfolio bid strategy where I’m going to most of the time keep the same constraint of target ROAS or target CPA, but I’m going to increase, for instance, the CPC cap. Okay? I’m going to really open the delivery on those keywords. But this process, first, I’m going to let the automation handle all the heavy work. And once a month, I’m going to reorder my structure entirely.

Frederick Vallaeys: Yeah. Great. I love it. So that’s your targeting for you, right? Improved targeting with automation layering.

Matthieu Tran-Van: To show some real numbers, because that’s what is interesting also, is that here, for instance, it’s a real e-commerce case where I have applied this exact strategy with Optmyzr’s rule engine. And as you can see on the screen, over 17 months of non-brand search campaigns, this specific automation steadily improved the return on ad spend while, of course, keeping the system pretty scalable, right?

So the takeaway that I want to make here is that automation is not only about losing control. It’s not because I let this automation run my keyword portfolio that I’m totally losing control. It’s about gaining this consistent performance without the constant micromanagement, I would say, right? So that’s the interesting thing. And again, that works, as we can see on this case study.

Frederick Vallaeys: Yeah, that makes sense. So you, as the human, you spend your time focusing on the strategy, testing out new things. And then once you’ve found what works, use an automation system to do that consistently for you. That’s honestly—but let me ask you, because one of the biggest things for me is like, I love exploring and experimenting with new stuff, but then once you’ve found the best way, then it gets boring, right? You want to move on to the next thing. So I think that’s where automation is so critical, because otherwise, like, you have these great insights and then you just never do anything with them.

Matthieu Tran-Van: Yeah. Exactly. It gets boring, but at the same time, it needs to be executed. So, again, that’s where automation is so important.

Frederick Vallaeys: Yeah. All right. So let’s talk about messaging.

Matthieu Tran-Van: Yeah. So next up is messaging. And here, this is where we turn DSA—so dynamic search ad campaigns—from actually boring to high-performing, completely on autopilot, right? So that’s what I wanted to share with you today. So if you remember our first episode, Fred, you already know how much I love DSAs, right? So I think that when they are built properly, they can not only generate highly relevant ads but also deliver tremendous performance.

So just as a reminder, the tactic that I shared the first time was to build highly granular campaigns directly from the Merchant Center feed, and again, using Optmyzr doing that. And in the examples that I shared, I had up to 52% increase in sales that were attributed to search and just with this campaign, right? And even your team—I was pretty honored when I saw that your team even reshared this tactic with your community. It was super interesting for me.

But however, since I shared this tactic, some advertisers came back to me and told me that they are still selling stuff online. Think about services or travel advertisers, etc. But most of them have not a Merchant Center feed per se, right? And here the good news is that even if you don’t have a product feed, you can still create one similar type of campaign from any business data. And that’s what I wanted to share again today here.

So all you need in the end, when you think about it, is a spreadsheet. It’s a spreadsheet that refreshes daily. And so you can have an advertiser updating this data on a daily basis, and you can feed that into the Campaign Automator of Optmyzr. And then you can instantly generate campaigns. For instance, the strategy that I used to do is generating campaigns by business priority. So I have different DSAs—P1, P2, P3, etc.—that is also going to create ad groups that are aligned with each offer page. Okay? With—that’s the beauty of it—with tailored ad variations that you can build at scale. And then the campaigns manage themselves automatically. No manual build, no endless edits, things like this. You let Campaign Automator do the work, and it works like a charm, right?

Especially when you look at the ads, right? Because this feature gives you—can give you thousands of very tailored ads without the headache, I would say. And it’s actually simpler than using ad customizers that you have in the Google Ads UI. It’s even more simple. So you can automatically include elements like what we see on the screen—like geo-targeting, price, ratings, features of the products, etc. And instead of spending hours writing ad variations or preparing those in your spreadsheet to implement them with Google Ads Editor, here you just let the system generate them dynamically and accurately for you. And that’s, again, tremendous.

Frederick Vallaeys: Beautiful. I love hearing that what we built is easier than Google Ads itself and more powerful. So yeah, this is amazing. Now I got a question here too. So these examples that you show on the screen, they’re in French, right? So basically, on the left side, you’re renting a camper. And the ad on the right is renting or is it—I could see if it’s renting or buying. It looks like it’s buying a Ferrari.

Matthieu Tran-Van: Yes, exactly. Exactly. That’s it. If you want to buy a Ferrari, Fred, then I have an ad for this. Okay?

Frederick Vallaeys: Well, you know, I’ve actually been pretty hot on Waymo lately. I don’t think I need any more cars to buy. I’ll just be driving around in a robo-taxi. But thank you. But my question is, when it comes to these different personas—almost like someone who’s going to buy a Ferrari is probably maybe not the type of person who’s going to go camping in an RV. They might be more of the five-star resort audience. So how do you think about the actual messaging beyond just making it relevant with the business data, but also kind of making it resonate with the customer?

Matthieu Tran-Van: Yeah, that’s a very good point. So here in those three examples, as you can guess, they come from three different accounts and three different advertisers. Okay? So of course, they are not marketing their product to the same persona because those are three different businesses. But to your point, I think it’s really relevant. And to be honest, I’ve never tested it, but that gives me another idea where you could actually combine, you know, audience targeting, right, with some specific DSA campaigns that target specifically some kind of offers.

So all your premium offers could only be served to a specific audience, etc., etc. And that’s pretty interesting because when I think about it, the way I would maybe do it is you take, you know, high-value, mid-value, low-value buyers. You import those as customer match lists. Then you look at the audience insights. It gives you maybe three, four, five very interesting audience segments that you can target for each range, let’s say, of services. And then you adapt the different offer page or product to each of those sets of audiences. That makes a lot of sense, and I’ve never tested it.

Frederick Vallaeys: Okay, maybe the next Automation Layering Masterclass episode we’ll get some results on that—another layer to bring into the whole system.

Matthieu Tran-Van: I would love to. I would love to. But still, you know, without even combining this audience thing, which, again, is super interesting and needs to be tested, here is the proof, right? In one e-commerce account where I spent about 150K in those DSA campaigns versus 65K in non-brand traditional, let’s say, search campaigns, the results were super interesting because those types of DSAs built with the Campaign Automator were very often—and they actually were—more efficient than the non-branded campaigns.

As you can see, the conversion rate is really higher. The conversions per impression is way better, as the return on ad spend is also way better. So for me here, the lesson is pretty clear. Automation is not only about being lazy, right? It’s not just because I did not want to write thousands of ad variations. Actually, you can see that it can be super strategic when it’s done right. It can be a real growth engine for the account. And here the strategy was to really align an offer with a certain semantic field chosen by DSA that delivers a very specific ad that is super tight with the keyword and the page that the user is going to be redirected to.

And so, of course, the hypothesis is true. You have a better conversion rate. It’s more efficient. The return on ad spend is better, and so on and so on. So that makes sense. You will say yes, that’s common sense. Yes. But here we see the numbers, and it actually works, right?

Frederick Vallaeys: Yeah, I like it. And I know we’re going to talk about profits later on. But what I like too about this particular slide is that you’ve included the metric of conversions per impression, which is not actually a Google metric, but you had to calculate that, right? And it’s really important because you might say, “Hey, look, my conversion rate is more than double and my ROAS is more than double.” And it’s like, that’s great, but maybe you sold two instead of 100. So obviously that’s not good for business, right?

But if you bring it down to conversions per impression, that is really a metric that I think will eventually connect to profitability because yes, in exchange for boosting your conversion rate, you may be buying fewer clicks, get fewer impressions along the way. But what ultimately matters is how many conversions you still drive. So really like that you’re bringing that metric to the table as well here.

Matthieu Tran-Van: Yeah. Yeah. For me, it’s a very important metric because it’s the one metric that truly encapsulates the entire value chain, right, from impression to conversion. And again, it gives us the true effectiveness of the campaign, right? So totally aligned with you on this one.

Frederick Vallaeys: All right. Bidding.

Matthieu Tran-Van: Bidding. So now let’s talk about bidding. Specifically here, I wanted to discuss with you how to move from, let’s say, smart bidding, right, to true profit bidding. And of course, doing it without hacking Google, right? So that’s the point. And again, in our last episode, I shared some specific tactics that improved the return on ad spend by optimizing consistently by 1% portfolio bidding strategies, if you remember. So every time that I hit my target, I’m going to challenge the algorithm to get just 1% higher ROAS. And this tactic actually works like a charm.

But the problem that we can highlight or the objection is that return on ad spend does not always equal profit, right? Because if you incorporate this notion of profit generation in your automation, then it can help you not only focus on return on ad spend or profit on ad spend, but also on margin, which is margin in dollars, right? Which is not a vanity metric. And that’s what we are actually looking for. So I wanted to come back on this bidding strategy and share with you a kind of improvement that I’ve been making on this tactic.

So just before showing you what I did, again, for us to really be aligned and clear on that, because what really matters when you think about it is not just the return on ad spend—it’s the relationship between the revenue per click and the incremental cost of a click, right? How much do you need to pay to get one additional click? So here is the visual on the screen, right? When your revenue per click—the RPC—is higher than the incremental cost of that click, of course you are earning money. You are making profit. But when the revenue per click is lower, you are losing money.

So the sweet spot is exactly where your revenue per click equals the incremental cost of a click. That’s where your profit peaks, basically. So when we understand this, for us PPC practitioners, we understand that the goal is not to maximize return on ad spend at all costs. It’s to maximize profits. And so that’s what I wanted to improve in my previous strategy.

Frederick Vallaeys: So did you want to react, Fred? Oh, I was going to react. Yes. So I was going to say profit is good, right? But there are some businesses that don’t necessarily care about profit. They might be in growth mode. So they might want to optimize revenue. But ultimately, I think your point is very valid. It’s just like, understand what is your actual business goal and then understand how that mathematically corresponds to how Google Ads works. And so what you’re showing here makes sense for probably the vast majority of advertisers. Let’s maximize that profit. So love it. Let’s see how you do it.

Matthieu Tran-Van: Okay, cool. So how do you operationalize this? Well, here with, I would say, a simple framework, right? So here it’s a matrix that you can see on the screen. And here you have four quadrants. So if your—in my case here, I’m working with POAS, so profit on ad spend—so if your profit on ad spend is below target and the ratio between the revenue per click and the CPC is trending down, then you are going to lower your bid significantly because you earn less and less. Not only do you not achieve your target, but also you are earning less profit. So you want to really decrease your bid significantly.

However, even if your profit on ad spend is below target, but the ratio between the RPC and the CPC is up, it means that you’re actually improving your profit generation. So you can just slightly, let’s say, decrease your bid with a smaller order of magnitude. Okay? And then you can replicate a little bit this logic when your actual profit on ad spend is higher than your target. And in that case, of course, you will increase your bid with different orders of magnitude depending on if your ratio between the revenue per click and the CPC is up or down, right? So it gives some nuances on how—by how much you should basically test or change your bids, right? That’s the new thing. And it’s not only about getting your ROAS 1% higher—it’s also having this smart intelligence on by how much I would also adapt the other constraints like CPC and so on.

Frederick Vallaeys: Okay. So basically what you’re saying is this is a more nuanced strategy than the “let’s improve by 1% every time.” So I’m going to guess it gets you to your results faster because you’re making a more strategic decision at every point.

Matthieu Tran-Van: Yes, exactly. So here, when we look at the results—again, that’s an e-commerce case where I’ve applied this exact rule just before. Maybe you can build this also in Optmyzr. So you see my matrix and those are the exact rules that I’ve been building and that correspond to each of the quadrants. So just by adding the specific threshold that I wanted and building some expressions in my rule strategy to actually measure and compare between two periods the ratio of RPC and CPC and so on, I could just replicate the exact framework that I was presenting earlier into—and operationalize it into Optmyzr. And the results, yeah…

Frederick Vallaeys: And I want to pause on that for a second. So, you know, people might have to make this really big on the screen to see what these rules in Optmyzr actually do. But one question I have is a couple of slides back, you talked about the incremental cost of a click, I believe.

Matthieu Tran-Van: Yes.

Frederick Vallaeys: Right? Talk about the incremental cost per click is not a Google metric.

Matthieu Tran-Van: Yes.

Frederick Vallaeys: And it requires the simulation tool in Google to actually derive that. So how do you go about that in a practical sense where you don’t end up having to manually calculate this constantly?

Matthieu Tran-Van: That’s a very good question, very well highlighted, Fred. So because the ICC is not a Google Ads metric, okay, and it’s not available also through the API, etc., I had to use the proxy of measuring the evolution of this ratio between the revenue per click and the CPC. Because think about it, right? Your ICC needs you to compare what has been the cost between two periods. And what has been the delta in cost and what has been the delta in clicks, okay? And you cannot operationalize this into Optmyzr with a rule strategy.

So you need to find another proxy, and the proxy is the following one. So if you look at a period—let’s say last seven days—what has been my ratio between my revenue per click, so conversion value divided by clicks, divided by my CPC? Let’s say it’s, I don’t know, two, right? And then you look at this exact ratio the seven days previous to this period, okay? So going back to 14 days earlier, and you remove—you offset seven days. So you compare last week versus the previous week, right?

So if this ratio has increased, well, it means that the delta that you have between the revenue per click and the CPC has actually increased. So you are making more profits. Back to the visual that I was showing, right? So that’s why I use this ratio. It’s because the ICC is not an available metric in Google Ads, and it’s not available in the API, unless I’m mistaken.

Frederick Vallaeys: No, you’re absolutely right. And that’s why I was curious how you do it. But so I think you pointed out two brilliant things. One, thanks to Optmyzr, it’s very easy to do these separate range comparisons in the rule engine. But then second, the fact that you’re leveraging ratios—and I wrote a post on that many years ago, like, why are ratios so helpful? It’s basically because it’s your gauge, right? It doesn’t tell you the actual thing, but it tells you if you’re getting better or worse.

And so I love how you’ve deployed the ratio metric that you’re calculating on the fly to help you steer the ship and then leveraging that and, again, layering it together with actual targets to get you to where you need to be. So yeah, no, love how you’re putting all of this together and sharing it with us.

Matthieu Tran-Van: Yeah. And ratio is really—it has a tremendous superpower. And as you say, it’s a kind of way to see what’s the direction you are going to take without actually giving you the exact number. But it’s a hint. It’s a very powerful directional hint that can help you make a better decision. And you see here the results also were pretty good because—so again, here is an e-commerce case where I’ve been applying this very precise Optmyzr rule. That’s the account itself. And just by dealing with smart bidding with these profit-based rules, you see that the account has doubled their profits without sacrificing profit on ad spend, actually. And that’s the beauty of it.

That’s the magic takeaway. You don’t need to outsmart Google in the end. You just need to guide it with consistency and the right indicator. And that’s where you see your profits and performance increase. So yeah, I was really happy with the results of this new strategy, actually.

Frederick Vallaeys: Yeah, I can imagine. Really good results.

Matthieu Tran-Van: Well, my client was even happier, to be honest.

Frederick Vallaeys: Hey, time for you to start your own e-commerce company and maybe start selling a product.

Matthieu Tran-Van: Maybe. Maybe. No, but I’m having too much fun, Fred, you know, managing campaigns and so on for so many businesses that I think I’m addicted to this.

Frederick Vallaeys: Well, good. And also, I mean, I think you did point out, you know, some of these metrics are not available in Google Ads. And so ultimately, that’s, again, where a smart human helping you with campaigns is really helpful. So if anyone watching is like, “Hey, we could use some help,” Matthieu is sharing a lot of stuff, but there’s also a lot of things that he can still bring to your account. And let’s add some of that to that number of spend that you manage.

Matthieu Tran-Van: Thank you. Thank you very much, Fred. And so, yeah, and finally, maybe I wanted also to look with you at maybe the next—what I call the next-level automation, right? And here it’s really automation taken to the maximum in this case. And my goal was to turn a lot of these different business data into a fully automated ad strategy end to end, okay?

And I decided to do this because I had the case of a retailer which was pretty difficult at the beginning because this case study—it comes from a retailer that looked like every other competitor. So basically, they sold the same SKUs as the competitors. They had seasonal demand because they are marketing swimming pool equipment and gear. So as you can imagine, it’s highly seasonal. Their conversion rate was heavily influenced by weather—again, because they are selling swimming pool equipment and gear, right?

And on top of that, the margin varied widely across more than 6,000 products. So no clear differentiation. A lot of external context can influence your sales, and you have a large catalog with different levels of margin. So the question here was, okay, how do we make Google Ads reflect all that complexity automatically and make that work for this advertiser? That was the initial challenge. Okay, sounds like an easy one. Talk us through.

Matthieu Tran-Van: Exactly. So my answer actually was not that easy, to be honest, because—

Frederick Vallaeys: Yeah. Yeah.

Matthieu Tran-Van: But it was not that easy because when I thought about it, I broke it down into four different steps, right? Starting from the insights, then looking at which business data do I have, how can I operationalize this into Merchant Center, and then finally into Google Ads. And so what I did is that we worked with the client where we started by gathering data, especially about competitive pricing. So we scraped competitors’ prices, etc. We collected the margin by SKUs. We also had a look at what were the inventory priorities, etc. And we implemented and translated all this data into, you know, smart labeling, let’s say, of custom labels in Merchant Center, okay?

Those custom labels helped me build a structure that was fully aligned to the business. So depending on the competitiveness, the priority scoring—because you have a lot in stock or not so much—and also the margin that the retailer had, every intersection of those three dimensions was giving one specific Shopping campaign. Okay? And of course, you can automate then all these custom label applications, right? And I then aligned portfolio bid strategies that were actually aligned on the actual margin that the retailer was earning.

And so as the output, you have an account where you have a dynamic campaign structure that is built around what I call this perfect trifecta between the right assortment with the right budget and the right bid strategy, right? So that gives you in the end—and because the business is highly seasonal and is highly influenced by the weather, I’ve then layered some geolocated Performance Max and Demand Gen campaigns that only promote the bestsellers. Okay?

And how do you find the bestsellers? Of course, thanks to the smart labeling tool in Optmyzr, where you can put some rules—you know, give me a label on all the SKUs that have given more than, you know, 10 conversions during the past quarter, and so on and so on. So you can add the rules that define for you your bestsellers. And so those Performance Max and Demand Gen campaigns were promoting those bestsellers just when weather conditions triggered the demand, right? So it’s sunny and it’s warm in this specific city—boom, you have your Performance Max and Demand Gen campaign that is activated. The day after, it’s raining, but it’s sunny in another city. We change automatically the geo settings thanks to Optmyzr to, you know, promote our bestsellers in this new area.

So this activation layer was just to reinforce even more the advertising strategy where the demand was highly likely to convert, right? So that was the idea. And all this system, all the campaigns, they can react without any manual intervention. That’s the beauty of it. And it also aligns the system—every decision, bids, products, geo, etc.—with the business data that you have in the input. So in this case, the humans just define the strategy. Exactly what you said earlier, Fred. Humans just define the strategy, and automation executes in real time. That’s what it does, basically.

Frederick Vallaeys: It’s brilliant. Yeah, so obviously not that simple. A lot of layers to this and a lot of thinking. But again, another brilliant illustration of how putting in all of these different capabilities and layering it all—I’m going to assume the results were good, right? So let’s take a look at that.

Matthieu Tran-Van: Yeah, the results were massive, to be honest. So just in comparison, we are comparing with the previous year. And in the previous year, the client had only one Performance Max campaign. So we are comparing—and during the same period, year on year—this system versus one Performance Max campaign. And the results were massive. So the revenue uplift—we had revenue uplifts of more than 100% in some months, of, you know, almost 300%, 159%, and so on. And all of this without extra manual work, right? Again, I want to highlight this. And the campaigns adapted automatically to margin, inventory, competitiveness, and seasonality. And yeah, return on ad spend were, you know, very good—at least the same as the previous year. But the volume of conversions that we got had nothing to see. So again, the lesson here is simple. Automation does not replace the strategy itself, but it can truly amplify it by executing it with perfection on a daily basis.

Frederick Vallaeys: Beautiful. No, I mean, that is the big takeaway here. Like you said, strategy, the machines automate it, multiple layers, and money comes out at the end. Money, money, money.

Matthieu Tran-Van: Exactly. Exactly. That’s it. Yeah. So if you want, just to wrap up with the key takeaways about those four plays, I think my three main messages—first is that, of course, automation frees you from babysitting anything in your account, so you can truly focus on the real strategy. And I think that in an AI-driven age, right, strategy—our strategy and creativity is our competitive advantage as humans, so we should focus our energy, efforts, and brain juice, you know, on this. So that’s the first thing.

Second, what I want to highlight is that the most impactful automation strategies are actually tied directly to some kind of business output—profits, margin, or demand signals. It’s not about clicks and impressions. If you can tie your automation to something that is truly meaningful for the business, you have a high likelihood to actually win, right?

And third, of course, the third message maybe is that tools like Optmyzr are not, of course, magic wands, to be honest. But they are orchestration layers, and they take your strategy and make it scalable, consistent, and, hopefully, profitable, right? But that’s what you need to remember about those tools. And that’s the superpower of those tools. Voilà. That’s it.

And so, of course, if someone in the audience is applying one of those four plays, I think I can bet that they’re not only going to scale their account with confidence, but also they hopefully scale their profits all along the way. And I’m also super eager to see if anyone, you know, tries one of those plays—if they want to share their learnings and results, I’m more than happy to have some feedback also.

Frederick Vallaeys: Great. Yeah. So anyone who’s watched this video, we got the comment section below. We would love to see in the comments which of these techniques you’ve tried, what kind of results this has led to. And maybe if your results were amazing as well, you should be a guest on Automation Layering Masterclass. So put that in the comments and let us know.

Now, in all of this, Optmyzr is obviously a very helpful tool, but you still need a smart human team to put in place the strategy. So, again, that’s where Matthieu—he is available. Matthieu, if people want to get a hold of you or maybe get an audit of their campaign or talk more to you, where can they find you?

Matthieu Tran-Van: Yeah, you can easily find me on LinkedIn. That’s the most convenient way of finding me, and you can reach out on the platform. I will be happy to read and answer any message that I receive there.

Frederick Vallaeys: Great. Yeah. And again, thank you so much for sharing all of these wonderful insights and sharing how you do it. Everyone, thank you for watching. If you’ve enjoyed this episode, which I’m sure you have, please like it. Please subscribe so that you can get updates when Matthieu and other guests come back and share more secrets about how to do automation layering. And with that, we’ll wrap it up here. Matthieu, thanks for being a guest. Thanks, everyone, for watching, and we’ll see you for the next episode.

Matthieu Tran-Van: Thank you very much, Fred. Thank you. Bye-bye.