Get some practical advice on how to structure your advertising to achieve increased profits as you’re growing your business. In this session, we explore issues around the accuracy of conversion tracking and how that impacts financial decisions, how you should set ROAS targets and budgets to maximize profitability, and how to handle situations where you have products with varying margins. This session will help you make better financial decisions about your firm’s digital advertising.
Presented by David “Psy” Deppner
Meet Magento New York, October 11, 2023
TRANSCRIPT
MMNY Host:
Our next speaker is David “Psy” Deppner, and he has been around for a minute or two. Worked on both, merchant side and the agency side. Real pro in terms of, e-commerce advertising. Advertisement. Today we’re going to be talking about, profitable growth. Right. So, I think there’s a lot of reservations out there in these conversations that I have with with merchants, with my customers.
About ROI. Right. If you’re going to make an investment, if you’re gonna make a plunge, you know, having some real confident forecasting. And so David’s gonna, you know, give it, provide his expertise on the matter. We’ll have, a few minutes for questions. After about 20 minute talk. So if you guys have any questions, we’ll get to those at the end.
So, without further ado, David.
David Deppner:
Okay. Thank you. Good afternoon, everybody. Okay. Let’s go ahead and get started. So the main topic today is going to be talking about advertising systems. But in particular we’re going to get into some questions about profitability and how to squeeze a little bit more profit out of the ad accounts that you may already have, or how to think about how to ramp them up to get them to be as profitable as you can get them.
Some structural things that might help you with that. I’ve already been introduced, but just real quick background. I was, director of IT at a packaging company back in 2008 as we launched one of the first Magento sites. And, I got experience there, with not just e-commerce, but also with running the ad. The ad accounts for that company.
Over time, we merged those all together into a larger, a larger department. And at a certain point in time, I realized that that was way more fun than working for a single company. Like managing advertising for multiple ecommerce merchants. And we ramped up Psyberware as a business to do that. So Psyberware basically focuses very specifically on e-commerce advertising.
We focus on product based companies, people moving products. It can be B2B, it can be lead based business, with salespeople closing deals after the fact, but basically some kind of e-commerce. We do a lot of B2C, a lot of B2B. The agenda today, we’re basically going to first go over some some basics about return on ad spend and how to think about some of the math behind this.
Not too deep on the math, but just some basic concepts about what the trade offs involved are. And we’re going to get into some, some details about, maybe how to structure campaigns in certain scenarios, in particular when you’re dealing with things like, a complex product catalog with lots of different margins. I’m going to talk about budgets over time and how to align some budgets with the opportunity available to you, and then get into some things about conversion tracking and improving accuracy there.
And we’ll wrap it up after that. So if there’s any, any one point I could leave you all with about the idea of what your return on ad spend should be. I want to talk about that here. I want to I want to leave you with the principle that there is an ideal return on ad spend for your business.
A lot of people don’t know what it is, and I want to talk around some of the ideas about maybe how to think about that. And so ROAS is one of these fancy, acronyms. It just return on ad spend. It’s just your, your conversion value divided by your ad spend. So if you spend a dollar and you get $5 back in revenue, we’re going to call that a five times return on ad spend.
Or we might say it’s 500%. I think it’s important to note that ROAS is not the same as a financial ROI calculation. Because we’re actually this is, you know, we’re seeing five times revenue, but we’re, you know, one, one, one of those times is actually just getting our initial investment back. So if we were talking about this in terms of ROI from a financial perspective, you know, that would be a 400% ROI, right?
A 400% return. This is, you know, the cynics among us might think that this is a way that the ad platforms like to inflate their numbers a little bit and make things look better than they really are. And the cynics among us would be right. But this is the tool we have when dealing with the ad platforms. They pretty much all standardized on ROAS reporting and target ROAS methods of controlling campaigns.
So this is the tool we have. And that we need to work with.
There’s a lot going on in this slide. I want to pause here for a minute and just make sure that the trade offs here are clear.
If you increase your spending on an ad account because you’re dealing with an auction, every additional click you buy costs more than the click you already bought. So we have a system where there’s increasing marginal costs as we scale. Because of that, we can be very profitable on each sale on a smaller ad account. But as we scale it up, as ad spend increases, return on ad spend decreases.
They have an inverse correlation. ROAS is not at all correlated with profit. What we actually see is as we move up from a, say, a smaller ad account that’s, you know, at point A on this graph right, they’re spending a lower level, they have a high ROAS. But if they increase their spend a little bit more to point B, their ROAS is dropping because their marginal costs are increasing to get that additional scale, but their total profitability is still going up as long as that marginal cost isn’t eating all of the marginal sales of all of the marginal product margin, right?
If you increase the spending a little bit more to point C on this graph, we get to a point where there’s a there’s an inflection point here where the trade off between increasing volume is basically getting a, counteracted by the increase in marginal cost of getting additional sales. So this is just like a theoretical point, right? The real world is way messier than this.
But there is a theoretical point where the ROAS is a perfect balance between those rising costs and the amount of profit that you can get on an additional sale. Right.
So if you think about that point C, if we spent a little bit more, we’re losing money. Our profit goes down. But also if we spend a little bit less, we’re losing money. Our profit goes down. It’s the opportunity cost of a lost sale we could have made that would have contributed more profit to the bottom line. That point is actually not easy to perfectly calculate.
But there’s a rule of thumb, and the rule of thumb is that it’s going to be somewhere around two divided by your gross profit margins. So just as an example, if you basically had two thirds cost on a product sale and you, you know, you got a 33% gross profit margins, approximately two divided by 0.33 is roughly six.
So your optimal return on AD spend with 33% profit margins would be roughly in that neighborhood. Might be a little higher, might be a little bit lower. People will have some different biases as well. They might want to factor into the the decisions they make for their particular business. But that’s roughly the point where you’re achieving that optimal profitability.
But there’s all sorts of exceptions to that, right? If you’ve got a situation where you’ve got higher costs than what you think they are, all right. If you got additional costs of sale, like you’re eating free shipping, right? You’re dealing with payment processing costs, things like that, then really your costs are a little bit higher. And you might want to factor those in, you know, don’t think about it as just being gross profit margin on the product portion of the sale.
But all of the variable costs associated with that sale. It’s easy to remember to divide it by your profit margins, but it’s really two divided by, you know what the contribution margin is after accounting for all of your variable costs of the sale. We also have situations where there’s multiple platforms claiming credit for a sale. You know you’re advertising on Meta, you’re advertising on Google.
You look at your analytics data and you see people are clicking on ads on both platforms before they actually make a purchase. But both of those platforms are claiming credit for that in the conversion value. So all of those sorts of scenarios where your costs are higher or you’re getting things with, your, your revenue actually being lower on each individual platform.
Those types of scenarios will maybe affect where you want to set your ideal target ROAS. You also might want to factor in the cost of management. And even if you’re not outsourcing it to a company that’s doing the management for you, you have internal costs for the employees, time that’s taking care of all this. And those are also advertising costs.
But on the flip side, there’s all sorts of scenarios where you might want to consider having a lower return on ad spend target, in particular businesses with, almost like a subscription aspect to them or literally a subscription aspect to them. If people buy something once and then continue buying supplies for that machine, or if people are, say you’re in a supplement company type situation, somebody who buys the first time has a very high likelihood of buying lots of times after that.
So you might want to instead do your math based on some some projections about lifetime value rather than the conversion value of the initial sale. When you know for sure that you’re actually getting new customers with a much higher lifetime value. Conversion tracking under reporting is also an increasing problem, like the amount of data, the amount of conversion value being reported by these systems is frequently just false.
It’s it’s maybe a little, you know, your real your real revenue might be higher because they’re losing some sales because of ad blockers. Right. We also have scenarios where, if you’re just not tracking certain sales, if your business is selling some things that have to go through a salesperson to close the sale, or if you are using a financing system where people get redirected away from your normal checkout to go through a financing process, and then that’s approved a week later and one of your salespeople closes the sale, you’re leaking conversion data out.
So those are scenarios where you might actually be want to be a little bit more aggressive with your ROAS targeting than what your conversion value would suggest otherwise. Last, I’ll just say ramping up faster. There are definitely situations with smaller ad accounts where the most profitable thing you can do is overspend for a few months early on to get data faster, so you can train the machine learning algorithms that are controlling your campaigns to achieve more profit in 2 or 3 months.
So there are all kinds of exceptions to, to to the rule of thumb, of setting your target ROAS around two divided by your gross profit margin.
Let’s talk just a little bit about how some of this might affect campaign structures. In the modern, advertising world, we have this increased use of machine learning to control campaigns and each campaign is like a collection point for a lot of data. And the more data these campaigns have to work on, generally, the better they perform. And in fact, depending on what bidding strategies you’re using, all the ad platforms have recommendations on how big campaigns should be to even be a separate campaign.
There are all sorts of reasons that people split up different campaigns in their ad accounts, and most of them are bad reasons. There are good reasons to do so. If you need to control things and have different ROAS targets on different campaigns, then you want different campaigns. If you need to target things differently by geography, you’re right. You might need different campaigns.
But most of the time, merchants lean towards wanting to split up campaigns based on their reporting requirements or to structure things around their product categories. That can be fine if you’re a really large company with a lot of conversion data in each of those campaigns, but we frequently see scenarios where people have a lot of very small campaigns or a few very large campaigns, and then the long tail of their their products that aren’t as popular, they just don’t get enough data to ever optimize.
Well. So combining some of these campaigns together is frequently a really good way to improve the performance of your account and improve the profitability. In particular. In particular. It’s it’s a very good idea to group campaigns together by the margins of the products that are being advertised in those campaigns. So this is this is a real example, from an account where they’ve got three different performance max campaigns that have the shopping, you know, portions of their shopping feed mapped to each of them.
And these campaigns are targeting products based on their profit margins. So you can kind of see we’ve got a campaign with a target ROAS of 500% or five times. That’s because it’s got products in there that tend to be clustered around a product margin of around 40%. Right. And if you think about it, the math two divided by 0.4, there’s five .4’s in two.
That’s why we come up with that number. That’s approximately correct for this. For this account, by tiering the products with different return on ad spend targets all together in campaigns, even irrespective of what class of product they are. Right? Like what type of product they are. This, just aligns the bidding strategy more with the opportunity of what the profit contribution could be on a particular group of products.
So talking about budgets in general and in particular budgets over time, this is one of the areas that has a a huge impact on profitability as well. If you can keep your budgets flexible so you’re not spending a fixed amount every month planned in advance, and instead you can use a target ROAS bidding strategy, for the version of that that’s available on the various platforms you’re working on, it allows the spend to fluctuate up and down dynamically with changes in opportunity day to day, month to month.
This by far is the way you will squeeze more profit out for the amount of money you’re spending. The. The reality for a lot of businesses is they need to plan in advance and set budgets frequently a year out. We’re working with a client right now. They’re setting their 2024 budgets right now in October, and they need to plan the entire year into next December.
This is pretty common. This will be approved by November, and it’s set in stone and probably won’t change. So that’s a challenge because we have to think about how we can project out what the opportunity might be in the future. Here’s a here’s a simple example of of maybe how we might go about thinking about that. The first graph here is showing you spend over time.
This is historical spend over time over a 12 month period. You can see this okay. This this particular account spends about $27,000 a month, most months. A little bit of fluctuation here and there, but it’s around that. It’s pretty fixed budget because that’s what is in their corporate budget. And that’s what they’re going to spend. And they’ve allocated a little bit added spend to ramp up in late November and through into December to capture more holiday sales.
When there’s more opportunity. But the second graph here shows what their ROAS is over time. And when you have budgets that are flat, you’re going to have ROAS that spikes up and down because there’s different levels of opportunity month to month. That opportunity doesn’t just come from people’s search volume or search preferences changing. It’s also coming from the fact that your competitors are responding differently over time.
So one thing that I note is that this company wants to spend more in November and December to catch more holiday sales, but all their competitors do as well. You can see there’s much higher opportunity to capture cheaper sales in the springtime for this business when the competition isn’t as fierce, but there’s a lot of demand for their products.
So the question is, if we have this fixed budget and we want to allocate it to be more profitable throughout the year, what do we do? It turns out that we can basically use the data that we have about what the response is month to month, for a given level of spend in order to sort of recast or model what would have happened if we had spent at slightly different levels.
Right. We know what the you know, we we know the response curve here. We know basically the slope of that line. And we can project out a little above, a little below, and think about what the what the what the what the trade offs are. Right. And by taking the same budget, same annual budget, if we can get consent from their financial department to shift that.
So it’s a little bit more some months, a little bit lower, other months we can kind of normalize that out. So they’re getting more of a fixed ROAS over time. They’re taking, you know they’re not overspending some months and underspending others they’re allocating it more. So they’re getting a little bit better performance year round. Right. They’re aligning their spending with the opportunity.
And in this particular case, just a little shift like this with no changes to the budget at all. It in this particular case resulted in about a 5% lift in total revenue at exactly the same budget. And that’s just about aligning the spending with the opportunity throughout the year. There’s still a little kick up in spending in December, November, December, but we’re putting a whole lot more of that into the spring because there’s a whole lot more opportunity there.
So conversion tracking. Conversion tracking is changing very rapidly right now. So I just want to wrap up with a few thoughts here about what you might want to think about. Historically, we would do conversion tracking straight out of the e-commerce system. We’d have some tags on the checkout success page that fire off conversion data in JavaScript to the various ad platforms, most more sophisticated sites.
Now we’re going to do that in Google Tag Manager or some other tagging implementation. But a lot of e-commerce sites are really lagging behind several years behind what’s going on right now. One of the things is we have increased cookie blocking, which is affecting the ability to or, you know, tracking blocking of the the click identifiers, which is affecting the ability to attribute a sale back to the ad click that resulted in it.
So one of the one of the solutions to that on the Google platform is to use what’s called enhanced conversion tracking. And most people have heard of it, but probably 50 or 80% of merchants haven’t actually implemented it. This system works by passing a unique identifier based on some information that Google knows, and we know we we create a hash of the user’s email address or some other information about them, and we can pass that along with some conversion data in Google Tag Manager over to Google.
And they can determine even if they can’t map the ad click directly, they can determine if that user if a user with that email address on Google clicked on an ad earlier that day that then resulted in that sale. And they can basically compare all that information and fill in some gaps in what would otherwise show up in the conversion tracking.
More accurate conversion tracking is then going to help the campaigns optimize a little bit better. And that slight improvement over time, particularly when your competitors have not implemented this, is going to give you a leg up.
Every platform has some version of this now. So on the Meta platform we’ve got this advanced matching. There’s a way you can pass information with your normal conversion tags. But there’s another way that you can flip the switch and just tell it to automatically look for email addresses. If somebody goes through the checkout, if you see somebody put an email address, go ahead and use that for trying to match users.
Very easy to set up on Meta. Increasingly, we’re switching to a world where we want to have duplicate conversion tracking going server to server on the back end that cannot be blocked by the browser. And this is actually pretty fun stuff. So a lot of people have heard about the Meta conversion API, and that’s one platform’s implementation of this and so on.
The Meta conversion API, we’re still firing the JavaScript tagging on the front end, but we know some of those are being blocked. In particular, Apple really likes to interfere with some of that stuff right now. So we also on the back end send conversion data through directly via API to Meta servers, and they can compare that with the front end data we sent.
Figure out what’s missing and match more of those sales to who clicked an ad just to figure out what what ad resulted in that sale.
I’m going to end the conversion tracking part here by just mentioning that idea about offline sales that don’t get counted. This can happen so many different ways with a with the conversion tracking just not firing for a sale. So could be finance forms. It could be lead forms on the site. It could be you’ve got a really awesome chat system and you’ve got, customer service reps, staffing that chat system and helping people.
And then that person completes the order on behalf of the user. So no conversion tracking is firing on the user’s browser. There’s all sorts of ways this can occur right. We need to think about ways to patch in some of this data to get a little bit more attribution for the ad platform. So again, the campaigns optimize better about who is actually converting.
Let’s say there’s like two key ideas here. One is you could use estimated values where you’re estimating the value of a lead, or you can actually upload real conversion data about what actually happened after the fact. The way to implement both of those is very different, and there are multiple ways that they can be implemented on different ad platforms.
Often on the same ad platform. There are some, some some of these solutions involve needing to track the click identifiers all the way through to a final sale and uploading that transaction data. Others are going to be easier where you’re passing some user information, but this is something that you need to solve. If you have this particular scenario, it’s going to really help the ad campaigns optimize more for these sales.
And frequently these are the larger sales. That’s why they have to maybe have a salesperson help them out with the issues. So just wrapping up now, I’m going to leave you with what are the key ideas here. Key idea is there is an optimal target return on ad spend. If you’re under that or over that, you’re leaving money on the table or losing money that you’re overspending on.
There’s an optimal level. Figuring out that optimal level for a business and targeting that consistently is going to definitely improve your profitability. Tiering campaigns by return on ad spend targets, by for different, for different groups of products with different variable margins is going to dramatically improve your profitability as well. It’s also about aligning the opportunity with the spending seasonal budgets, exactly the same principle, aligning the opportunity with, the spending that you’re actually doing.
And last, if you modernize your conversion tracking and take care and take advantage of these additional features, you’re going to have much better information to optimize and make decisions on. And you’re going to be you’re just going to understand the whole situation a whole lot better. So with that, I’m going to wrap it up. If there are any questions, I’d be happy to, happy to answer them.