Presented by David Deppner
Meet Magento New York, October 17, 2017
Competition for online advertising is driving up your costs, and it’s not going to get any easier. What can you do to improve your PPC advertising and get even more profitable? This presentation covers some common misconceptions that lead advertisers into lower returns, and explores strategies that will boost your bottom line.
If you're interested in this topic, then you might also be interested in my blog post on Why Return on Ad Spend (ROAS) Bidding Kills Ecommerce Profit.
My name is David Deppner. I’m the President of Psyberware and we do ad management for mid-sized Magento merchants. Previously some of you may know me from ClearBags, where I was Vice-President of IT and Ecommerce, and in that role I worked with Magento since 2008. I managed that organization through several major transitions, from a previous ecommerce platform to Magento 1 and then later on to Magento 2.
In the middle of all that, there was a point where I took over the management of our online advertising from the previous Vice-President of Marketing, who left the organization. Before that point, I really had a negative view of pay per click. It seemed to waste a lot of money, it seemed to not get a lot of results, and it seemed like most of the people on the marketing sides of various organizations and PPC agencies didn’t really understand how to drive profit. So I got really interested in the question of how to maximize profit coming from pay per click.
Today I’d like to talk about just a few things but I’m mainly going to focus on the financial side of this. It’s a huge topic and we’re just going to focus on the tip of the iceberg here. We’re going to look at some things about whether ROI is a useful metric, why you should never hit your budget, what you should have taken away from Econ 101 way back in high school or college that can really help you out here, and wrap up with just a little bit of discussion about the Key Performance Indicators most businesses use and maybe what you should use instead.
I have heard this a lot of times from a lot of ecommerce merchants – “I used to get 25 times ROI on my ads, now I’m only getting 5 times ROI. Is this even worth it? Is this even profitable anymore?” Frequently people don’t even know what is profitable. On the flip side of that, your typical PPC agency is now giving messages like this: “We guarantee you that you will get at least a 10 times ROI if you work with us.” I think this is really sloppy thinking, and I’m going to go into why that is first.
The thing to understand is that when people say “ROI” they really mean “Return on Ad Spend.” This is just a simple ratio of your income divided by your ad expense. So “10 times ROI” just means that I had 10 times the income compared to the amount that I spent on my ads. I want to point out that this isn’t ROI. This is a ratio of two income statement accounts. It has nothing to do with Return on Investment at all. It misses a lot of information. So what is ROI? We’re here in New York. Think about stock market investments. I spend $100 on some stocks, later on maybe I get some dividend income, I get some capital gains, I sell that and I end up with $200. That’s not a two times ROI, that’s a 100% return on top of the original investment. That’s ROI. ROI relates your returns to an asset that you invested in.
So which of these should you focus on in your pay-per-click? ROI or ROAS? The thing is, these are both very different. They give you different information. So let’s dive into a couple of examples that sort of get at this a little bit further.
If we were to take an example like “5 times ROI” – aka Return on Ad Spend, or ROAS – that might look like this. A business brings in $100,000 in revenue, spends $20,000 a month in order to get that, so you know at first glance the way people present this it almost seems like there’s $80,000 in profit. But that’s not true at all, because this leaves out all the other costs of business. The actual ROI can’t be figured out from this. The actual profit can’t be figured out from this.
As an example let’s say that the business had another $40,000 of non-advertising costs and we had $40,000 in net profit. Well, now we can at least say that we’ve got $20,000 in expense, we’ve got $40,000 in net profit, we’ve got 200% ROI? Again, not really. We’re dealing with income statement accounts here.
If we were really going to get to the ROI of a pay-per-click account, we would have to look at that pay-per-click account as though it were an asset. And I’d like to propose that it really is an asset. In the same way that your Magento installation is intellectual property, it is a business asset. You have invested in building this thing that is generating income for you. Your pay-per-click account is, as well. You’ve encoded all sorts of business rules in that pay-per-click account that basically turned it into an automated marketing system that’s always making sales for you, 24/7. If you were going to get at the ROI of that, you’d have to go through and you’d have to look at your original investment and your working capital tied up in it. You’d look at the profits coming off of that, and now you could start to work out an ROI.
The thing I’ll say, though, is that’s not actually useful. That doesn’t help you get to more profit. That’s just recording what has happened in the past. And there’s an old adage in business that I think is very relevant here – would you rather have 10% of $100 or 5% of $1000? The percent ROI is completely irrelevant. Ten percent of $100]is not more money than 5% of $1000. Ten dollars is a lot less than fifty.
We need to move beyond just thinking about things in terms of ROI.
ROAS really has absolutely no connection to profit. I’ve subtitled this slide “It’s Really More Of A Budgeting Method,” and the thing to think about here is that when people say “We’re going to get you a 10 times ROI” what they’re really saying is “We’re going to spend 10% of your revenue.” This can become a little bit of an interesting chicken and egg thing. Are they spending that 10% to get the 100% of revenue, or are they really just spending up to 10% after they get the 100% of revenue?
A profitable pay-per-click account is like a printing press printing money. They are constantly working, spitting out an endless stream of cash for you. Budgets and how we approach budgeting have a huge impact on how profitable they are, because as soon as you hit the budget the printing presses stop. If you’ve got ad accounts that are working for you, spitting out money constantly, you hit 5 o’clock on a Friday and your budget’s been hit, the ads turn off, they stop running, you’re no longer making money.
The simple answer about budgets is that you should actually set them higher than they need to be. And I’m going to get into the question of where they need to be. But budgets themselves should not be the mechanism that determines what you spend on your pay-per-click accounts.
So if you don’t set your spend based on budget, how much should you spend? This is one of the central questions that has interested me for quite a while, and the answer to it goes back to high school economics. Really this is something that anybody who took a basic economics class got exposed to a long time ago. There’s a principle about the theory of the firm, and profit maximization for a firm in a competitive market. You probably remember from high school, you had all these crazy slides and graphs, and nobody likes these. I like them. I’m kind of a closet econ geek. I love this stuff, but I recognize that most people hate these things and have really bad memories, so I want to try to build up to this concept with a very hands-on pay-per-click example instead of using abstract graphs.
Pay-per-click advertising is always purchased in auction markets, and a central characteristic of an auction is that in order to buy one more of anything, you have to out-bid every other competitor that wants to buy it. This has the effect over time of steadily pushing the prices up on PPC. It also means that during a given time period – like during a given week or a given month – if you want to buy more clicks or more conversions, you have to incrementally spend more and more in order to get it.
Let’s just look at a simple example of how cost per acquisition might rise. You buy your first conversion in a week and maybe that costs you $10. In order to get two conversions that week, you’re going to have to spend more on that second conversion. You’re going to have to increase your bids. The second one costs you maybe $20. Maybe the third one costs you $30. If you ramp up because you want to get five conversions this week, you’re going to have to set your bids at a level where you’re spending $50 in order to get that fifth marginal conversion.
Now, this really matters. Most of us think about cost per acquisition as an average, and the average of these five conversions is $30. If we’re in a business where we know that we can spend up to $30 to get a customer, but we’re actually spending $40 to get some customers and $50 to get some, we’re actually losing money on some of our sales. But that’s obscured by the fact that we’re measuring this using an average cost-per-acquisition metric.
A lot of businesses make this mistake. A lot of businesses overspend because they’re obscuring the rising marginal cost of acquisitions in an average cost-per-acquisition metric. That marginal cost per acquisition is far more important in making profitability decisions. That’s not reported anywhere. You have to get at that yourself by digging into the numbers, sometimes extrapolating it from other things. But the marginal cost per acquisition is what matters.
So let’s go back to our earlier example. A hundred thousand dollars in revenue, $20,000 in ad costs, $40,000 in non-ad costs, and you got a $40,000 profit. How much should we spend on advertising? It’s easier to think about this in terms of individual conversions. So if we took those same figures and we just broke it down in this example to say there’s $100 average order value, so that would be a $20 cost per acquisition – average cost per acquisition – you’ve got $40 in non-ad costs, and we’re pulling out $40 in profit. In order to look at the micro-economics going on here, we actually need to back out the ad costs. We need to look at what is the profit coming in from a conversion before our advertising expense. So in this example I’m just going to take out the $20 ad cost and add that back into the profit. So before we spend any money on advertising, we’re making $60 per conversion.
This graph has a lot going on. I’m going to take a minute here. I want to make sure it makes sense. So I’ve put a line across just showing where that $60 profit per conversion is, before the ad cost, and then used the same bar graph I was using before to show how with each additional conversion your cost per acquisition is rising. And so you can see when you’re buying one conversion for $10, you are actually pulling in $50 in profit off that conversion. Your next conversion at $20 is pulling in $40 of profit. Once you get up to your fifth conversion you’ve only got $10 in profit but you’re still making profit. You hit that sixth conversion, it’s a wash. If you buy a seventh conversion for $70, you’re spending more than the amount that you can spend profitably.
This right here is what gets lost when you’re looking at averages across all of these numbers instead of understanding what happens due to your marginal cost per acquisition. Your marginal cost of acquisition shows that it’s profitable to spend all the way up to 5 or 6 conversions. You could get into “Should I buy that 6th conversion or not? I’m not making any money.” Well, I think you might as well so your competitor doesn’t get it.
So this is where these line graphs in economics come from. It’s just building up all that data and putting it in a few lines that illustrate the simple point. You got marginal profit on one line, your marginal cost on the other, and the point where those two lines intersect is the point where you’re going to get the most profit that you possibly can. And you can actually model this very precisely with the data that’s available in your pay-per-click accounts. That point where you’re going to achieve maximum profit is basically the point where you want to set your bids. You set your bids, not based on your budget, not based on any theoretical average cost per acquisition that you’re trying to target, you set your bids to that point where you’re going to increase your profit to the maximum possible point you can get out of your account. And notice at that point, if you set your bids to go higher and buy more conversions you start losing a little bit of money at the margin, but also if you spend less than that, you’re not going to be making as much profit as you could have. The term in economics is the opportunity cost. You missed out on that opportunity.
I think this is a profound idea that can be a little bit disconcerting to people, that one of the primary mechanisms by which you get profit in your account is to set your bids well. If you are setting your bids based on things like your target cost per acquisition, which is really a target average cost per acquisition, or target ROAS, or if you’re just setting bids $3 here, $7 there and guessing, you’re above or below this point. I have seen many pay-per-click accounts for clients where just doing a financial analysis and setting the bids properly doubled the profit overnight. Just changing bids – no other changes.
Please – ignore your average cost per acquisition. CPA doesn’t matter. Marginal CPA matters a great deal. Remember – there are only two things you need to know to do this. You need to know the profit of your conversions before your ad costs, and you need to know your ad costs at the margin.
I’ve already talked about why ROAS is not a good metric to use. I’m going to make a couple more points about CPA here, but also a lot of people set KPIs based on achieving budgets or ad spend or achieving certain revenue goals. I just want to put the nail in the coffin on all of these. Remember that KPI means Key Performance Indicator. A KPI should either influence your performance or it should measure performance that you’re getting. But it should be directly related to the real underlying performance, and none of these metrics relate to profit directly.
Here are a few of those crazy graphs, just to illustrate a point. If you plot the Return on Ad Spend at various points along that example that I used before, you can see that if you’re trying to increase your ROAS like when people say “I want 10x ROI or a 20x ROI” you’re actually frequently moving down to a point where you’re leaving profit on the table. You’re leaving money on the table that you could have gotten if you got closer to the margin.
Cost per acquisition has no correlation with profit, either. People who are trying to reduce their cost per acquisition to a certain arbitrary point are missing the mark. If you were bidding at that point of maximum profitability and you tried to drop your cost per acquisition to save money, you’re going to move down from that point of maximum profitability and leave money on the table.
Reducing ad spend? Exactly the same issue as reducing cost per acquisition. It’s not actually correlated to the point of maximum profitability.
And increasing revenue beyond that point has the same problem. You just start increasing your costs faster than you’re increasing your marginal revenue. It doesn’t work.
I want to start to wrap up here with this take-away, that profit is actually the best KPI to measure profit by. None of the other proxies that marketers and advertisers use in this industry are really adequate. They’re not as good as understanding your business model, understanding how pay-per-click auctions work, and then doing everything you can to squeeze every last dime out of them.
So this presentation really has been about boosting your bottom line in the context of pay-per-click, but I realize we really just scratched the surface of this and we didn’t go into a lot of detail. There’s a lot more behind the scenes if you’re dealing with probability and statistics, dealing with more economic principles, but at least take away a few things from this. Ditch the KPIs that most people are using – ROI, ROAS, and Cost Per Acquisition don’t help. Remove your budget constraints so you’re not limiting the profit that you’re able to get out of your accounts. And take the time to understand just the basic micro-economics of profit maximization. Make sure that the people working on your pay-per-click accounts do as well. That allows you to bid at the point where profits are maximized. That can have an immediate and dramatic increase on your profitability. But ultimately, your pay-per-click manager, whether that’s an in-house employee or an outside agency, absolutely has to understand your business model. They have to understand your cost structure, or they actually can’t make profitable decisions inside the PPC accounts. There’s frequently a disconnect between the financial side of businesses and the marketing side of businesses. Marketers frequently focus just on that top line growth. They don’t understand the costs, the marginal costs, profit margins, all of those things that go into making profitable decisions. They actually are far less profitable than they could be.
I’d love to talk to you if you’d like to chat more or get in touch. Send an email or reach out on Twitter. I’ve written a number of blog posts on these topics that are at Psyberware.com/blog, if you’d like to dive a little bit deeper into some of these ideas, and I’ll be posting several more that relate to this talk over the next couple of weeks, as well.