In a word, “No!”
In two words, “Hell no!"
In eight words, “No, because of the law of diminishing returns."
CPA, or Cost-Per-Acquisition, is a metric based on averages. It’s simply the average cost per conversion. If you’re not really clear on what this means for your decision making, it can lead you into making mistakes that will kill your profitability.
Part of my frustration with CPA is that it’s used so widely.
Merchants frequently set a budget for new customer acquisition for their marketing departments.
Google wants to capitalize on this and just have you set a CPA and let them manage your bids in order to achieve that target.
- Ad agencies like to report on this metric, showing how their CPA is below the client’s target cost.
All three of these players have it wrong.
Online bid-based advertising is a context where you will see declining marginal returns. This means the first clicks you get are cheap. As you (and your competitors) fight for more clicks, you have to increase your bid and drive up the cost. The central fallacy with CPA-based decision making is that your costs go up as you try to scale, but you’re averaging those increased marginal costs across your entire result set.
What matters isn’t the average cost, but the marginal cost.
A Simple Example
Let’s say that you currently spend $1000 per week to get 1000 clicks. You’re getting 10 conversions, so that’s a 1% conversion rate and your cost per acquisition is $100.
But let’s say that your company is willing to spend up to $150 to get a new customer. If you’re getting 10 conversions at $100 each, you know you could get more if you were willing to increase your bids and use all of your budget.
So, you increase your spending to $2000 per week by increasing your bids. The thing is, you are already getting 1000 clicks per week for $1000. As you go past that level of activity, you’ll only get more clicks if you’re willing to outspend other companies that are currently bidding more than you are. As you raise your bids, the additional clicks cost more than the previous average. So perhaps you end up getting an additional 500 clicks for your additional $1000, at the same conversion rate. Now you’re getting about 15 conversions per week, but spending $2000. Your CPA is now $133.33.
But here’s the rub: the average cost across all your customers is now $133.33, but that’s bunk. You were already getting 10 customers per week at $100 each. You’re actually getting an additional 5 customers, at an additional cost of $1000 per week. So your _marginal_ cost per each additional conversion is actually $1000/5 = $200. Your 11th conversion is less than $200, sure, but your 15th costs even more than $200… and possibly even much more.
If you were willing to spend up to $150 to get a new customer, and now you’re spending $200 or more to get each additional customer, you’ve made a bad business decision. If you’re making decisions based on CPA metrics, that fact will get lost in the averages. You’re averaging some customers that cost you far too much with some customers that you’re getting at a great price, and the $133.33 CPA metric will lead you into losing money. If you’re willing to spend up to $150 per new customer, the average cost where the next marginal customer costs $150 will be far lower than that amount.
When you’re in a situation with diminishing returns, you can’t make decisions based on a metric that’s an average.
You have to make decisions based on the effects at the edge. With AdWords, the cost per acquisition increases as you scale. Any reporting based on that average is also flawed, if those reports are used to justify additional poor financial decisions.
Why Merchants Get it Wrong
Ecommerce companies get this wrong because they use simplistic metrics for their marketing departments. Too often, goals are set based on Marketing needing to get a certain number of customers at a certain cost. In our example, perhaps we earn $500 in profit on every customer, but we need to keep our Marketing budget in line, so we say we’re willing to spend $150 to get each customer, knowing that leaves $350 in gross profit margin to cover our product costs and overhead, and still turn a net profit for the business.
The Marketing folks then turn around and treat that $150 target as the real goal, forgetting that the real goal is company profits. And, because of some sloppy thinking, then get a bunch of customers at $100 each, and a few more at more like $200 each, but for a total average of $133 each, and they pat themselves on the back that they were able to deliver customers at an even lower CPA than the company’s goal.
Why Google Gets It Wrong
Google doesn’t actually get it wrong. They get it right…for them.
Google’s motive is obvious. If they convince you to turn on CPA bidding, and you put in a $150 CPA figure, they’ll happily take all of your money, right up to the $150 you’re willing to spend per new customer. You’ll get some customers at $5 per conversion, and some at $500 per conversion. But it’ll average out in the end, so Google will happily deliver what you say you think you want, without pointing out how much you’re spending for that last marginal customer. It’s good for them. More revenue. Even more, it’ll force your competitors to also up their bids. Even better (for them).
I love Google. Google has made the world a better place. But let’s be real. They’ve engineered this game to be a never-ending escalation of bids between all of us that ultimately benefits them more than it benefits us. Good for them (buy their stock!), but be very careful with your own business decisions.
ProTip: Never, ever, under ANY circumstances, EVER, let any Google employee “optimize” or “accelerate” your account based on “best practices”.
Why Ad Agencies Get It Wrong
I think most ad agencies get it wrong because they’re trapped between clients expressing goals in terms of CPA and Google reporting based on CPA. It’s such a common goal from clients and such an easy metric to report on, that agencies have no choice. It’s the lingua franca. So they unwittingly play a part in driving costs up for merchants and optimizing AdWords accounts based on a fallacious premise.
It’s easy for ad agencies to just do what clients ask for, and forward along simple metrics in their monthly status reports.
What Should You Do?
Online advertising decisions have to be based on one simple metric: Profit. Does this ad spend increase the company’s bottom line? Does it generate more profit? Or does it lose money?
If you’re currently spending $1000 per week getting 10 clients at a CPA of $100 and you’re willing to spend up to $150 to get each new customer, then you have to go a bit deeper and do some financial analysis which is a bit more difficult than a simple CPA calculation.
The key questions for you are:
What was the marginal cost of that 10th customer? Was it under $150?
If so, how much further can I push the bids and stay profitable?
- If I’m earning $500 in gross profit margin, does my $150 target even make sense? (Hint, “no”).
Optimizing AdWords to generate the most possible profit takes a different kind of decision making. It takes an understanding of the cost structure of the company, how results diminish as bids go up, and how competitors respond to your decisions. If you’re making decisions based on simplistic metrics like CPA, you’re not maximizing your profitability. You’re either spending too much, or you’re leaving money on the table.
If you’re making decisions based on CPA or working with an agency that’s reporting their “success” in terms of CPA metrics, you’re not maximizing your profits.