Fundamentally, with pay-per-click advertising, the idea is to balance the cost per click that we’re willing to pay with the profit per click that we expect to achieve. As advertisers, we attempt to find the right level of spending that maximizes the business profit. And since every business is different, it requires us to understand a bit about the business model. In reality, we don’t need to know all the details. But we do need some simple rules of thumb when making decisions.
Microeconomics 101 – Declining Marginal Returns
PPC advertising systems built on auctions, like Google Ads, all have the property of exhibiting declining marginal returns, primarily because the cost-per-click goes up as you buy additional clicks. For example, you might get $500 in profit from the first $100 you spend, but only $250 in profit from the next $100 you spend, and then maybe only $125 from the next (for a total of $300 in ad spend and $875 in profits). Now, the last $100 you spent would have been worth it because you would have still covered your ad costs. But, if you continued on and the fourth $100 you spend only brings in a $50 in profit, the profit is no longer exceeding the ad cost, and so that would be a bad purchase. This is illustrated below.
Microeconomics 102 – The Flaw of Averages
This reality can be obscured by averages if you are lumping all of that together and not paying attention to your marginal results. This is one of the biggest decision-making fallacies people make in this industry. In the previous example, if you spent $400 total and got $925 in gross profit, you might think that still made sense, because after the ad cost, you still made $525 in total net profit. But at a lower spend of $300 you would have received $875 in profits, for a contribution margin after ad cost of $575. The lower spend was the optimal point to achieve maximum profitability.
This is why making decisions based on averages is often bad. For example, Cost Per Acquisition, Average Order Value, and Return On Ad Spend metrics are all broad averages that obscure the profitability effects at the margins. Any business goals expressed in terms of these metrics will drive decisions which reduce profitability either by spending too much, or missing out on easily achieved profits.
Microeconomics 103 – Business Overhead Is Not Material
The only costs that are material in these marginal profitability decisions are costs that are directly attributable to an additional sale. That’s because we’re making micro-decisions about whether to bid a given keyword at, say, $1.50 or $1.60. The result of this decision might be that we get an extra 5 sales over the month, but has no impact on business overhead.
The only relevant costs are the costs associated with those additional sales. For most ecommerce companies that would be cost of goods sold, the expense of free shipping for businesses that absorb those costs, and potentially some other minor direct costs associated with a marginal sale. Businesses using activity-based costing shouldn’t include costs in these figures that spread overhead out to individual orders, since these don’t actually increase at the margins. In the long run, all business overhead is obviously material. But when making many tiny marginal adjustments each month, business overhead is not relevant.
Microeconomics 104 – A Dollar of Conversion Value Is Never a Dollar in Profit
Google Ads has no idea what your business model is or what your product margins are. So it happily reports on Total Conversion Value of all Conversions attributed to Google Ads. For a pure-play ecommerce business where we need to make decisions based on the marginal profitability of sales achieved through direct-response advertising, we need to deflate this figure by the costs associated with those sales in order to make decisions about how to optimize for increased profit.
But many businesses may actually be on the other side of this. Instead of deflating the Total Conversion Value to work with a lower gross profit margin, we may want to make decisions based on a multiple of the Total Conversion Value. Let me go through a couple of cases of what we’ve done with other businesses and talk about how we think about different situations.
Case 1: Direct Response Ecommerce
We have a client that only sells online. They sell a commodity product with no brand loyalty. People search for it as a generic product, buy it, and they will likely never hear from that customer again. They have 50% gross profit margins approximately. So for them, $1 in revenue is actually worth 50 cents in gross profits. If they are spending 55 cents at the margin to get $1 in revenue, they’re losing money (because it’s bringing in 50 cents in gross profit which doesn’t cover the ad costs and result in any contribution margin). But if they spend 44 cents for $1 in revenue, that’s still worth it.
Case 2: Growth Company
Another client focuses on maximizing the long-term profitability of their business, and has no short-run cash constraints. The gross profit margin is high—well over 50%. Happy customers frequently turn into repeat purchasers. This company really wants to get a lot of new customers and keep them coming back. Since the long-term revenue stream from customers is significantly higher than the first purchase, and this company wants to maximize shareholder value over the long term, we worked with them to model lifetime value of new customers into the business, on average, and express that in terms of lifetime value of the profitability of a new customer (not the revenue).
We worked with them on a system to match their transactions from Google Ads with their back-end ERP, so they could model the new vs. returning customers from Google Ads. Then we were able to work out the value to the business profit of each dollar of Total Conversion Value reported in Google Ads. Initially we revisited this math quarterly, but it has been fairly stable, so now we run the numbers annually and tweak the multiple a bit if necessary. Since there are no short-term cash flow constraints, we are often spending more to acquire new customers than they are making in gross profit on the first sale. But this is not a problem as it has greatly accelerated their business growth, and they break even after a few months on the repeat purchases, and after that they earn a huge amount of profits.
In reality, we are undervaluing their marginal profitability on new customers because they also have high word of mouth. So we’ve seen a surge in new customers from Google Ads followed by a surge in new customers from word of mouth.
Case 3: Bricks and Clicks
What if your ads are driving ecommerce sales as well as sales in physical stores? We might need to make a few assumptions to develop a reasonable model that we can use in decision-making, then run with it for a few months and check the results to test our assumptions.
For example, let’s say that gross profit margin on sales is around 60% on the site. So, we would start out with that as our baseline. But we also estimate that for every dollar of Google Ads revenue online, we’re influencing another $5 in revenue through retail. Maybe we’ll arbitrarily say Google Ads gets 50% of the credit for that, and the retail sales have a lower 40% gross profit margin.
\[$5 \times 50\% \times 40\% = $1\]We then use a model that $1 in Total Conversion Value is actually driving approximately 60 cents in online gross profit margin and another $1 in offline gross profit margin for $1.60 total. So we would make Google Ads profitability decisions understanding that $1 in Total Conversion Value is actually worth 1.6X in gross profit to the business.
What Is A Dollar Of Total Conversion Value Worth To You?
If you want to focus only on direct response advertising to achieve direct ecommerce sales, you can simply work with the gross profit margin of products, on average, or a ballpark estimate of COGS and other direct costs of marginal sales. But every business is different. If you sell products that are more like subscription services, with lots of repeat purchases you would want to steer towards a lifetime value model. If you have a lot of retail stores and web traffic drives sales to them, a completely different model where you anticipate greater conversion value than you measure through online purchases would be most appropriate.
Obviously those numbers are just examples to illustrate how to think about this. But it does illustrate how we can make a few reasonable assumptions and work out a rational model for making decisions designed to drive increased profit and growth to the business. Some of the assumptions will prove wrong in the future, but it’s still worth attempting to make them. Run with them for a period of time, and then see from the data how much they need to be adjusted.
Ultimately, growing your business profitably through online advertising requires a sophisticated understanding of your own business model and your financial model for how the online advertising contributes to your business. You have to go beyond simplistic thinking that doesn’t actually drive profit, and approach this with a bit more rigor than we typically see in this industry.