Choosing goals and targets are some of the most fundamental decisions in paid search. Unfortunately, they are often chosen somewhat arbitrarily and rarely re-evaluated.
Many traditional advertisers will often set CPA targets based on product revenue and margins alone, without taking incremental CPA and diminishing returns on profit into account. Let’s consider the following example:
An advertiser with a $100 product and 25% margins has $75 in profit per conversion. He or she wants search to be a profitable channel and decides on a $50 CPA target with the goal of maximizing conversion volume at this CPA. While ultimately profitable, the problem with this approach is that the advertiser in question is optimizing towards conversion volume at a static CPA target, rather than a level of profitability.
An alternative advanced search strategy uses the concept of ‘incremental CPA.’ To break it down simply, additional conversions gained through bid increases come with higher and higher CPAs. It is somewhat intuitive but often overlooked when CPA targets are decided.
How Incremental CPA Works
Let’s take a look at performance by rank for a set of queries with similar behavior in our example account. For the sake of simplicity, conversion rates have been standardized. “Profit” = (Conversions*$75)-(Cost)
You can see here that our keywords in position 1 are below our CPA target. However, in this case the account is losing profit by bidding aggressively to position 1. The additional 101 conversions driven in rank 1 are coming at a cost of $8,439.
If we’re looking to maximize profit, the incremental CPA should be no greater than the break-even on post-margin profit – in this case, $75. This happens somewhere between positions 1 and 1.5, as indicated on the x axis below.
Use Profitability Curves
It’s still useful to target CPAs since they feed nicely into existing bidding strategies, and you cannot bid for profit directly. The next step is to figure out at what CPA profit is maximized; you can do this by plotting profit at different CPAs (the values of the x axis, below) and looking for the maximum profit value of a trend line.
What’s Really Going On
The main factors controlling the incremental CPA and profitability here are the differences in the rates of increasing CPC and CTR by rank, as shown in the graph below (the x axis values are for ad position):
In this case, the CPC is increasing much faster than the CTR at the highest positions, and profit is decreasing at those positions. This won’t always be the case, and in situations where the CTR is increasing faster than the CPC, the profitability curve will be shifted.
When to Optimize for Profit
The profit-maximization approach isn’t for everyone! It definitely has caveats:
– The main drawback is the very large data requirement – you need a lot of steady state performance data at various ranks on high-volume queries to map CPC and CTR accurately.
– Including user lifetime values and the availability of products with different revenue values and margins makes this much more complicated.
– At the end of the day, a flat CPA target based solely on margins is profitable for a business. The amount of additional money you can make/save here may not be worth the effort involved.
– Remember that growing businesses may opt to take a slight hit to profit in order to drive more overall transactions or orders. Incremental CPA wouldn’t matter here.
That said, if you have a lot of data and are concerned only about the profitability of your business, this is the best road to take.
Questions? Thoughts? Please leave a comment!