Price Elasticity: Why Your CPA Is As Broke As A Joke

marc-grabowski-better“Data Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media. 

Today’s column is written by Marc Grabowski, Chief Operating Officer of Nanigans.

Not all customers are created equal. And if you are acquiring customers on a CPA basis, you are making three major mistakes:

  1. You are missing the customers who represent the highest lifetime value
  2. You are only able to afford customers during supply rich periods
  3. Somebody is taking advantage of a margin on your spend and is incented against acquiring your most valuable customers.

CPA based bidding is one of the biggest farces in advertising. It assumes each customer generates the same value, when the reality is some purchase once and others purchase repeatedly. CPA based bidding incentivizes targeting lower quality customers who monetize once, versus those who deliver value over the long-term, encouraging opportunistic buying.

But there’s more to the story – price elasticity. In advertising, price elasticity refers to the ability to bid up or down on inventory as market conditions change, like during the holidays. Advertisers with rigid customer acquisition models like CPA are limited to supply abundant markets. Advertisers with models that allow for greater price elasticity, such as advertisers optimizing for the actual revenue generated from their campaigns, can take advantage of inventory in both supply-abundant and supply-constrained markets.

The graphs below show just this – an advertiser with rigid CPA-based bidding lost out on significant new registrations during the holidays when inventory demand was high, while an advertiser optimizing instead on revenue was able to continue to generate high-quality registrations during the same period.

CPA-Based Advertiser

As shown in Graph 1a and Graph 1b, this CPA-based advertiser was unable to adjust to spikes in market costs during the holidays. When prices rose (11/6-11/8, 11/16, 11/21 and 12/1), registrations plummeted.

graph1a graph1b

The registration drops happened when inventory prices increased. The advertiser’s rigid CPA pricing structure assumed all customer acquisitions were valued equally, regardless of the fact that some acquisitions were generating far greater return over time.

Graph 2 further highlights the point above. The advertiser was unable to spend, as their hard CPA was more rigid that the market tolerated.


So, not only does CPA based customer acquisition suffocate the ability to reach your most valuable customers, it suffocates the ability to take advantage of market conditions. And this can lead to an advertiser under-valuing its best customers during peak sales periods.

Revenue Optimized Advertiser

In contrast, an advertiser optimizing on the actual revenue generated from their ad campaigns was able to continue to spend and drive incremental registrations and revenue when costs spiked between Black Friday and Cyber Monday (see graph Graph 3 below).


The price elasticity in revenue-based optimization allowed this advertiser to absorb a cost increase of 21% on Cyber Monday vs. the prior day, all while achieving a single day return on ad spend (ROAS) of 131% on Cyber Monday – higher than their 121% ROAS on the day prior (see Chart 1 below).

In other words, revenue increased at a higher rate than price.


This case study makes clear that CPA is not a proxy for lifetime return. In fact, it discounts differences between high and low value customers, particularly those that often result from seasonal shopping. Revenue optimization, on the other hand, allows advertisers to take advantage of more inventory, regardless of changing market conditions.

So as you assess your advertising budgets in 2013, consider the opportunity loss that CPA-based optimization represents and the incremental revenue and value that would be achieved with revenue-based optimization.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!


  1. Fantastic post. Completely agree with you Marc. Have been telling this to many, but it falls on deaf ears. There is a serious lack of inventory / pricing / ROI understanding amongst not just marketers but many digital folk.

  2. We at Optimal agree, good analysis. But some of the market might not be as capable of doing this analysis or want to — smoothing of results. I think it’s crazy myself, but psychology like this exists:

    An exec at one of the DSPs we compete with on FBX told me that many of the DSP’s customers knew they were making big profits from buying their media at a fixed CPA, but they valued it because it removed some of the variability from daily results which left less to explain to bosses, clients, or partners. He told me: “If you want to hit a $40 cost per sale, and the DSP starts running it, sometimes you get a $12 cost per sale and the next day a $90 cost per sale – a lot of these agencies can’t handle this variability. But this is life. Sometimes it’s just easier to have a fixed CPM or CPA or number of impressions. Many of their systems are not set up for variable amounts.”

  3. Alan Brody

    Interesting, i assume the above data is not from RTB inventory/buying?

  4. It’s always great when numbers speak, and great article to explain it all. Now only if we could get people to start adopting real attribution models beyond CPC, CPA and last touch. Sadly, I don’t see that happening for most large advertisers. The CPA pressure is reflective of the pressure on the CMO from Wall Street to deliver greater sales today rather than build for tomorrow.

  5. Peggy Reinders

    This model works great for retailers that have the ability to optimize based on revenue. How do you see this working in CPG where there is no easy direct tie to sales?