Features

The Death of CPM

Written on
Apr 15, 2014 
Author
Brian Barnum  |

ADOTAS – As the success of ad tech has continued to surge in the past year, one of the questions that has been on everyone’s mind is: Are the old metrics still working?

In general, the consensus seems to be “no.” As user behavior changes, our understanding of how to measure to success is shifting accordingly. This also impacts how we think about the value of display ads and how we should pay for them. One of the biggest rocks we have to move is whether or not we should continue to measure and pay for impressions the same way. Currently, advertisers purchase impressions by the thousand, which is why our standard unit of measurement is CPM.


Implicit in the CPM metric is the assumption that advertisers need to buy impressions by the thousands so that their ads will reach just a few potential buyers. The idea is that if you reach enough people, you’re bound to at least get a few conversions. As has always been the case with advertising, you’ve got lots of opportunities to hit eyeballs, but no guarantees about how many will convert to buyers.

Essentially, trying to reach a potential customer with display advertising is like looking for a needle in a haystack. The problem is you’re not just looking in the haystack – you have to buy the whole haystack first. In other words, you’re charged for every thousand impressions, even if you only reach one (or zero) possible buyers.

Working Within the Limits

For B2B, this is an unreliable tactic but historically, this has been a reasonable gamble for B2C. There are many potential consumers out there and it’s much easier to drive a sale. (Purchasing Oreos doesn’t require that much deliberation.) On the flipside, that’s why we never saw many B2B commercials on TV or full-page displays in magazines: The likelihood that you’ll catch a potential customer in a wide net is much greater for B2C than it is for B2B.

But even B2C companies have leveraged different types of targeting, and as technology improves, our ability to target is growing exponentially. Advertisers, especially those in B2B, have the ability to choose quality over quantity and they are willing to pay for it. The metric that matters to them isn’t “how many thousands of people saw my ad.” It’s “how many of the right people saw my ad and furthermore, did they engage with it?” Since it’s easier than ever to focus on quality, CPM is no longer a metric that advertisers can feel comfortable bragging about. After all, having to admit that you reached 999 of the wrong people isn’t too impressive.

Ending the Ambiguity

Of course the fact that we know we’re reaching the wrong people is a breakthrough. After over a hundred years, the old adage from John Wannamaker, “half my marketing dollars are wasted; I just don’t know which half” is no longer accepted as a necessary truth. CMOs, CEOs and of course, CFOs, want to know how effective advertising spend is, and whether or not it’s actually helping deals close.

With programmatic technology and third-party data, the level of detail is finally available. But the question is: How SHOULD we pay for ads now that my audience is far less random? As a CFO asked to approve advertising, I want to know the ads are creating value and part of the cost of sale. I don’t want to buy a haystack, and I want to see metrics that tell me more about the success of my programs and the growth of my business.

I’m not alone in this, and the economics of advertising are in fact shifting. Because we’re now have the technology to target the specific customers we want, it follows that we should only have to pay for that segment. We’re willing to pay for quality, so we need a model that enables us to do that. As that model begins to take off, it may kill off CPM for good.

This is especially true in B2B, where long sales cycles require continuous engagement. Running one campaign won’t have much of an impact on overall decision-making. No matter how many impressions I buy, it’s very hard for me to connect ad spend to deals that are closing.

Understanding the Goals

It’s starting to take shape, especially in B2B. You’re trying to build relationships and consensus within a specific account over a period of time. Alternatively, in B2C, you’re trying to make sure that a particular message reaches as many people as possible. In B2B, if you message reaches as many people as possible, 85% of them will likely be consumers who are of little value to your business.

What you want is to know that you’re talking to the right audience, delivering the right message at the right times throughout the sales cycle and building a relationship throughout. Whereas once you had buy thousands of impressions to reach only a few, now it seems fair that you should be able to pay based on who you’re trying to target and what you’re trying to achieve.

New Measurements

Once you’ve allocated budget to deliver messaging to a particular target, our success metrics are no longer ambiguous. Furthermore, the seamlessly integrated into our overall evaluation of broader marketing programs. We look at:

  • Engagement: Has the targeted account visited my site more, and consumed more pages while they’re there.
  • Conversions: Have they moved to the next phase of buying cycle via hand-raise, RFP or booking a meeting?
  • Stakeholder awareness: As the team moved through the pipeline, do they demonstrate an understanding of the messages they’re received.
  • Pipeline growth: Ad activity can be measured against CRM data to determine whether ad budget has to be properly allocated to accounts that are driving value.

While this shift can feel uncomfortable to people have been doing things the same way for years, the work will ultimately pay off. Publishers and media buyers will be able to deliver better value, brand advertisers will have the power to directly impact growth and continually optimize their programs, and of course, CFOs will be more than happy to approve ad budget because they know they’re getting needles, not haystacks.





Brian Barnum is Chief Financial Officer at Demandbase, responsible for the company's finance and accounting, legal, and corporate development functions. Prior to joining Demandbase, Barnum served as the CFO and COO at the Rubicon Project, and prior to that had been the COO and CFO of Business.com, where he oversaw the acquisition of the company by RH Donnelley (now Dex One Corp). After the merger, Barnum was named President of RH Donnelley Interactive. Barnum was previously CFO of Rent.com, where he prepared the company for an IPO while managing a mergers and acquisitions process that culminated in an acquisition by eBay. Before Rent.com, Barnum had been the CFO of Transaction Network Services, where he executed an IPO, a secondary stock offering, and consistent growth as a public company.

Barnum earned his BA in Economics from Stanford University and his MBA from The Anderson School at University of California, Los Angeles.

Reader Comments.

Interesting article, especially from my perspective as a B2C publisher. I couldn’t agree more regarding your comment of advertisers wanting to know they are connecting with the right audience, delivering the right message at the right time.

What I’d love you to expand on is what new model would be a likely candidate to replace CPM. While the wrong model, CPM is easy to understand. What is the model that acknowledges quality over quantity? How do we transition to it in a way that will provide a predictable revenue stream for the publisher?

Are CMOs, CEOs, and CFOs, really ready to pay for quality? Will they pay a premium price for solid engagement that converts to business growth?

The risk for publishers in what you suggest is that poorly executed ads wouldn’t have to pay for lack of engagement. The advertiser then blames the media rather than a lousy pitch. The publisher loses because the ad space was taken up by a bad ad which is wholly outside of their control.

Understand that I’m not a fan of flat CPM pricing. However, an alternative new measurement must take into account the reality that not all ads are born winners. Any new model should take this into account. Any suggestions?

Posted by Major Webuser | 12:45 pm on April 15, 2014.

Hi Major-

I do have thoughts on the questions you raise.

First as to what will replace CPM? Well CPM may always be a transactional reality – we need a “currency” to transact in and lots of systems are built around CPM. The point I am really making is that data driven buy-side technology platforms can handle that and create other “currencies” that meet needs of specific advertiser verticals and define the “exchange rate” of that more relevant currency into the mechanical reality of transacting in a CPM environment based on the costs of units relevant to the activities in that vertical.

Let me give you an example, my company, Demandbase. We service B2B customers, mostly mid-sized and large enterprises that are selling to other mid-sized and large enterprises. At any point in time our customers are probably targeting hundreds, perhaps several thousand target accounts, or other companies – they don’t sell to individuals but to accounts and these accounts typically have long and complicated purchasing processes that involve a number of people across the organization. We have proprietary IP that allows us to identify and target advertising at the company level, and then serve ads to the people from that company out on the web. It’s very valuable to our customers because the vast majority of the traffic out there is not only not B2B at all but consumers and most of the B2B traffic will still not be one of their target accounts. With our proprietary data and our real-time technology, we can specifically target users from target accounts or from precisely defined target segments of companies.

Our customers think in terms of target accounts, have their CRM systems built around targeted accounts, have their sales compensation defined around closing target accounts – why shouldn’t they be able to advertise and measure in that same “currency”? We offer our technology on a subscription basis that customers pay for based on the number of accounts they want to target, not on CPM, and provide all the measurement and analysis against results from those target companies, because that is what makes sense in B2B.

On the topic of ad creative – I agree it’s critical and has been under-developed in digital. I think of creative as what Madison Avenue is all about – sending the “right message” that captures the attention of the audience. Data and technology now allow us to send the message to the “right audience” at the “right time” and play a role in knowing enough about the user to provide the opportunity to send the “right message,” but the creative has to deliver. Now that the technology has allowed us to really have an ongoing dialogue with a relevant audience that should care about what we say, this should allow the ad industry to focus on nailing the creative that allows that dialogue to occur and to capture our audiences’ imaginations.

Posted by Brian Barnum | 5:18 pm on April 15, 2014.

Leave a Comment

Add a comment

Tags: , , , , , , , , , , , , and
Article Sponsor

More Features