• by Ari Rosenberg , Op-Ed Contributor, February 28, 2013

Three CPMs walk into a bar: Print, TV and Online.  The Print CPM goes to order drinks and when he turns to ask the others what they want, he only sees the TV CPM standing next to him.  The Print CPM asks the TV CPM, “Where did the Online CPM go?” 

The TV CPM responds, “The toilet.”

All bad jokes aside, CPMs reflect a medium’s perceived value. While other media like television have experienced significant growth over the past ten years, “ours” has completely nosedived over that same time period.

The issue of excess inventory deflating site eCPMs (Total Site Revenue divided by Total Site Impressions “available for sale” x 1000) is not being ignored here -- but every medium can influence its own supply. When we collectively made the decision early on that our content wasn’t worthy of consumers paying for it, we chose to open up the supply floodgates that can now never be closed.

Today we talk a big game about big ideas and big data.  We expound on the power of native ads, RTB and now programmatic reserved, but have any of these innovations increased site eCPMs -- or have they just changed which hands collect the nickels and dimes our inventory is worth in the eyes of advertisers?  How long can we ignore the truth: Our CPMs are embarrassingly low.  Recent reports have online display CPMs rising off the floor, thanks mostly to the adoption of viewable impressions, but any modest CPM increases we do achieve through any of these innovations will be eradicated by our commitment to repeating prior mistakes that deflate them. 

My former boss Frank Smith taught me the valuable business lesson never to surface a problem without offering a solution.  So here are the mistakes I am referring to, and how we can fix them.

Mistake 1: 

We still have too many advertisers appearing at once on each page view, so clients don’t see a greater perceived value when looking at their own ads.  This causes them to look at performance reports instead to determine our value, and our eCPMs suffer.

Solution:

The IAB creates an “advertiser seal of approval” that can only be earned by publishers adhering to certain quality standards that make their sites advertiser-friendly.  In this case, publishers earning the seal must serve no more than three display ad messages per page view (this includes logos), and each page view is served on a 100% share of voice to one single advertiser. 

For those who see this solution as too dramatic, I argue this particular problem of advertisers not seeing value when they see their own advertising as a dramatic problem in need of a dramatic about-face solution.  I also ask you to visit Grantland.com, an example of a site that only serves one advertiser per page view.

Mistake 2:

We still sell ads we know annoy our site visitors.  We serve these UOAs (“units of annoyance”) because these ads artificially drive performance metrics.  These include pop-ups, which are back (did they ever really leave?), and any ads that trespass from their defined area to either cover up or push down the very content users are trying to view.  Interaction rates then accidentally increase when users furiously try to squelch these ads like mosquitoes, often missing the “x” mark we condescendingly offer as help. 

This particular mistake causes a complex problem. 

First, these ads displace responsibility for ad interaction rates from the actual creative messaging and places it on the physical “ad unit” itself.  Thirty-second spots on TV and PG4Cs in print don’t drive engagement -- the creative does. But in online we have these lines between ad units and creative crisscrossed. 

As a result, these rich-media ad units, sold initially at slightly higher CPMs based on their visual glory, not only hurt our relationship with our users, but further reinforce that our perceived value should be based on performance reports -- hence eCPMs’ decline. 

Solution:

This solution is a tough sell, given the IAB’s recent announcement that three of the five “rising star” ad units for video display will be “overlays” covering content users are trying to view, but here goes: under this “advertising seal of approval” program, publishers earning the seal will not serve any ad units that pop up (including survey pop-ups ironically telling users “your opinion matters”), or any ads that move beyond their defined areas, including all forms of “overlays” or “push downs” -- unless these ads are 100% user-initiated.

Now the creative message will have to shoulder the proper amount of responsibility to deliver interaction rates, and users won’t hate us. 

Mistake 3:

We are not helping our efforts to improve our medium’s perceived value by continuing to stuff low-rent advertisers into our glutton of inventory, and then convincing ourselves these low-quality ads don’t lower our perceived value.

Solution:

Running the ad networks out of town hasn’t worked, so how about publishers earning the IAB “advertising seal of approval” agreeing to only run these low-quality ads between midnight and 6 a.m., just like when television runs their infomercials?  Performance may (or may not) suffer for your lowest-paying advertisers while clients and buyers for premium advertisers who visit your site during the workday won’t see this:

The underlying and self-inflicted problem with online publishing and the lack of perceived value we deliver to advertis ers, is how poorly we treat the users who visit our sites.  We act entitled to their attention, certainly to their personal data, and we grab either any way we want with little regard to how it impacts how users feel.  This translates to less time and engagement on our sites -- proven by the high bounce rates that no one ever talks about -- and the entire ball of perceived value starts rolling in the wrong direction from there. 

In other words, we pretty much shit on users if it makes us more money from advertising, and that’s why our eCPMs are in the toilet.  Until we start treating site visitors like kings and queens instead of pawns, we will continue to be paupers in this game of selling ads.