By Michael Iantosca, Chief Revenue Officer, Integral Ad Science

“I can’t be bothered.”

That’s often what I hear from agency execs when we discuss selecting a viewability measurement vendor. To be honest, it’s not difficult to understand why. After all, choosing the right viewability measurement partner is hard work, a task that requires marketers to spend hours researching and interviewing vendors to find the best fit. And why bother investing all of that time and energy in the selection process when you already know that four of the six media partners on your biggest campaign use the same vendor, and none of them are going to budge without a substantial effort on your part?

Unfortunately, the wrong measurement partner can be a costly decision requiring agencies to pay potentially hundreds of thousands of dollars for impressions that another vendor would have determined to be outside the user’s view. While it might be slightly inconvenient for agencies to play an active role in choosing a vendor for their next viewability campaign, it’s a big mistake for them not to.

If you were buying a house, you wouldn’t let the seller select the inspector, and when doing your taxes, you don’t give the IRS free rein to pick your accountant. So, why allow a party with different interests than yours dictate the terms of your ad buy?

As an example of how drastically your outcomes can differ based on your vendor choice, consider first that not all measurement rates are calculated equally. There are many reasons a vendor might not be able to measure the viewability of an impression. For instance, some technologies have difficulty determining the viewability of custom executions, others are hit-or-miss on native ads, and new types of ads pop up all the time that can create challenges. But some vendors subtract “unmeasurable” from the total, shrinking the denominator and implying a near-perfect measurement rate. By asking vendors to provide how they calculate measured rate, you can get a better sense of just how much of your campaign they will be able to verify as viewable.

Of course, it’s important to follow up your question of how much vendors measure with one asking them how well they measure. One major discrepancy between vendors is the point at which they start counting the one second that at least half of an ad needs to be in view in order for it to count as viewable. Many vendors favored by publishers embed their technology within the ad server, starting the clock once the ad container is loaded. However, as you’ve probably noticed, the actual ad creative is usually the last element to appear on the page. This means that unless you opt for a vendor that starts its countdown once the creative appears, you risk paying for impressions where all a consumer ever sees is the empty box your ad is supposed to be in.

Last, but not least, you’ll need to make sure that you and your publisher partners settle on a vendor not only capable of detecting fraudulent activity but invests heavily in building strong, cutting-edge solutions that are capable of marking all impressions served to bots as non-viewable inventory. Even an ad placed smack-dab in the middle of the page is completely worthless if there isn’t a real human being on the other end to see it. Bot programmers are smart and go out of their way to make ads served to non-human traffic appear viewable. You would be surprised how frequently they are able to fool vendors who don’t make fraud a priority. In fact, just last week, we were told that our in-view rate was 50% lower than a rival viewability vendor. When we looked at the numbers, we found that the other vendor did not detect that over 55% of the total impressions served on the campaign were shown to bots. They only detected 5%, a huge miss for them and our mutual client.

In a way, having a vendor that doesn’t have robust fraud detection that discards all fraudulent impressions defeats the entire purpose of transacting on viewability, which is that you only have to pay for the ads people are able to see. For instance, if you bought 1 million impressions at a $1 CPM and your vendor measured the impressions 70 percent viewable, you’ve saved $300 you would have wasted on non-viewable ads. But if half of those “viewable” ads were fraudulent, $350 of your $700 total spend is still going to a bunch of bots, meaning you wasted more money than you saved by transacting on viewability in the first place.

For brands to receive real value from viewable deals, agencies need to make a point of vetting their prospective measurement vendors. While publishers are by no means out to get them, it stands to reason that absent agency input, ad sellers will act in their own best interests and select measurement technologies that are more likely than not to mark their inventory as viewable. Marketers who cannot be bothered to engage in this process risk paying a premium for “viewable” impressions that nobody actually sees.

The Cynsiders column is a platform for industry leaders to reach out to colleagues, followers, and the public at large. In their own words and in targeted Q&As, columnists address breaking news, issues of the day, and the larger changes going on in the ever-evolving world of television, video and digital. Cynsiders columns live on Cynopsis’ main page and are promoted across all daily newsletters. We welcome readers’ comments, queries, and column ideas at [email protected].