Can Sales Lift Measurement Reduce Ad Fraud Risk?

Marketers debate how precisely brands should target their digital ads. Meanwhile, numerous studies have proven the value and efficiency of being able to target people based on very specific data — such as the brand of soap they buy or the age and make of the car they drive — showing it to be more effective than general demographic targeting. However, ad fraud can undermine the surgical-strike efficiencies of well-targeted digital advertising without the proper measures in place to uncover and mitigate fraud risk.

Conservative estimates show ad fraud is more than twice as costly as tax refund fraud, costing advertisers in excess of $8 billion per year, making it the number-one cybercrime today. We live in a world where, according to Bloomberg, “Fake traffic has become a commodity. There’s malware for generating it and brokers who sell it.”

Ad fraud – the pay-per-click advertising crime of using a person, automated script or computer program to imitate legitimate web browser ad-clicking to generate revenue – is well-established in the world of desktop browsing, but is gaining a foothold on mobile browsers as well, and can occur with in-app advertising.

So, what can be done about it? What can a brand advertiser do to reduce the risk of paying for ad fraud or ads otherwise not viewed by a human? Unlike solutions that focus on the symptom and not the real problem, sales lift measurement holds surprising potential to help marketers manage fraud risk .

How Measuring Sales Lift Reduces Fraud Risk

As I tackled in my recent article in Martech Advisor and we discussed on page 10 of our recently released 2017 Benchmarks Report, the best way to manage fraud risk is for a brand to:
  • Insist on closed-loop sales lift measurement (measurement of an online ad’s direct effect on off-line sales, using in-store sales data) to define a campaign’s success.
  • Make sure the closed-loop measurement methodology considers both sales lift and media cost.
The latter ensures that you are comparing apples to apples when buying media on a CPM basis across multiple platforms, and helps advertisers identify fraudulent activity and other unusable impressions.

This is one of the key benefits of using in-store sales data to calculate campaign success by a precise ROAS – Return on Ad Spend.

By calculating ROAS for each platform that your brand places media with, you can then objectively compare performance across platforms. Then, for example, if a platform has an unusually low ROAS compared to others running the same campaign and creative, it signals possible fraud: wasted impressions.

Wasted impressions occur for many reasons, not just fraud. They can also happen if targeting is poor or inaccurate. And, yes, they can occur when a substantial percentage of impressions weren't viewed by a human with the potential to purchase, either because of delivery issues or fraud, either of which inflate media cost relative to the incremental sales generated.

With ROAS measurement, you have a comprehensive and consistent metric you can apply across all platforms. ROAS puts the power in the hands of the advertiser to identify easily if any platforms have targeting or delivery issues.

What I love about this approach is that it changes the focus from the symptom — fraudulent ad impressions — to the real problem: that bots don’t ever go in a store and buy your product.  Advertising to them is a waste of your money. 

While you can insist ad platforms implement fraud protection measures, the truth is it’s a bit like playing Whack-a-Mole®.  Just like with spam and identity theft, bad actors will keep finding new ways to rip you off faster than fraud companies can put measure in place to stop them.

Sales lift measurement won’t prevent ad fraud. But implementing it across all of the publishers you run ads with, performed by independent measurement companies like Nielsen Catalina Solutions or Kantar Worldpanel Shopcom, empowers you to mitigate fraud risk and shut it down before it costs you big time.