On July 27, the Wall Street Journal reported that P&G cut $100 million out of their second quarter digital ad spend with no adverse effect on their business.  They cut ads targeted by bots and placements that were inconsistent with brand equity.

On the one hand, as the biggest advertiser in the world, P&G’s views on ad spending are highly respected and carefully watched.  If they cut digital ad spend over concerns with bots and weak brand image alignment, then we need to pay attention.  On the other hand, a $100 million cut by P&G only represents 4% of their total US ad spend.  The digital ad cut may be part of a healthy “test and refine” cycle.

Still, P&G’s decision raises concerns for all advertisers about click fraud. Click fraud is a highly contentious and complex issue. You have:

• Advertisers who want to buy real clicks from publishers that support their brands, not detract from them.
• Publishers like Google and Facebook who have an oligopoly and are paid for clicks, not law enforcement.
• Ad agencies who are accountable for buying good clicks, but their fees are often tied to ad purchase volume.
• US government agencies with limited resources to focus on bots and click fraud.
• Foreign governments that struggle with arresting a citizen who may be committing a very difficult-to-prove crime in another part of the world.

In this piece, we first address the complaints that P&G and many others make about click fraud.  Then we explore how planning and oversight can be used to fight the problem.

Estimates Vary Wildly

A study commissioned by WPP published in March of 2017 estimates the cost of digital ad fraud could reach $16.5 billion in 2017.  This is more than double the estimated $7.2 billion estimate for 2016 made by Association of National Advertisers (ANA).

Another study released one month later by WhiteOps and the ANA estimates the 2017 cost at $6.4 billion, which is a 10 percent decline from 2016.  This estimate is more than 60 percent lower than the WPP estimate.

Wild variation in click fraud cost estimates does not inspire confidence, but we can conclude it is real, pervasive, and hard to measure.

It is Real and Pervasive

A survey shows that 78 percent of brand marketers are concerned with click fraud and bot traffic.  Further, 80 percent of brands in the WhiteOps/ANA study deploy some form of bot countermeasures.  Everyone participating in digital advertising should know that click fraud is a real and pervasive risk that must be guarded against.

It is Hard to Detect and Measure

It is very hard to measure digital ad fraud with precision.  Much of the problem is intended by the perpetrators; their survival depends on being sly.  For example, WhiteOps discovered and then reported that the Russian scheme Methbot went undetected because several internet registries were hacked to make its fraudulent sites seem legitimate.  Bot-makers are constantly adapting and inventing new ways to avoid detection.

Fraud measurement is further complicated by the fact that there are two general types of digital ad charges: “view-based” and “click-based” fraud.  Digital ad fraud is not just about measuring the value of a click, but the value of a view-through as well.  Assessing view-through value requires rigorous A/B testing, or the use of large and costly control campaigns.  One way to eliminate the latter complexity is to only pay for clicks.  Then fraud detection can focus on which clicks convert into actual views by real humans on buyer journeys.

Weak Digital Strategy

Too often, the concern about click fraud is a smoke screen for a weak digital media strategy.  When clicks do not generate sales, a convenient excuse is to blame click fraud.

The perfect digital ad strategy is based on:

1. Knowing exactly what future customers want to read, see or hear.
2. Crafting tactics that provide it exactly when and where they want to receive it.

It is impossible to create a perfect digital strategy, but the more you approach perfection, the lower your exposure to click fraud.  Risks escalate as you move toward a shotgun digital ad strategy.  It exposes you to a wild ocean of unseen activity, including those with malicious intent.

Segmenting and targeting are critical to a good strategy, as are understanding the buyer’s journey and the various personas they may assume along the way.  The first line of defense against click fraud is a well-crafted digital advertising strategy.

Connect Clicks to Results

A good digital strategy defines success measures, sets performance targets, and specifies how the measures will be tracked.  Since it is so easy to start a digital ad campaign, it is very tempting to skip these critical steps.  By taking the time to specify them up front, you lay the foundation for measuring which clicks supported your campaign objectives, and which ones did not.  In other words, a click is not a click.  Some are well-aligned with your digital strategy, some produce actual results, and the rest are suspect and may include click fraud.

For every click, you want information about whether it was in your target or not.  Just knowing you have a click on your ad is woefully inadequate in the wild world of digital.  You need information about who clicked on your ad.  And then, when they went to your site, did they visit more pages, spend more time per page, subscribe to your mailing list, download coupons, click on a map, watch a video, or make more purchases than the average site visitor?

If you are not trying to detect and measure these things, you’re not really implementing a digital strategy, which is all about somehow converting anonymous clicks into useful intelligence about what you just paid for.  Multiple martech disciplines are required to make this happen.  In retail, maybe beacons or point-of-sale (POS) information are required.  Maybe it means detecting the clicker’s URL and attempting to tie it to a target market, past customer or past site visitor.  And then, you can repeat these steps for each ad with each digital publisher or referral source.

If you don’t have these capabilities or access to them, your exposure to click fraud escalates.  As it turns out, much of the effort involved in protecting yourself from click fraud is really nothing more than doing a great job of establishing your digital marketing strategy and carefully tracking its success.  The closer you get to tracking a click and connecting it to your strategy, the harder it is for bots to continue their charade and trick you into thinking they represent something legitimate.

Weak Alternatives

What many companies are starting to learn the hard way is that top-of-funnel metrics like “reach”, “lift”, “frequency”, and “awareness” are poor substitutes for measures that look deeper into the sales funnel, or align more closely with real buyer’s journeys.  Offline measures from surveys and panels are often disconnected from active customers.

While online analytics is still an imperfect science, there is a lot to be learned from the highly granular and in-the-moment world of digital advertising.  So, our advice is to focus on what you can do about click fraud.  Use it as a motivator—a silver lining—and take your digital strategy to the next level.

David Rodnitzky is founder and CEO of 3Q Digital, a position he has held since the Company’s inception in 2008. Prior to 3Q Digital, he held senior marketing roles at several Internet companies, including Rentals.com (2000-2001), FindLaw (2001-2004), Adteractive (2004-2006), and Mercantila (2007-2008). David currently serves on advisory boards for several companies, including Marin Software, MediaBoost, Mediacause, and a stealth travel start-up.

Frank Grillo is the CMO of Harte Hanks. He is a passionate advocate for bringing the human back into marketing by better understanding and facilitating customer journeys.