Do you still need FouAnalytics if ...

If you have already turned off Facebook Audience Network ("FAN") in your Facebook campaigns, turned off Google Search Partners ("GSP") in your Google search campaigns, turned off Google Video Partners ("GVP") in your YouTube video campaigns, turned off LinkedIn Audience Network in your LinkedIn campaigns, and turned off Pangle (TikTok's audience network), you would have avoided 90 - 99% of the obvious ad fraud already. If you've moved to an inclusion list (instead of a block list) you would have also saved a ton of work (you no longer have to add bad sites and apps to blocklists every week. That doesn't work well anyway because there's a limitless supply of bad sites and apps that bad guys use to steal your money.) The charts for campaigns A, B, and C below show examples of campaigns measured by FouAnalytics in-ad tags that are already well-optimized away from fraud and bots (very low dark red) and well-optimized towards humans (very high dark blue).

So the question is, if you've done all of the above already, do you still need FouAnalytics? Probably not. But would you like to know more about the humans that did click on your ads and arrived on your landing pages? Were those users attentive (they went on and did something on your site)? And did these users perform a conversion event -- e.g. buy something from an ecommerce site, do a zipcode search for a local auto dealership, etc? If you are interested, read on.

Attentiveness on advertisers' landing pages

Advanced advertisers continue to use FouAnalytics in-ad tags to measure their ads in order to ensure the ads are still going to quality publishers and shown to humans. Having FouAnalytics in place means they can immediately see any issues that arise and take corrective action right away. But for the most part, their campaigns are running in a good "steady state" of low fraud and high humans because of how strict their media buying controls are already. These same advanced advertisers also use FouAnalytics on-site tags on their landing pages so they can 1) verify the number of clicks arriving from paid media, 2) check that the clicks are from humans, and 3) see whether these humans were "attentive" on the landing pages. The slide below shows "human + attentive." The left side shows that most of the clicks that arrived are dark blue (humans) and the right side shows 75% of the users had touch events and 60 - 65% of the users clicked something else on the page -- i.e. they were attentive on the landing pages. Attentive is the opposite of "high bounce" and "low time-on-site."

Obviously, humans need to navigate around the site in order to "convert" (buy something on ecommerce sites or do zip code searches for services in their local area) or look up nutrition information for products they will eventually buy from offline grocery stores. Attentiveness on the site is a step towards conversions, both online and offline. Greater attentiveness also gives you clues about the effectiveness of the different digital channels you use -- for example paid search versus display versus paid social, etc. Attentiveness is estimated from human interaction events such as clicks, mouse moves, page scrolling, and touch events.

The table above shows 5 different advertisers and the attentiveness of the users coming from various paid media channels, compared to "direct traffic" (users who visited the site directly, yellow highlight). We also show the percentage of mobile users (mobile:1) and the percentages of bots and humans. The column of percentages on the left shows the ratio of traffic from the various sources. In all cases, you will notice that direct traffic is the largest source, by far -- 87%, 51%, 98%, 71%, 89%, respectively. This is a good sign. It tells me these advertisers are not over-spending on paid media. In contrast, the bad habit of "desperately buying traffic for the site" is what directly contributes to high bot traffic (high bounce rate, low time on site) that is entirely useless because bots don't convert. By looking at the RELATIVE attentiveness of users from various paid channels, the advertiser can further optimize their campaigns by allocating more budget to digital channels and tactics that drive more attentive humans to the site. Simple, right?

Cookieless and clickless conversion attribution

Once you know your ads are optimized away from fraud and towards humans, and you know the humans that clicked through to your site are attentive, what else can you do with FouAnalytics? You can analyze conversions and the rate of conversions ("conversion velocity") without cookies and even without clicks. The following example comes from an Australian agency that uses FouAnalytics to measure and optimize all campaigns for all clients. They also use FouAnalytics on the landing pages to measure conversions even for offline media such as TV ads and billboards. The data below comes from 3 types of clients -- automotive, financial services, and tourism.

Obviously there's no cookie-based targeting or attribution with TV ads and billboards. There's also no click throughs from the ads to the site (if you're getting click throughs from CTV ads, you need to seriously question the vendor). So how do we do attribution and assign credit for conversions to these forms of media without cookies or clicks? Simple, with time and location. This agency runs TV ads in certain cities and not in others. We can see that the "conversion velocity" (numbers of conversions per users arriving directly on the site) is higher in the cities where ads are in-flight, compared to those cities where ads are not running. This obviously requires a level of discipline and sophistication that does not exist when advertisers are over-spending in every channel, every market, at all times (ads are always running everywhere in all channels).

Side-bar: To be clear, DON'T follow P&G's example, especially in digital channels. Everyone praises P&G for their advertising savviness; but that's only because these vendors have something to sell, and they think flattery will get them some of P&G's gigantic budgets. P&G can cut 90% of their digital budgets and see BETTER outcomes than they do now. Their current levels of spending in digital channels are single-handedly propping up criminal enterprises worldwide and all of the disinformation purveyors, and all the crappy brand safety vendors that sell them reports showing their ads are not going to such criminals and disinformation outlets. If P&G came to me right now to ask to use FouAnalytics, I would decline them as a client. 'Nuf said. You can do (significantly) better than P&G in all advertising channels, not just digital.

So what?

The most advanced advertisers and their agencies use FouAnalytics in-ad tags to monitor and optimize their digital ads where possible. They also have taken actions to avoid 99% of the fraud to begin with -- e.g. by turning off audience networks and by using very short inclusion lists of real publishers. They also use FouAnalytics on-site tags on the landing pages so they can double verify that the clicks are human and the humans are attentive. Greater attentiveness leads to greater conversion velocity. They can even optimize their budget allocation, to spend more on digital channels that drive more attentive humans and greater conversion velocity. These advanced advertisers can even do attribution without cookies or clicks. But do you know what is most satisfying to them?

What is most satisfying to these advanced advertisers is KNOWING -- and having analytical PROOF -- that ad fraud actually IS someone else's problem. ;-)

Further reading: https://www.dhirubhai.net/in/augustinefou/recent-activity/newsletter/