Full Funnel and Outcomes Measurement with FouAnalytics

Humans convert; bots don't. But you all already knew that. So why are so many marketers still reporting on useless vanity metrics like impressions, traffic, and clicks? Well, those are easy to measure and, conveniently, they are large numbers so they look impressive when you show your boss. But as you saw from this previous newsletter -- Marketers, DON'T optimize your campaign for "performance" -- clicks are not always what they seem. This is especially true if you can't tell apart which clicks were from bots and which were from humans. Reprising the following chart, you can see just how few humans (dark blue) there were in examples A and B below. D had far more human clicks but even then it was only 28% of the total number of clicks. You can read the rest about verifying clicks in the previous article.

Full funnel measurement with FouAnalytics

In this post, I will talk about some examples of how advertisers have used FouAnalytics to measure the full funnel - from the ad impressions, to the arrivals on the landing pages, and finally to the conversion events.

IN-AD tags - These FouAnalytics tags go into the ad impressions themselves. This tells you where your ads ended up and whether a bot or human caused it to load. Using FouAnalytics provides an extra level of detail that placement reports don't give you. Even if you had asked for domain level placement reports in the past, fake sites pretending to be mainstream sites will be comingled in the data and you won't be able to pick apart fake marthastewart.com impressions from real ones, because the domain in the bid request is the same. With FouAnalytics measurement you will be able to see where your ads actually ended up -- e.g. on the fake site, even though the bid request said it was marthastewart.com. Furthermore, the fraud verification vendors you are paying for don't catch this. You don't have to believe me, I can show you if you do a side-by-side measurement test with FouAnalytics.

ON-SITE tags - These FouAnalytics tags go on the websites and landing pages to which the ads click. The purpose is to help you verify the quantity of the clicks arriving from paid channels and the quality of those clicks. Almost daily, I hear from advertisers that the clicks reported by their ad platforms far exceeds the clicks that they could see with their own site analytics, like Google Analytics or Adobe Analytics. There are two problems here which cannot be distinguished with current analytics -- either the clicks didn't arrive or they did arrive but were filtered out automatically by Google Analytics or Adobe Analytics. While filtering bot clicks is useful in some scenarios, in this case it creates discrepancies that are hard to explain. Using FouAnalytics on the landing pages will give you the details to troubleshoot this problem. FouAnalytics records all clicks, bot or human. It does not discard anything; this way you can see which bots came to your landing page, how much there was, and where they came from. Knowing this, you can better tune your paid media, upstream.

On-site tags are also useful for certain paid digital channels which do not allow tags in the ads themselves -- for example, search ads, YouTube ads, Facebook ads, Taboola, Outbrain, etc. don't allow tags in the ads themselves. FouAnalytics on the landing page let's you judge the relative quality of the clicks being delivered by each of those paid channels. For example, you can easily see the differences in quality of the clicks coming from 8 different paid sources. Dark red means bots/fraud and dark blue means humans. With just this information you can optimize your media by allocating more budget to paid channels sending more dark blue humans and reduce budget to those sending more dark red bot clicks. Simple, right? Right, if you had the right analytics in place ;-)

Humans convert, bots don't - in charts

Now that you have optimized your media using FouAnalytics in-ad tags (found bad sites and apps and added them to block lists) and optimized your budget allocation using FouAnalytics on-site tags (shifted spend to sources sending human clicks and away from sources sending bot clicks), you can take one final step and confirm which sources are converting most efficiently for you. To illustrate this, I will show you a series of charts from over the years, by clients in different industries and verticals.

In the first example above, campaign A starts with better media (more dark blue). This leads to more arrivals on the landing page. Note the relative quality - higher dark blue and lower dark red than in campaign B. Then note the conversions donut chart - it's mostly dark blue. This illustrates that humans convert (i.e. buy). Even in campaign B, only humans convert. But because ads were wasted being shown to bots, and fewer humans clicked through to the site, the number of conversions was far lower. In this example, you can see that starting with better media means your outcomes can be 30X higher. So even if your media were more expensive on a CPM basis, it is worth it. Using FouAnalytics you will have the data to make those determinations yourself.

In this next example (chart above), we compare the clicks arriving on the landing page from four different display ad sources. You can clearly see Display 1 and 2 send the largest numbers of clicks; and the dark blue tells you the portion of those clicks that can be confirmed as human. These humans went on to convert. The "human conversion rate" is calculated by taking the number of conversions, divided by the number of human clicks. Note that Display 1 converts at 40%, which is 8X higher than Display 2. Hint: Display 1 is Facebook display ads (with FAN turned off). When used properly, Facebook ads work well.

In the third example above, we track the number of calls made to special phone numbers to track call conversions. You can see the correlation between higher dark blue (humans) and more calls. You can also see the comparison between organic and paid channels. More humans means more conversions and higher rate of conversions.

Finally, the last example below shows another direct comparison between a paid media source and organic traffic, in terms of conversions. You can clearly see organic traffic converts at a 4X higher rate (66% versus 16%) than paid media sources. I've always said that if a human clicked on an organic search result and came to your site, they actually wanted to be there. So they are in the right mindset to convert. That is why I also said if you had a dollar to invest, invest it in creating content that attracts organic visitors rather than spend it on paid ads. This is one example that illustrates this and supports my recommendation.

That said, you can see that paid media drove a larger quantity of conversions -- 1,043 vs 222. So you should use paid digital media to scale and grow. For those who think I am against digital advertising, I am not. I am just against BAD digital marketing, like what is being done today. Most advertisers, especially the biggest brands, are doing it all wrong today, because "scale and grow" to them means buy as many ad impressions as possible on programmatic exchanges, for as low cost as possible. Those are mostly bot impressions. Furthermore, they haven't even deployed FouAnalytics to see if the ads were shown to humans or bots, so they are completely blind to whether it is working or not.

After a ten year roundabout in the wilderness, isn't it time to get back to real digital marketing and use real analytics to "see fou yourself" -- i.e. help yourself measure what is really working or not? If you want to use FouAnalytics, please message me.

More FouAnalytics Newsletter posts: https://www.dhirubhai.net/today/author/augustinefou