Retargeting? Just stop it.

Retargeting? Just stop it.

Large advertisers bought into the concept of retargeting because on the surface, showing ads to users that had previously visited the advertisers' sites kinda makes sense. In theory, if someone visited your site, they "expressed interest" in your brand, right? Yeah, kinda makes sense.

But while that makes sense in theory, in the real world, fraudsters have taken advantage of this by first having their bots visit advertisers' websites and THEN go to cash-out sites to cause ads to load, so they can earn the higher retargeting CPMs. Read this one more time. Bots go to advertisers' sites first, to collect cookies; the ad networks then assume these users "expressed interest" and re-target them with ads, thinking it will drive better outcomes for advertisers. Look at the "retargeting" and "remarketing" rows in the data grids below. They have a lot more dark red than the other sources of paid traffic.


No better outcomes, if you're retargeting bots, duh.

How do we know there will be no better outcomes? If you're retargeting bots and not humans, you won't get any more outcomes. Look at the following data from FouAnalytics, measured with in-ad tags applied specifically to a retargeting campaign. The color coding shows 11% dark blue (humans) -- about 1 in 10 retargeted users were humans. The rest is either confirmed bots (dark red) or likely bots (light red). So you are showing ads to bots that visited your site first. That's who you're retargeting. You're retargeting lots of bots and only a few humans.

But don't take my word for it, you can also see this retargeting problem in the data reported by the DSP. FouAnalytics collects macros from the DSP as well, and the data grid below shows "custom_user" which is the identifier that the DSP assigns to the unique user. Notice anything wrong with this? You can ignore the color coding for now, just look at the counts. 835 ads were shown to the first user, 770 to the next, 587 to the third and so forth. Do you think you should be showing this many ads to the same user in your retargeting campaign? Do you think that user would be more likely to buy from you, or be more likely to be pissed off at you after seeing 770 ads over and over again? And this is the DSP's data, not FouAnalytics data which corroborates that you are showing far too many ads to the same user (bots in this case).

Well, you don't have to worry so much because most of these users were bots, and bots don't get pissed at you. They LOVE when you show them more ads because they make more money, especially at the higher CPMs you paid for retargeting. ;-) Most of these retargeted users were bots. That's why the title of this article is "Retargeting? Stop It!"

How FEW of these users in the data grid above are marked blue? Very few. But did you notice that the custom_users that were blue were shown hundreds of ads? Yeah, these humans are now pissed at you for following them around the internet with your ad. You may have experienced this yourself too. For example, after you look at a product on Amazon, that exact product shows up in display ads no matter what site you visit. You think that's creepy AND you think that's a dumb waste of money because you may have already purchased the item. Yeah, that's retargeting -- the bots love that you retarget because they make more money; the humans hate that you retarget them because they've already purchased the item. Either way, you're wasting most of your budget.


What to do if you insist on doing retargeting campaigns

But if you insist on doing retargeting campaigns, at the very least use FouAnalytics in-ad tags to measure your ads. This way, you can see where they are going and identify the cash-out sites that are using bots to steal your retargeting budgets. In the data below, we can see different "sources" -- the sites where the retargeted ads went. Find the ones that are the highest in red and dark red, and add those to your block list. Leave the ones that have some blue or dark blue (because those are the sites where retargeted human users go). You don't need to add those to your block list. By doing this analysis with the data from FouAnalytics in-ad measurement, you can make your retargeting campaigns more effective.

Want to have a go at this? Reach out to me and I will set up a FouAnalytics account for you.


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

David Culbertson

Agile SEO | Google: Ads + Analytics + Search Console + Data Studio + Tag Manager | Conversion + Ecommerce Optimization | Shopify | WooCommerce | WordPress | Google Business Profiles | Piwik Pro

7 个月

Counterpoint: At least for the Google Ads network, stay of top of the site/app exclusion list. Even better, build a white list of sites/apps for retargeting.

Malcolm Auld

Marketer, advertiser, educator, author, commentator, keynote speaker, Host of The BIG Marketing Show - You get better results, or else...

7 个月

Yes Dr. Augustine Fou How bad are the burn marks left on prospects by inept retargeting? Nobody is tracking the negative effect on prospects, just the CTR on the ad: https://www.dhirubhai.net/pulse/how-surveillance-marketing-damaging-brands-reputation-malcolm-auld/

回复
Dhiraj Gupta

Anti-Ad Fraud Expert | Co-Founder of mFilterIt | Building Technologies For Brands To Save Their AdSpends & Protect Brand Reputation

7 个月

Infact avoid all performance ?? rogrammstic retargeting channels !! They only spam to high jack organic traffic

Craig McDaniel

President at Sweepstakes Today LLC

7 个月

Many times, I have seen the bad guys simple change the domain name or URL. There are a number of variations to the game the bad guys play, but any suggestions how to stop them?

回复

Classic marketing school graph... effective reach & negative exposure:

  • 该图片无替代文字

要查看或添加评论,请登录

Dr. Augustine Fou的更多文章

社区洞察

其他会员也浏览了