blog: Don Marti


Behavioral targeting and investor relations

19 January 2019

Eric Shih writes, on the Digital Content Next site,,

Leveraging data to make more informed ad targeting decisions is a breakthrough versus previous methods where ads were un-targeted. Personalized ads are a win for all parties. [emphasis added] It is better for:
  • Users (connects them to more interesting and relevant ads)
  • Advertisers (results in higher return on investment)
  • Publishers (delivers higher CPMs and increased revenue)

Hold on a minute. That doesn't look right.

The win for all parties is clearly bogus from the user and publisher point of view. Behavioral targeting means that users see more low-quality and deceptive ads, because behavioral targeting gives a long-running structural advantage to deceptive advertisers. As a potential customer, I also win as ad budgets go into supporting context that I care about, such as news and cultural works, and lose as ad budgets go into behavioral targeting with negative externalities.

From the publisher point of view, behavioral targeting creates near-infinite saleable ad inventory on low-value and fraudulent sites, and forces publishers to contend with those sites for ad money. For users and publishers, the less behavioral targeting the better. But what about for marketers? Isn't behavioral targeting a win for them?

It looks like the answer is: not so much. Here's a must-read piece on research by Prof. Byron Sharp.

Personalised advertising may be one thing but getting people to respond to even micro-targeted ads is a whole other ball game. However, analysis of 3.1 million ad exposures shows that such adverts generate low click through rates (CTR). Furthermore, some of the responses to such ads are counterintuitive – with a higher CTR coming from ads mismatched to personalities and lower vs. the overall industry average for Facebook ads.


But if micro-targeting is arguably so ineffective, why do some many marketers use it? Sharp and Danenberg highlight several reasons:
  • Marketers often do things based on theory/logic rather than evidence. The worst myths, the longest lasting, are those that sound plausible.
  • Micro-targeted campaigns can boast of high ROI, largely because they are so small, reaching people who had a high likelihood of buying anyway. Marketers see the high campaign ROI from micro-targeting but fail to realise that the overall return to the company may well be lower.
  • It’s fashionable!!!

Read the whole thing. Not surprising if you follow the underlying #behavioralEconomics. And Dave Trott agress with the fashionable point: It's right because everyone's doing it.

There might be another reason, though. Maybe part of the problem is that marketing science is hopelessly mixed up with investor relations. After all, adtech firms and ad agency groups are publicly traded companies. And creative ad work, hiring writers and artists and marking up what you pay them, is the kind of business model that stock markets get bored with. Margins are low, you can only grow as fast as you can hire, and your assets can quit and go work somewhere else.

Choosing to place an ad in a quality context is more cost-effective, but again, it doesn't scale. If your business is putting good ads in good places then the people who do good work have market power. But psychographic models and the underlying data sets are more investor-friendly. Even if it takes torturing the data and putting up with fraud to make adtech look effective. Mathemagickal woo-woo is scalable, more like the intangible assets of a software company. Markets see promise of big margins and high revenue/employee.

How much do investor-focused messages about the effectiveness of behavioral targeting companies interfere with marketer-focused messages about the effectiveness of behavorial targeting in campaigns?

The case against behavioral advertising is stacking up | TechCrunch

After GDPR, The New York Times cut off ad exchanges in Europe — and kept growing ad revenue

Why Online Ads Haven't Built Brands

‘GDPR will ultimately be good for the industry’: Guardian CRO Hamish Nicklin on 2019 plans

Google and publishers edge closer to reaching common ground for GDPR standards