Dive Brief:

• The algorithm for Facebook's ad platform is serving up bizarre and sometimes inappropriate ads from a company called Wish, according to a report from The Daily Beast. Wish is an e-commerce business akin to Amazon or Alibaba that keeps its products appealingly cheap by shipping directly from Chinese manufacturers.
• Wish is also a heavy spender on Facebook and quick adopter of the site's new marketing tools, including dynamic ads, where businesses can upload their entire catalog with specific products or carousels of products being served to users that match the right data profile. The issue is that Wish sells a lot of things, many of which could be construed as offensive such as sexually-suggestive items and clothing referencing illicit drugs. Inventory for these ads has started to appear on what are otherwise brand-safe apps like Timehop, where it is frequently flagged by users, per The Daily Beast. 

• Part of the issue stems from how some of Wish's more unusual products, such as plastic nostril holders and cat blindfolds, garner high clickthrough rates based on curiosity and shock value alone, boosting their visibility, according to The Daily Beast. This has mushroomed to the point where there is an entire online subculture dissecting ads served by Wish, including through a Twitter account called Weird Wish Ads.

Dive Insight:

Facebook's Wish conundrum points to how out of control some algorithm-based models have become and how relying on automated technologies for activities like targeting ads can ultimately hamper brand safety. Even so, algorithms are increasingly viewed as a necessity when marketing on digital channels that reach audiences that would be practically impossible to sort through manually. 

If companies like Facebook can't sort out these problems and more carefully manage how their massive stores of data are leveraged, they risk not only further damaging trust with advertisers but also inviting heavier regulation from the government and larger outcry from consumer protection advocates. This has already been apparent with revelations around the prevalence of fake news, much of it funded by Russian operatives, that was spread on Facebook and other social media platforms to influence last year's presidential race.

Some of Facebook's competitors recognize these weaknesses and are attempting to offer sturdier solutions for marketers. Snap Inc. recently redesigned its video messaging app Snapchat to steer away from algorithms and vet more of its content in-house. Both Facebook and Google are also hiring more human staff to review content, but if cases like Wish continue to appear, it might signal a more fundamental flaw in their structures that needs to be addressed.