‘A crack in the wall’: Meta accused of inflating Shops ad performance

Meta has been accused of inflating the return on ad spend (ROAS) for ads running on its Shops network, in serious allegations from a former product manager of the Shops ads team.

The claims, first reported in Adweek, were made by Samujjal Purkayastha this week in a London employment tribunal after he was laid off by Meta earlier in the year.

Meta has rejected the allegations outright, saying “we have full confidence in our performance review processes.”

On Linkedin, Purkayastha lists his former role at Meta as “Senior Staff Product Manager”, which he held for five years from 2020. He is now working for Amazon’s Prime Video, also based in London.

In the tribunal complaint, Purkayastha reportedly says that a Meta internal review found the ROAS on Shops ads had been inflated by 17-19% through the inclusion of shipping and tax in the final revenue number.

This is reportedly not standard practice in Meta’s other ROAS calculations, nor in the ROAS numbers of competitive products.

Shops in Instagram and Facebook keep customers within the app, rather than pushing them to external merchant sites. The implication is that Meta is incentivised to show Shops ads as outperforming other kinds of advertising because it owns those ecosystems.

The Shops product was launched on Facebook and Instagram in 2020, expanding through the period that Purkayastha claims the inflated numbers were provided to merchants. ROAS is a simple metric that divides sales by the cost of a campaign.

Example of a Meta Shop interface

Global media lead at TrinityP3 Stephen Wright said the allegations have serious implications for marketers in Australia.

“It’s one of the first things to publicly come out of the walled gardens that casts doubt on the data. It’s one of the first cracks in the wall, if you like. 

“If the alleged complaint has substance it shows a willingness to increase yield and profit through the manipulation of data in a questionable manner.”

The fact that digital platforms companies like Meta control all data gathering and reporting was a threat to advertisers, Wright said. In a written statement to Mumbrella, he described the alleged inflation as a “crude tactic” compared to new ways of making performance appear better.

“Of increasing concern for advertisers is the danger of sophisticated AI tools providing an arsenal of high tech weapons deployed in secrecy. The legal principle  ‘Caveat Emptor’ or ‘Buyer Beware’ has never been more relevant.”

In a written response to Mumbrella’s inquiries about the significance for Australian marketers, Louder founder Andrew Hughes described how the Shop ads format might have automatically included shipping and tax in the purchase value.

“It’s important to clarify that the alleged ROAS inflation only applies to Meta’s Shop Ads, which are fundamentally different from standard Meta ad formats. These ads run through Meta’s native checkout experience and require a direct integration with platforms like Shopify, so the way purchase values are tracked may not be something advertisers can directly configure.

“From a technical standpoint, ROAS is calculated using purchase value data sent via pixel or API. But if Shop Ads automatically pass tax and shipping through from the commerce platform, then it’s difficult to argue this is simply user error. Including those values in ROAS would inflate performance by definition, and the decision to do so sits more squarely with Meta.

“The broader point is that not all Meta ad products operate the same way. Grouping them together risks oversimplifying the issue. For marketers, this highlights why it’s important to understand exactly how performance data is being generated, especially when it’s tied to platform-owned commerce flows.”

In Purkayastha’s tribunal claim, cited in the Adweek article, the former product manager says the ROAS metric was part of an internal Meta performance target.

Samujjal Purkayastha (Linkedin)

“In addition to the ROAS performance metric being overstated by nearly a fifth, it meant that, rather than having exceeded our primary target, the Shops Ads team had in fact missed it once the figure was reduced to take account of the artificial inflation.”

Purkayastha says he was concerned about the ROAS inflation and brought it to the notice of his bosses at Meta several times, implying this was the reason for his dismissal. The tribunal judge did not agree out of hand, denying his application to immediately be reinstated to his job, and the case is proceeding.

Mumbrella contacted Meta in Australia for more detail and background on the case. It released the following statement:

“We are actively defending these proceedings and note today’s decision by the Employment Tribunal rejecting the Claimant’s application. Allegations related to the integrity of our advertising practices are without merit and we have full confidence in our performance review processes.”