Facebook’s eCommerce Division Faces Exec Departures, Frustrated Retailers

Facebook, lawsuit, FCA, Meta

Meta Platforms, the parent company of Facebook, isn’t making as much progress on eCommerce as employees thought it would at the time when the pandemic hit, The Wall Street Journal reported Tuesday (May 3), and some retailers aren’t happy with that.

Retailers have been “frustrated” by this, saying the service doesn’t have some basic elements like the ability to display products in different modes if they’re not sold directly through Facebook and Instagram.

At least five senior executives have left in the last six months, an anonymous source told WSJ.

But Meta is confident in the project.

“Building a full-fledged commerce platform is a multiyear journey — one that became a company priority less than two years ago,” said Joe Osborne, a spokesperson for Meta. “We’re proud of our progress and in the teams working to build out these experiences.”

Meta has reportedly been adding to its efforts in eCommerce early in the pandemic as it debuted features like Facebook and Instagram Shops. The shops allowed brands to post their catalogs directly on the apps.

In addition, Meta had begun preparing for various privacy changes Apple was looking into putting in place, with changes letting users opt out of having activity on their devices tracked by companies like Meta, which was crucial to how Meta looked into the efficacy of their ads.

See also: Facebook Owner Meta Gives AI Community Access to Large Language Model

PYMNTS wrote that Meta has added more access to its large language model, including the AI community, to add in research.

The model is the first 175-billion-parameter language model to be available for the AI research community.

The “large language models” refer to natural language processing systems trained on large amounts of text, which can answer reading comprehension questions or make new text.

Called the Open Pretrained Transformer, Meta said this would boost the abilities of research to understand how large language models work.