Meta Tests Takes Prediction Market App With Employees

By Daniel Okafor

Meta Platforms, the parent company of Facebook and Instagram, has begun an internal employee test of Takes, a prediction market app that uses virtual points called Cred rather than real cash, according to The Block’s November 13 report.

The trial moves the project beyond the directive stage announced in June, when Meta CEO Mark Zuckerberg instructed employees to develop a prediction market app. At that point, the internal project carried the name Arena.

Takes now lets employees stake Cred on the outcome of a specific event. A correct prediction earns points; an incorrect one costs the user those points. The report gives no number for the initial point balance, event categories or employee participation.

Meta Platforms has not disclosed a launch date for Takes. The company also has not said whether it will add a real-money betting function.

What does the points-only model change?

The points-only design separates Meta’s employee test from real-cash prediction markets such as Polymarket and Kalshi. Users stake Cred on the outcome of a specific event, earning points if their prediction is correct and losing those points if it is wrong.

That structure allows Meta to test a market interface without introducing cash-backed stakes during the trial; it does not establish whether users will place real funds in a later version. Virtual points also do not expose users to the same financial consequences as cash.

The report identifies no employee count, point allocation, event list, trading volume or retention data. Its description of the mechanics stops at the basic rule: users stake Cred on an event and gain or lose points according to the result.

Those omissions prevent a conclusion on adoption or product fit. The trial can show whether employees understand the app’s basic proposition but cannot by itself establish demand for a consumer service or willingness to risk real money.

The source also gives no indication whether Meta will preserve Cred, convert it into another reward or remove it in a later version. Nor does it identify a regulatory strategy for real-money betting. Those decisions remain unresolved rather than announced.

What happens next?

In June, Zuckerberg’s instruction gave employees a specific product mandate: build a prediction market app. The subsequent use of the name Takes differs from the internal Arena label, although The Block’s report does not explain when or why the name changed.

Meta has not said when the employee trial will expand beyond staff, whether outside users will receive access or whether a public release will carry Cred. A launch without real cash would position the app as a forecasting or entertainment product rather than a cash-backed wagering venue.

A real-money version would raise separate operating questions about cash handling, eligible events, user verification and the legal treatment of prediction markets in the jurisdictions where Meta might offer it. The source names no jurisdiction, regulator, filing or timetable for such a change.

The most important unresolved commercial question is whether employees use points merely for an internal experiment or continue trading once Meta opens the product. The report provides no adoption data, so there is no basis for assessing demand beyond the fact of the trial.

Meta’s confirmed steps remain limited: Zuckerberg issued the June directive, the internal project first carried the Arena name, and Meta now tests Takes with employees using Cred. The launch date and any move toward real-money betting remain unannounced.

That leaves Meta with a functioning internal experiment but no public deadline. The next structural checkpoint will be a disclosed expansion in access, a stated launch schedule or a decision on cash-backed trading; without one of those, Meta has no public timeline and the employee test remains an internal points experiment.