For as much press as workplace flexibility has received in recent years, infamous cases of flexibility clawback have also made headlines. We’ve all read about how companies like Yahoo, Aetna, Bank of America, and IBM, just to name a few, have restricted or revoked their remote work programs, much to the astonishment of observers accustomed to hearing only success stories about flexibility. The hype around these companies' so-called flex failures is proof of something we’ve coined, the “flexibility glass cliff.”

You may already be familiar with the term glass cliff, which is typically applied to female leaders who break through the glass ceiling only to find that their lofty new positions were doomed from the start. These women are then blamed for their organization's failings, and this unjust blame is exacerbated by sexism as critics spin their misfortunes as evidence that women "just can't hack it."

Similarly with flexibility, the glass cliff comes into play when doomed-from-the-start flex initiatives are regarded as failures and then swiftly revoked because they “just didn’t work.” The companies mentioned above including Yahoo and IBM are among the big-name brands that have required their remote employees to return to the office in recent years, reasoning that keeping workers in close physical proximity "fosters teamwork, idea-sharing, and quicker decision making" and is "the best way to build a strong culture, increase engagement, and fuel work relationships."

But that is a flawed assumption. Without data on flexibility, it’s virtually impossible to say that removing access to flex will automatically yield better business outcomes. In fact, data shows that flexibility itself is rarely the source of a company’s performance issues; the culprit is often poor implementation. (Our research actually shows the contrary, that employees with access to flex are happier, more productive, have higher net promoter scores, and are less likely to quit.) The mistake these companies made wasn’t embracing a culture of flex, but rather failing to gather data—before rolling out their remote programs—on their existing flexibility supply and demand and the impact those flex gaps have on business risks and outcomes.

In the case of IBM or Yahoo, for example, this data might have revealed that the majority of employees didn’t need to work remotely 100 percent of the time, but would have been happier and more productive with access to location variety or even TimeShift. It’s important to remember that having access to flexibility isn’t the same as having access to the right kind of flexibility, and that can ultimately make or break the success of a flexibility program. It’s why one-size-fits-all programs typically underperform or fail entirely, especially when they are replaced by yet another one-size-fits-all solution, such as a company-wide clawback.

While we can never know exactly what happened when these companies decided to cancel their flex programs, one thing is for certain: flexibility data could have gone a long way in preventing them from falling off the flex glass cliff.