Your team looks busy. Reports are getting written faster, emails go out quicker, and everyone seems to be moving. That’s not the same as everyone actually getting more done, and new research suggests that gap is one of the bigger productivity leaks most business owners don’t realize they have.
What’s actually happening
Workday’s research gave it a name: human middleware. It’s what happens when employees end up as the connective tissue between AI tools that don’t talk to each other, copying data from one system into another, checking whether two apps agree with each other, rewriting prompts to give an AI tool context it doesn’t have on its own, and fixing outputs that almost worked but not quite.
The scale of it is the surprising part. Workday found that a quarter of workers spend seven or more hours a week doing exactly that, and 62 percent spend at least half their working time translating between siloed systems instead of doing the work those systems were supposed to support. As a Workday VP put it, “too many employees are serving as the human middleware between disconnected AI systems.” That’s not a fringe problem. That’s most of a workday.
Why it hides itself so well
Here’s the part that should actually worry you. Despite all of this, around nine in ten workers in the same research still reported a strong sense of progress and ownership over their work, and nearly half said AI had genuinely sped things up. People don’t feel the drag, because moving information between apps still feels like work. It looks like productivity from the outside and it feels like productivity from the inside. It just isn’t creating much value.
That’s the trap. A business owner watching a busy team has no easy way to tell the difference between someone solving a problem and someone acting as a translator between two apps that were never built to talk to each other.
What to actually do about it
Before you add another AI tool, look at how information already moves through the ones you have. Where does data live, which systems can actually see each other, and where are people still manually carrying information from one place to another. Workday’s research points at a few common culprits worth checking for yourself: approval processes that are heavier than they need to be, AI access or training that’s inconsistent across the team, workflows that were built before the AI tools existed and never got revisited, and data that’s inconsistent enough that nobody fully trusts it without checking by hand.
Adding more tools on top of that doesn’t fix it. It usually adds another seam for someone to sew shut by hand. If you want a clear picture of where your team is actually spending that time, that’s exactly the kind of strategic IT review we do.