A Nudge Is Not a Diagnosis
“No one can whistle a symphony. It takes a whole orchestra to play it.”
We (Groops) are about to put a tool into the world that measures teams. So when I read Gartner's prediction that companies will start buying technology to fix how their people communicate, I have two reactions.
The first is relief. I have spent years arguing that cohesion is performance infrastructure rather than a soft cultural nicety, that it sits in the gap between strategy and execution where most of the value quietly leaks out. Now the analysts are saying it out loud and the category has a name. Awesome.
The second is skepticism and discomfort, because I know what is about to get sold under that banner. Gartner calls the emerging set nudgetech, tools designed to bridge the professional communication gap and restore effective collaboration (Gartner, 2026). The prediction is right and the problem is real. But a nudge is a small behavioral prompt, and if there is no psychological model sitting underneath it, a nudge is a guess delivered with confidence. It might change the behavior but not the root cause.
The Myth of the Better Tip
Underneath most workplace technology is an assumption that people underperform because they do not know what to do. Leaders think that if they close the information gap, better behavior follows.
Gartner's own examples make that assumption visible. AI that prompts an employee to use email rather than text based on a client's preference, reminders to managers about a direct report's working style, custom communication tips generated on the fly (Gartner, 2025). All of that is genuinely useful and none of it is the thing that is broken.
Most people already know what good communication looks like. They know they should raise the concern, name the disagreement, tell the client the timeline slipped but something else is stopping them.
What Actually Stops People
Gartner found that fewer than one in three employees feel comfortable being vulnerable with their colleagues, and that across generations, few teams experience high cohesion (Gartner, 2026). Turn that around and it means roughly seven in ten people are carefully managing what they say to the people they work with most closely. That is not a skill problem. That is a read on risk.
Amy Edmondson's research on hospital teams gave this its clearest shape. She found that what distinguished teams was a shared belief about whether the group was safe for interpersonal risk taking, and that this belief predicted whether people asked questions, admitted mistakes, and surfaced problems early (Edmondson, 1999). What looks like a communication failure is often a quiet calculation about what honesty costs here.
A tip cannot solve a calculation. Someone who has already concluded that speaking up will be held against them will follow the nudge, switch to the recommended channel, adopt the recommended tone, and still say nothing that matters. The communication metrics improve and nothing changes about what the team actually knows about itself.
The Credibility Problem
This one really hit home. Gartner reports that only 14% of employees think algorithms cannot give fairer feedback than their managers (Gartner, 2026). It is an awkward sentence to read, so flip it. Nearly nine in ten employees believe a machine would judge them more fairly than the person who manages them.
I do not read that as enthusiasm for automation. Nobody wakes up hoping software will rate their performance. I read it as a verdict on manager credibility, and it makes more sense next to a second finding from the same research, that six in ten employees say they are not getting the coaching they need to do their core job (Gartner, 2026).
People are not asking for a bot.
They are asking for fairness and for someone to invest in their development, and enough of them have stopped expecting either from a human that they will take their chances with software. They are settling for the breadcrumbs.
That is the real market Gartner is describing. Not a communication gap. A trust gap that has gotten large enough to look like a technology opportunity.
Why This Matters Now
Gallup's midyear data offers the counterweight. Employees whose manager actively supports how the team uses AI are engaged at 48%, compared with 30% for employees without that support. It is the widest gap in the whole report, and it is worth sitting with, because both groups have the technology. Everyone in that comparison has access. The only difference is whether someone above them made sense of it for them (Gallup, 2026).
That is a diagnostic act before it is a communication act, and it is exactly the part nobody is resourcing. Only about one in five CIOs say they are focused on limiting the harm AI might do to employees' work or their wellbeing (Gartner, 2026).
So the tools are arriving, the people buying them are largely not watching what happens to the humans on the other end, and the one thing that reliably predicts whether any of it lands is a leader who understood the situation first.
The Groops Take
Every tool that touches a team is making a claim about what is wrong with that team. Most of them never say the claim out loud. A nudge assumes the problem is behavior. A diagnostic assumes the problem is knowable. A coach assumes the problem is worth sitting with.
We think the order matters more than the tooling. Understand the pattern, name it back to the team itself, then change one thing. Skip the first step and you are optimizing behavior you have not understood, which is how organizations end up with better-mannered meetings and the same unspoken problem underneath them.
Something to Try This Week
No tool required for any of this. Four steps, one week.
Ask one anonymous question. "What is something you wanted to say in a meeting this quarter and did not?" One question, genuinely anonymous, no follow-up fields.
Read for pattern, not for people. You are looking for the shape of what goes unsaid rather than who said it. Resist the urge to identify anyone, because that instinct is the same one your team is already protecting itself against.
Say one thing back out loud. Name what you heard and what you are going to do about it. This step builds the safety, not the survey. A question with no visible response teaches people that speaking up disappears into a void.
Change one thing, and let the team pick it. Their choice will be smaller than yours would have been, and it will stick.
Want to Measure This in Your Own Team?
This is the reasoning behind what we built. Groops is a Team Intelligence System, and the Cohesion Diagnostic sits at the front of it because we do not think you can meaningfully change a team you have not first understood.
Here is how the pieces fit together. Your team takes the assessment. You get results that name the pattern rather than rank you against a benchmark. Then you sit down with a Groop Guide, a real person, to work through what those results mean for your specific team and to decide what to change first. From there the AI Coach carries the work forward between sessions with nudges shaped by what your own diagnostic surfaced, and your team keeps access to a live virtual team coach for the moments that need a human in the room.
The AI Coach is grounded in the same training our Groop Guides receive. Same psychological frameworks, same understanding of what happens to teams under pressure, same view of where the leverage sits. That is deliberate. It means the nudges your team gets are not generic communication tips. They are informed by what the diagnostic already knows about you, and they hand off to a person when the moment calls for one.
This is what we believe a real partnership between AI and humans looks like in team optimization. Not software standing in for people, and not people doing work that software can carry well. Each doing the part it is genuinely better at, and a clear line between them.
We are opening a small group of design partner spots now as we build out the larger platform. Design partners get the full diagnostic for their team (team and individual report on the platform), the results session with a Groop Guide, ongoing AI Coach support, and live team coaching, along with a direct hand in shaping how the platform develops. The decision window is open. If you want to know what your team is actually carrying rather than what it reports on an engagement survey, this is the moment to be part of building it.
References
Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350-383. https://doi.org/10.2307/2666999
Gallup. (2026). Employee engagement remains flat as AI adoption accelerates. https://www.gallup.com/workplace/712433/employee-engagement-remains-flat-adoption-accelerates.aspx
Gartner. (2025, January 8). Gartner identifies top nine workplace predictions for CHROs in 2025 [Press release]. https://www.gartner.com/en/newsroom/press-releases/2025-01-08-gartner-identifies-top-nine-workplace-predictions-for-chros-in-2025
Gartner. (2026, January 8). The future of HR: Clear the hurdles, reap the benefits. https://www.gartner.com/en/articles/future-of-hr