Here is what 85% of leaders told Microsoft researchers: they find it challenging to trust that hybrid employees are being productive (Microsoft Work Trend Index, 2022). Meanwhile, 87% of employees said they are productive. Satya Nadella called this gap "productivity paranoia" and told leaders they need to get past it.
Three years later, most of them have not.
Instead, they watch the green dot. They check who is online at 9:01 AM. They pull Jira velocity reports and treat story points like a stock ticker. They count pull requests and call it "output." They measure everything that is easy to measure and almost nothing that matters.
This is not a management philosophy. It is a coping mechanism for leaders who do not know what healthy teamwork looks like from the outside.
The Four Fake Signals
Let's name them. These are the metrics remote leaders default to, and every single one fails the same way: it measures activity, not effectiveness.
1. The Slack Green Dot
The green dot means one thing: someone interacted with Slack within the last 10 minutes. It disappears during exactly the kind of deep, focused work organizations value most. An engineer in a flow state shows as "away." A person scrolling channels and procrastinating shows as "active." The signal is inverted (Remote Report, 2021).
People know this. So they game it. Mouse jigglers, auto-clickers, Slack bots that post on schedule. A cottage industry exists to keep the green dot lit while the person does literally anything else. When your metric is this easy to fake, it is not a metric. It is theater.
2. Jira Velocity
Goodhart's Law states: "When a measure becomes a target, it ceases to be a good measure." Jira velocity is the textbook example. Teams inflate estimates so completed points look impressive. A task that used to be a 3 becomes a 5. Velocity goes up. Actual throughput stays flat. Nobody says anything because the chart looks good in the all-hands (Jellyfish, 2025).
Worse, velocity punishes teams that invest in quality. Refactoring does not produce story points. Mentoring a junior developer does not produce story points. Paying down tech debt does not produce story points. The work that prevents future fires gets deprioritized because it does not move the number that leadership watches.
3. Pull Request Count
More PRs do not equal better engineering. A developer who ships one well-architected PR that solves a systemic problem is more valuable than a developer who ships twelve small PRs, each of which patches a symptom. But PR count does not distinguish between the two.
PR count also incentivizes splitting work into tiny increments not because smaller PRs improve review quality (they do, to a point) but because more PRs look more productive to whoever is staring at the dashboard.
4. Email and Message Volume
ActivTrak's analysis of 443 million hours of work activity found that after AI tool adoption, time spent on email increased 104% and chat/messaging surged 145% (ActivTrak, 2026 State of the Workplace). Meanwhile, average daily focused time dropped by 23 minutes per user.
More messages. Less thinking. If message volume were a proxy for team health, the most dysfunctional teams would be the healthiest. They are usually the ones generating the most back-and-forth because nothing gets resolved the first time.
What Actually Predicts Team Health
Google spent two years studying 180+ teams in Project Aristotle, looking for the variables that separated effective teams from mediocre ones. They expected to find that team composition mattered most. Hire the right people, get the right result. That was wrong. What mattered was how the team interacted (Google re:Work, 2015).
The five dynamics they identified: psychological safety, dependability, structure and clarity, meaning, and impact. Psychological safety came first. Teams with high psychological safety were rated as effective twice as often by executives, brought in more revenue, and retained members at higher rates.
None of those dynamics show up in a Slack activity log.
Real team health signals are behavioral. They look like this:
Who speaks up when the stakes are high. Not who talks the most in stand-ups, but who raises a concern when a deadline is unrealistic or a technical approach has a flaw. Silence under pressure is one of the clearest signs a team lacks psychological safety.
How conflict gets navigated. Does the team address disagreements directly, or does everything route through a single decision-maker because the group cannot handle tension? Conflict avoidance looks calm on the surface. Underneath, it corrodes decision quality.
Who bridges disconnected groups. Every organization has silos. The people who move information and trust across those silos are disproportionately valuable. They rarely show up in output metrics because their contribution is relational, not transactional.
Whether new information changes behavior. A healthy team updates its approach when it learns something new. An unhealthy team plows ahead with the original plan because changing course feels like admitting failure. Adaptability is a signal. Stubbornness disguised as consistency is noise.
The Measurement Gap
Here is the problem: behavioral signals are hard to observe in remote and hybrid environments. You cannot read the room when there is no room. Managers who led in-person teams developed intuition for these patterns over years of shared physical space. Remote work stripped that away, and nothing replaced it.
That gap is why leaders reverted to activity proxies. The green dot is a terrible signal. It's just a visible one, and visible beats invisible when you're anxious.
The real question is whether you can make behavioral signals visible without creating a surveillance apparatus that destroys the trust you are trying to measure.
This is where QuestWorks sits. It is a flight simulator for team dynamics, a platform where teams work through realistic scenarios together and their interaction patterns become data. QuestDash surfaces behavioral callouts: who stepped up, who bridged groups, who adapted under pressure. It is available to everyone on the team, not just managers. Leaders get a separate weekly team health report.
The design matters. This is not keystroke logging. It is not screenshot capture. Players opt in, participation is voluntary, and the data reflects how the team collaborates, not how many hours they were online. HeroGPT provides private coaching that never shares upstream. HeroTypes give teams a shared vocabulary for working styles, visible to all.
Compare that architecture to watching a green dot and hoping it means something.
Moving From Activity to Interaction Quality
If you want to stop relying on fake signals, here is a practical shift.
Replace "hours online" with "interaction quality." Do your meetings end with clear decisions and owners, or do they end with "let's take this offline" (which means nobody will)? Track decision velocity, not attendance.
Replace "velocity" with "what shipped and what it did." Stop measuring points. Start measuring outcomes: did the feature reduce support tickets, did the migration land without rollbacks, did the new onboarding flow improve 30-day retention? Outcomes are harder to game because they connect to reality.
Replace "who is active" with "who is connected." Organizational network analysis shows that the structure of communication matters more than its volume. Teams with bridging connections across groups outperform teams where everyone talks to the same five people. You want to see breadth and reciprocity, not just frequency.
Replace "sentiment surveys" with "behavioral observation." Surveys tell you how people feel about the team. Behavioral data tells you how the team functions. Both matter. Most organizations have the first and none of the second. QuestWorks fills that gap without the surveillance tradeoff that makes behavioral observation radioactive in remote contexts.
Microsoft's own researchers put it directly: the antidote to productivity paranoia is "judging performance based on the quality of the outcomes and the impact of the work, not by how many emails were sent and how early in the morning they started" (Microsoft WorkLab, 2022).
The tools exist. The data exists. What is missing, in most organizations, is the willingness to stop watching the easy numbers and start reading the hard ones.
Try it free
