How to pilot an employee-owned tool (a buyer's note)

We are a vendor. You should read this with that caveat in mind. But we are also research-literate enough to know what a good pilot of an employee-owned workspace looks like, and we would rather you run one well — and walk away if it doesn't work — than buy something on the back of a slick demo and feel burned in eighteen months. The market for honest People-tech is small. We would like to grow it.

This is a buyer's note on what to evaluate, what to ignore, and what to refuse to compromise on.

The thing you are actually testing

A pilot of an employee-owned workspace is testing a specific theory of change. The theory is: when employees have a private, autonomy-preserving space to capture their work, reflect weekly, and prepare for 1:1s, the resulting behavior pattern produces engagement, retention, and visibility outcomes that are better than what your current programs produce. The mechanism is well-grounded in the literature on autonomy (SDT), progress visibility (Amabile), structured reflection (Di Stefano), and manager-conversation leverage (Gallup, Kahn). The product is the implementation. The pilot is the empirical test in your environment.

This is not a survey replacement evaluation. It is not a productivity tool evaluation. It is not a Slack-or-Notion replacement evaluation. The question is whether the structural design of the workspace pulls the mechanism the literature points at, in your specific organization, with your specific employees.

Keep that question in mind through the six months. Everything else is implementation detail.

What to look for in adoption

Adoption is the first signal. If employees do not actually use the workspace, no further analysis matters. The number you want to see is the fraction of seats that captured at least one win, challenge, or reflection in a given week, sustained over time.

A healthy pattern for a six-month pilot looks roughly like this. Week one: 60-80% activate, drawn by curiosity. Week three: drops to 30-40% as the novelty fades. Week six: stabilizes at the floor — the fraction of employees who actually find the workspace useful enough to return to weekly. That stabilization is the number that matters. If it is below 30%, the design is not producing enough value to overcome the friction. If it is above 50%, you have a real adoption pattern.

What is the floor doing across teams? Some variance is expected — engineering will look different from sales which will look different from operations. Wide variance within a single team suggests the manager either champions or undermines the workspace, and that is itself data.

What to look for in cadence

Adoption is necessary; cadence is sufficient. A team where 80% of seats captured a win this month but only 20% drafted a weekly reflection is using the workspace as a scratch pad. That is fine but it is not the mechanism. The literature on structured reflection only applies to people who are reflecting structurally.

The number to watch is the fraction of active seats that completed a weekly reflection in a given week. Healthy pilot teams cluster at 50-70% of active seats reflecting weekly by month three. The Di Stefano field experiments on reflection identified an 18% performance improvement in their specific context; the mechanism requires the habit, not just the tool. The cadence number tells you whether the habit is forming.

Open-loop closure rate is the third cadence signal. A team where employees flag items in 1:1 agendas and the items get closed with outcomes is having functional 1:1s. A team where loops pile up is signaling a manager problem you can address — even though you cannot read the loops themselves. Watch the ratio.

What to refuse to compromise on

This is where most pilots go wrong. The vendor offers, in the name of the pilot, to bend the privacy boundary for "evaluation purposes." Admin access to one team's content. A sentiment analysis run over reflections, returned as aggregates. A demo of what the dashboards "could" look like if you wanted them. Do not accept any of these.

The reason is structural. The mechanism the literature underwrites depends on employees actually trusting the privacy guarantee. The moment the pilot makes the guarantee conditional, the underlying behavior changes. Employees write performatively. The signal you measure during the pilot is the artifact of the bent guarantee, not the mechanism. The pilot is no longer a test of what we said it would test.

A vendor willing to bend the privacy boundary to close the sale is signaling that the privacy boundary is marketing, not architecture. The right pilot is the unbent version, even though it is harder to sell and harder to evaluate. If the unbent version cannot produce the engagement signal you want, the bent version was not going to either; it would just have produced a prettier dashboard you could not trust.

What to evaluate qualitatively, at the end

Six months in, run honest qualitative read with a sample of pilot employees. Ten people, separately. Ask:

• Did you use the workspace? At what cadence?
• What did you use it for that you weren't doing before?
• Has anything about your 1:1s, your work conversations, or your sense of your own progress changed?
• Do you believe the privacy guarantee? Why or why not?
• Would you keep using it if your organization rolled it out broadly?

The last question is the integration test. If a meaningful fraction says yes, you have a tool worth scaling. If most say "I tried it for a month and forgot," you have a design problem. Either way, the qualitative read is what tells you whether the quantitative signal is real.

We will be honest with you in this conversation. If the pilot doesn't produce the signal in your environment, we will tell you so before you renew. We are trying to build a serious product for the long run; we are not optimizing for short-term close rates.

What success looks like

A pilot that worked produces, at the six-month mark, four things. A stabilized weekly active rate above 30% of seats. A weekly reflection cadence above 40% of active seats. A team-level open-loop closure rate that has stayed above 60%. And qualitative read from your pilot employees confirming that the privacy guarantee is real to them and the workspace has produced value they want to keep.

If you see these four numbers, you have evidence the mechanism is pulling in your environment, and a credible basis for a broader rollout. If you see two of them but not the others, you have a design or rollout problem worth diagnosing. If you see none of them, the bet did not pay in your environment, and you should walk away. Walking away after a pilot that did not work is the cleanest possible outcome — for both sides — and we will not try to talk you out of it.

This is the only honest pitch for a pilot of a tool that explicitly stops trying to read your employees. It is also the only one that respects you as a buyer.

How Nela Helps

Use Nela to log your wins, track your challenges, and build a private 1:1 agenda from your own evidence for your next conversation. Your data is owner-only at the database — enforced by Postgres Row-Level Security, not just hidden in the UI — and only you can read it back through the app. Request pilot access.

Further reading

• Di Stefano, G., Gino, F., Pisano, G. P., & Staats, B. R. (2014). "Learning by Thinking." Harvard Business School Working Paper 14-093.
• Gallup. (2015). State of the American Manager.
• Frey, B. S. (1993). "Does Monitoring Increase Work Effort?" Economic Inquiry.