Most smart-building pilots end with a dashboard that everyone likes and nobody can use to make a decision. The devices worked, the data arrived, but no one can say whether the building is better, or whether rolling the system out to ten more floors is worth the money.
A pilot is an experiment. It should be designed to answer one question clearly. This brief sets out how owners and property teams can frame a pilot so that, at the end, the decision to scale up, change course or stop is obvious.
Choose one outcome
Start from a problem the business already feels, not from a technology. Good pilot outcomes are specific and owned by someone:
- Reduce complaints about washroom cleanliness on the busiest floors.
- Cut the number of manual patrols the facilities team makes each day.
- Find water leaks in plant rooms before they reach tenant areas.
- Show whether meeting rooms and floors are used enough to justify their running costs.
One outcome is easier to measure and easier to explain. If several teams want different things, run them as separate pilots or agree which one comes first.
Measure the baseline first
A result only means something compared with what happened before. Before installing anything, record how the outcome looks today, for long enough to cover normal variation such as weekdays, weekends and peak periods.
Use data the business already trusts where possible: complaint logs, work orders, patrol sheets, cleaning records or utility bills. If no baseline exists, a short manual count is better than none.
Pick a representative site
The pilot area should look like the places you would roll out to next. A showcase floor with new fit-out and enthusiastic staff will flatter the results.
- Choose a typical floor or building, with its normal mix of users and problems.
- Avoid areas that are about to be refurbished or change use during the pilot.
- Make sure the operating team on site has time to take part.
Define success before installation
Agree in writing what result would justify scaling up, and what result would mean stopping. For example: "If complaints on the pilot floors fall by a clear margin over the baseline period and the cleaning team says the alerts are useful, we will extend to the remaining floors."
Include practical measures as well as the headline number: how often staff acted on alerts, how many alerts were false, and how much maintenance the devices needed.
Run it long enough
A pilot needs time to settle. The first weeks are usually spent adjusting device positions, thresholds and workflows. Plan for a settling period and then a measurement period that covers the same kind of conditions as the baseline.
Keep a short log of changes made during the pilot, so the results can be explained later.
Make the scale-up decision easy
At the end, the pilot should produce a short report that answers four questions:
- Did the outcome improve compared with the baseline, and by how much?
- Did the operating team use the system, and what would they change?
- What did installation and maintenance actually involve?
- What would it cost, and what would change, to roll out further?
If the answers are clear, the decision usually is too.