Skip to main content
GullySystem

Pilot Launch Execution

Pilot launch execution is the controlled release of a first version to an invited group. It covers invited access, close monitoring of the early sessions and a read-out that decides what happens next.

Choosing who gets access

Access is limited to people who have the problem the product addresses.

  • Invite codes or approved mobile numbers, so access is granted and not discovered
  • People who genuinely have the problem, not friends being supportive
  • A mix that includes the awkward case, such as poor connectivity or an old phone
  • A clear end date for the pilot, told to users at the start

Agree the finish line first

Before anyone is invited, the exit rules are written down. Every rule is assigned to one person, and the decision date is fixed in advance.

A few pass or fail rules

Pick a small number that can be checked objectively, such as how many invited users complete the core task and how many return in the second week. Long wish lists dilute the verdict.

A severity rule

Define what counts as a blocking fault, for example lost payments or lost records. A common gate before widening access is a run of quiet days with no blocking fault still open.

A time limit

Pilots that never end become unofficial launches with no accountability. A fixed end date, with one agreed extension at most, keeps the group honest.

A decision owner

One person makes the call at the read-out. Shared ownership means the loudest voice wins.

The first two weeks of a pilot

Someone is watching

Errors, stuck records and half-finished journeys are reviewed the same day and not at the end of a sprint.

A direct line for users

One number or channel where a pilot user reports a problem and gets an answer from a person.

Manual rescue is allowed

If a booking fails, a team member calls and fixes it by hand. The failure is still logged.

Daily notes

What broke, where each user stopped and what was asked for. They are written the same day, not reconstructed later.

Controls that limit damage

A switch beats a scramble. Features can sit behind flags that let you show them to one group and hide them from another. If a feature misbehaves, you turn it off without a new release. Flags need housekeeping, though. Each one gets an owner and a removal date, otherwise old switches pile up and nobody remembers what they do. Widen access in steps, and hold each step for about a week before the next, since some faults only appear after sustained use.

The read-out at the end

A pilot closes with a decision. The read-out covers signups, completed journeys, repeat use, payments where payment was part of the test and the reasons people gave for stopping. Four outcomes are open. Go ahead, keep running for a set time, change the idea, or stop and redirect the remaining budget.

How the pilot looks from each seat

Three seats matter in a pilot: the person running it, the invited user and the developer watching. Each sees something the others cannot. This page follows one ordinary pilot day from the point of view of all three seats. Their views differ at first.

Pilot lead at the start

The lead sends a short note to the invited group with the day’s focus and a contact for help. Sessions are logged by user name. Afterwards the lead watches the first few from a shared screen.

Invited user at work

The user tries to complete a real task in the product. They hit a snag and message the lead. A reply arrives quickly, and the snag is written on the shared list. Replies stay short and kind.

Developer reading the log

The developer scans the day’s error log and the session list. Two errors repeat for one user. A fix is prepared and tested before the next day’s sessions begin. Quiet days still get a careful read.

Lead at the close

The lead writes a three-line summary of what worked, what failed and who was affected. The summary goes to the founder. Decisions about widening or pausing the pilot rest on those summaries. Decisions stay on paper.

Rules the founder sets before the pilot opens

Control is what makes a pilot safe. The founder decides in advance who joins, what is watched and what ends it. These rules are best written before the first invite goes out. Write them early, in plain words, and share them.

Who is invited

List the invited users by name and the reason for each. Filling it with friends proves little in practice. Include at least a few people who would have to pay for the product later. Names and reasons go in one list.

What gets watched

Pick the two or three events that matter, such as sign-up completed or first order placed. Watch those daily. A long list of numbers without a purpose gets ignored by everyone. Weekly figures hide short problems.

What stops the pilot

Agree the conditions that pause it, such as a payment error or lost data. Say who may press pause. A shared rule beats an argument held while a customer waits. Write the triggers beside the names.

How existing customers move over

If customers already use a spreadsheet or a messaging group, decide how they transfer. Moving ten users by hand is fine. Plan the order so nobody loses their history. Keep a note of who has moved.

Make an informed decision

Agree the evidence before signing off this part.

Use a representative scenario to decide what this part of the system must show and where review is still needed.

The read-out at the end

Can the same receipt be allocated without counting the money twice?

Take an unallocated amount in “The read-out at the end” and apply it to the intended balance. Review the remaining amount and the original receipt reference together.

Background reading: ERPNext: payment allocation

FAQ

Frequently asked questions

How long should pilot launch execution run?

It depends on your business cycle. The pilot should run long enough for that cycle to repeat at least twice. A booking product may show results quickly, while a monthly billing product needs a month end and the start of the next.

What if pilot users like it but nobody pays?

That is a finding, and a useful one. Where willingness to pay is part of the question, payment collection is built into the pilot so the result is measured and not assumed.

Do we tell pilot users they are on a pilot?

Yes. Users who know they are early give more feedback and accept rough edges. Hiding it means the first bug reads as a failure and not as work in progress.

How many users should a pilot have?

Enough to see the journey repeated by different people, yet few enough that you can talk to each one. That means a few dozen invited users and not thousands.

What is the difference between a pilot and a beta?

The words overlap. A pilot means a small, chosen group using the product in real conditions with close support. A beta is a wider, less supervised release. The label matters less than the exit rules.

Can we charge pilot users?

Yes, and it is wise when price is one of your unknowns. A small paid pilot shows real willingness to pay, while a free one shows only curiosity. Keep the terms simple and in writing.

Talk to us

Tell us what you need.

Tell us what your team does today, where work gets stuck and what you want to change.

What happens after you share your requirements

  1. We review your brief and the systems you already use.
  2. We clarify the workflow, missing information and the work to consider first.
  3. The proposed scope sets out the deliverables, responsibilities and dependencies.
  4. Review the scope and commercial terms before deciding whether to proceed.

Start with your requirements

Describe your current process, the people involved and the change you need. Start with the details you know; we can clarify the rest together.

Share your requirements