How many performances should a subtitle pilot include?

There is no responsible fixed number. The pilot must include enough variation to test its named uncertainty—for example different seating areas, audience arrival patterns, operators, or show conditions—without expanding beyond the team's ability to support every promised performance.

Should ticket buyers be told that subtitles are being piloted?

Yes. Describe what will be available, how to access it, any material limits, and the supported alternative if the preferred route fails. Calling it a pilot should create informed expectations, not excuse an unreliable access promise.

What evidence supports moving from a pilot to a full rollout?

Look for successful audience entry, sustained readability in representative seats, manageable support demand, reliable cue operation and recovery, and staffing that can be repeated without exceptional help across the variations the rollout must support. Usage counts alone do not prove those conditions.

When can an organization skip a subtitle pilot?

A separate pilot may add little when the same audience journey, venue conditions, operating roles, and recovery path have already been demonstrated in representative local use. Even then, a staged confirmation is safer when a new venue or audience promise introduces a material dependency.

Subtitle Pilot or Full Rollout? Test the Audience Promise, Not Just the Software | SurtitleLive Planning

Short answer

Run a subtitle pilot when it can answer a named uncertainty under representative conditions. Move to a broader rollout only when the audience promise, staff responsibilities, reading conditions, and recovery path have been demonstrated across the variations the rollout must support—not merely because the software ran for one performance.

Subtitle Pilot or Full Rollout? Test the Audience Promise, Not Just the Software

A small launch is not automatically a useful pilot. If it happens in the easiest room, with extra staff, invited users who already know the access link, and no credible failure rehearsal, it may prove little about the service that ordinary ticket holders will receive. The scale is smaller, but the uncertainty remains.

A responsible pilot starts with a public-facing promise: which audience will receive which text, where they will read it, what guidance and alternatives will exist, and what the team will do when the preferred path fails. It then tests whether that promise survives real seats, real arrival patterns, real rehearsal changes, and the staffing level that later productions can actually sustain.

The purpose is not to make audiences absorb the risk of an unfinished rollout. It is to limit the scope of uncertainty while giving pilot participants a supported, honestly described service—and to make expansion, revision, or stopping equally legitimate outcomes.

Treat the Pilot as a Decision Instrument

Name the uncertainty: audience entry, sightlines, language demand, cue operation, support load, or recovery
Choose conditions that expose that uncertainty instead of protecting the preferred solution from a fair test
State what evidence would justify expanding, revising, repeating, or stopping before the first performance
Give one accountable owner the authority to act on the result rather than letting the pilot continue indefinitely

Choose Scope by Consequence and Reversibility

One-production pilot: use when the team still needs to prove the full journey from preparation and audience guidance to live recovery
One-venue pilot across several shows: use when the room is understood but repertoire, operators, or audience patterns vary
Staged rollout: use when the workflow is credible but different venues or departments should enter in controlled waves
Full rollout: use only when representative evidence already covers the important variations and owners can support them at scale
Not ready to promise: use when no bounded scope can yet provide readable text, usable access, adequate support, and a credible fallback

Write the Pilot Charter Before Tickets Go on Sale

The exact audience promise and any limits that must be communicated before booking or arrival
The seats, languages, performance conditions, and audience journeys the pilot must represent
How participants can report difficulty without having to justify or publicly disclose an access need
How informed consent, data handling, and fair compensation will work when people contribute time or lived expertise to research
The front-of-house, operator, and backup roles that will exist under normal—not exceptional—staffing
The minimum evidence for expansion and the stop conditions that trigger repair before another performance
When the decision will be made, who owns it, and how affected audiences will hear about any change

Measure Access and Dignity, Not Adoption Alone

Scan counts and viewer sessions can show that some people reached the service. They do not show who abandoned the process, avoided it because the instructions were unclear, could not read from their seat, or needed help but did not want to identify themselves publicly. Evidence should include attempted access, successful entry, sustained readability, help requests, alternatives used, and recovery from the most credible failure.

Feedback is strongest when it includes the people the service is intended to support, without treating their time or lived expertise as free quality assurance. Explain what is being tested and how information will be used, obtain informed consent, provide fair compensation where appropriate, keep participation optional, offer a working fallback, and close the loop on what changed. A pilot label must not become permission to advertise access while transferring avoidable failure to the audience.

SurtitleLive can support a bounded pilot using the same prepared cues for projection and browser-based audience delivery. It cannot decide whether the tested seats, guidance, support model, participant safeguards, or evidence are representative. Those remain production responsibilities.

Expand, Revise, Repeat, or Stop

Expand when the promised experience worked across the representative conditions named in the charter and the team can reproduce it with ordinary staffing. Revise when a specific barrier has a credible repair. Repeat only when the next test answers a different uncertainty; repeating the same protected pilot is not accumulating useful evidence.

Stop or narrow the promise when a critical audience path remains unusable, support demand exceeds the operating model, or recovery cannot be rehearsed reliably. That decision protects trust. Quietly broadening a pilot because no major complaint arrived is not the same as proving that the service worked.

To compare the delivery modes a pilot may need to test, continue with Mobile Subtitles vs Projection Surtitles. For training readiness before expansion, continue with How Much Training Does a Subtitle Rollout Require?.