Adding a second location is where hardware-based translation becomes genuinely painful. Everything has to exist twice: the transmitter, the receivers, the volunteer who knows how it works, and the person who remembers to charge the batteries on Saturday night. Each site develops its own quirks, its own failure modes, and its own reason why it did not work last Sunday.
The equipment problem is the visible one. The coordination problem is the one that actually decides whether translation survives at a second site.
What changes when the hardware is a phone
The duplication disappears, because attendees bring the equipment. Nothing needs charging, nothing needs counting back in, and a site that gains fifty people one Sunday does not need fifty more receivers.
What remains is genuinely a coordination question: who starts the broadcast at each site, which languages each site offers, where the codes are displayed, and who notices when something is wrong. Those are people questions, and they are worth planning as carefully as churches once planned the hardware.
Sites are rarely identical, and the mistake is assuming they are
The common error is treating campuses as copies. In practice one site sits in a neighbourhood where three languages matter and the other serves a congregation that is almost entirely local. Or the city-centre site fills with students and the suburban one with families.
Set the languages per site based on who is actually in that room, and review it after a few months. Neighbourhoods change, and a site's language mix will drift in ways nobody predicted at the launch meeting. The figures will tell you before the leadership notices.
One caution in the other direction: keep at least one thing identical across sites, which is how a person joins. A member who moves between campuses should not have to relearn anything.
Live video links between sites
Many multi-site churches broadcast the sermon from one campus to the others. This creates a specific technical decision that is worth getting right the first time.
Take the audio for translation from the source — the room where the preacher actually is — rather than from the receiving room's speakers. Audio that has been through a video link, out of a speaker, across a hall and into another microphone has lost a great deal, and every step adds delay. Translating from the original feed gives better text and gets it there sooner.
The practical arrangement is one broadcast, started at the source site, which every campus joins. The receiving sites do not run their own translation of the same sermon; they display the same code.
When each site preaches its own sermon
Where campuses teach independently, run a separate broadcast per site. Two useful consequences: each site's usage and languages are measured separately, which is exactly what you need for the review above, and a problem at one site does not affect the others.
Overflow rooms and the rooms you forget
The same approach covers rooms nobody puts on the plan: an overflow hall, a parents' room, a foyer with a screen, a crèche, a corridor where people stand at Christmas. Anyone within reach of the code can follow, in any language — including the language being spoken.
That last point is quietly useful. Side rooms usually have the worst speaker feed in the building, and a parent in the crying room can read along in the main language and follow properly for the first time. Churches add translation for the immigrant families and discover the parents' room uses it most.
Practical arrangements
- One operator per site. Do not try to run a remote site from the main one. The person who can see the room should hold the controls, because they are the only one who will notice that the microphone was never switched on.
- Two trained people per site, not one. The single-volunteer model is how translation quietly stops at a campus: they move house, and nobody else knows how it works.
- Keep the signage identical. Same card, same wording, same place at every site.
- Agree who is told when it breaks. One name, one phone number, at every campus.
- Review usage per site each term. The languages each location needs will diverge, and the point of measuring is to act on it.
A word on the volunteer problem
Multi-site churches usually underestimate this. Central staff can set up an account, choose the languages, and print the cards. They cannot be in three buildings at nine o'clock on a Sunday morning.
The thing that makes it survive is that starting a broadcast has to be genuinely simple — a task a rota volunteer can do correctly on their third-ever attempt, half asleep, ten minutes before the service, without a manual. If it needs a specialist at each site, it will work at the main campus and lapse everywhere else within two months. That is the honest test to apply before rolling anything out to a second location.
Keep the cost proportional
Because usage is measured in hours actually broadcast, a second site running one service a week costs roughly one service a week — not a second full set of everything. This is the practical reason a small church plant can offer the same languages as the main campus, which under the old model was never affordable.
Set up your second site
InMyTongue supports multiple halls under one church account, each with its own code, its own language list and its own usage figures, so every campus can be set up for the room it is actually in. New accounts include free credit. See how multi-site works.