FOR MUSEUMS

Interpretation that scales with the collection.

Museums are the core Convo use case. Art, science, natural history, history, university, small-to-mid municipal — the shape of the problem is the same. You have a collection visitors come to see, interpretation that has historically been expensive to produce, and an audience that increasingly arrives expecting it in their first language and the ability to ask follow-up questions. Here’s what changes when that work runs on Convo, and where to start.

WHO THIS IS FOR

Most museums, honestly.

When I say Convo is built for museums, I mean a wider range than the phrase usually implies. Art museums with permanent collections that rarely turn over. Science museums with rotating special exhibits where the interpretation has a shorter half-life than the gallery. Natural-history museums with specimen-rich halls where families want to know what they’re looking at and why it matters. History museums and historic-house museums where the narrative does most of the work. University museums with scholarly source material that already exists in the form of catalog essays and didactic panels. Small municipal and regional museums where one curator wears five hats and there hasn’t been a budget for a real audio guide in a decade.

What ties them together: a collection people travel to see, an audience that no longer all speaks the institution’s home language, and an interpretation budget that has not kept pace with what visitors now expect. Most museums I talk to are sitting on excellent written source material — wall cards, catalog entries, exhibition essays — and have never been able to make it audible without commissioning a project. That gap is what this is for.

Where this isn’t the right answer: if your visitor experience is built around timed-entry guided tours led by a docent, and the docent is the product, Convo is less interesting. We complement a docent program; we don’t replace one. And if your institution is a research archive with no walk-in public — Convo doesn’t help.

WHAT CHANGES

The math of interpretation changes.

The first shift is cost and time. A traditional audio guide is a project: scripts, a recording studio, a voice actor, a post-production round, then the same again in every additional language. Convo drafts each stop from the source materials you already have, voices them, and translates them into ten languages — English, Spanish, French, German, Italian, Portuguese, Chinese, Japanese, Korean, Arabic. A first gallery, fully built, is about two weeks. Subsequent galleries are faster because the voice and the editorial register are already established. Practical detail: /implementation.

The second shift is multilingual as the default, not a special project. Every tour ships in all ten languages from day one. An edit to an English stop re-voices the same stop across every language in roughly a minute. You stop having an English tour and a translation backlog; you have one tour, ten ways. Detail: /multilingual.

The third shift is the conversational layer. After listening to a stop — or instead of listening to one — a visitor can ask a question, in voice, text, or by pointing their camera at the object. The guide answers from the source materials you uploaded, declines when it can’t, and logs the exchange so you can see what visitors actually wanted to know. Over a quarter, that log becomes a curator artifact: the gaps in your wall text, the questions you didn’t know your visitors were asking. Detail: /visitor-qa.

The fourth shift is the update cadence. A correction — a wrong date, a re-attribution, a new acquisition next to an old one — is a Tuesday, not a project. You edit a stop, the audio re-renders across every language, and the QR card on the wall still works. This matters most for science and natural-history museums with rotating exhibits, but every museum has had the experience of living with a recorded mistake for a year because re-recording cost too much to justify.

Sub-type, briefly: art museums lean hardest on the multilingual layer and the Q&A — international visitors, deep curiosity, long dwell times. Science museums lean on the authoring speed for rotating exhibits and the multilingual layer for family audiences. Natural-history museums look a lot like science museums. History and historic-site museums lean on the Q&A — narrative tours generate the most questions. University museums lean on the source-grounded answering because their material is already scholarly. Small museums lean on cost and time-to-launch — the pilot is free, and two weeks is shorter than most internal budget cycles.

What hasn’t changed: a curator still decides what the tour says. The tool drafts, voices, and translates; the curatorial judgment about how a piece should be framed stays with the institution. The curator workflow is at /authoring; the product overview is at /product.

WHERE TO START

One gallery. One quarter. Then decide.

The right way to evaluate Convo is to run it next to whatever you have today and let your visitors decide. Pick one gallery — a permanent room you wish had better interpretation, or a special exhibit opening in the next quarter. We build the tour in about two weeks, free. The QR code goes up at the gallery entrance alongside your existing signage. Your curator keeps editing access. Visitors do what visitors do.

At the end of the quarter, you have actual evidence: how many visitors used it, what they asked, how long they stayed in the gallery, what they said in feedback. If it worked, you expand it — another gallery, another site, the whole institution. If it didn’t, you keep everything we made and the experiment cost you a curator’s time and nothing else. Most institutions don’t want to make a software decision before seeing the thing work in their building. The pilot exists so they don’t have to.

For the operational detail of how those two weeks go, see /implementation. For the budget side and what the tiers look like after the pilot, see /pricing and the cost pillar at /resources/buying-and-cost. If you’re a multi-site institution and want to see what the rollout shape looks like across sites first, there’s a long-form piece at /resources/operations/multi-site-museum-tour-management.

COMMON QUESTIONS

What museum teams ask first.

Yes, and it gets easier per site, not harder. The first site does the heavy lifting on voice, tone, and source materials. Each subsequent site reuses the voice and the writing patterns and only adds what’s genuinely different. Curators at each site keep editing rights over their own tours. There’s a longer write-up of the operational shape of this in the resources hub under multi-site management.
You write what the tour says. The visitor guide answers only from materials you’ve uploaded — not the open internet — and it declines when it can’t ground an answer in your sources. Every visitor question and response is logged in the admin so you can review what’s being asked and how it’s being answered, and edit the source materials if anything is off. For repatriation language, contested provenance, indigenous attribution, or anything where the wording itself is the curatorial decision, you should write that wording. The tool doesn’t override it.
No rip-and-replace. Most institutions run Convo in one gallery alongside their existing audio guide for a quarter or two. Visitors with rented devices keep using those; visitors with phones can scan the QR. You compare what people actually do. If Convo wins on usage and on what visitors leave the gallery understanding, you expand it as the existing contract renews. If it doesn’t, you’ve only changed one room.
Yes. The pattern of "an object with a story, a visitor who wants to understand it, an institution with deep source material" is the same. Science museums and natural-history museums often benefit even more from the multilingual layer because their audiences skew family and international. Specimen interpretation, exhibit narratives, and "what is this thing and why is it on display" are all things the tool handles well.
If you have at least one gallery worth interpreting and someone on staff who can spend a few hours over two weeks reviewing drafts, you’re the right size for the pilot. The pilot is free, so the question isn’t whether you can afford it — it’s whether you have a curator’s time. A small historical society with one exhibit is a real Convo customer profile.
WHAT WE’RE ASKING

Pick one gallery.
Give us two weeks.