A director walks into the planning meeting for a 12-week show and asks the obvious question: is it even worth producing an audio guide for something this short? By the time we've written, recorded, and translated, the show's almost over.
For most of the last twenty years, the right answer was no. Today the right answer is almost always yes — with a few honest exceptions worth naming up front.
Why the math used to make this a no
Audio guide production was a 3 to 6 month process. You locked the checklist, drafted scripts, ran them past curators, hired voice talent, recorded in a studio, edited, localized into two or three languages, did QA on hardware devices or an app build, and trained front-of-house staff on distribution. Optimistically: 12 weeks. Realistically: 16 to 24.
That timeline killed the case before the budget conversation started. A show running 12 weeks needs a finished, locked checklist four to five months before opening — exactly when curators are still negotiating loans, swapping pieces that didn't clear conservation, and rewriting the wall texts. Late changes meant either re-recording (expensive, slow) or shipping a guide that referenced objects no longer in the room.
Then there was the cost. $20,000 to $50,000 for production, paid before a single visitor walked in. For a show drawing 80,000 visitors with a 10% adoption rate, that's $4 to $6 per actual user. If you charged $5 for the guide, you broke even on the lucky shows and lost money on the rest. Compare that to a permanent collection guide at the same price, amortized across hundreds of thousands of users over five years. The per-visitor cost was pennies.
Everyone in the field understood this implicitly. So temporary exhibitions got a printed brochure, maybe an audio descriptions track for accessibility compliance, and a hope that the wall labels did enough work. The most ambitious, most marketed, most talked-about programming a museum did got the thinnest interpretation in the building.
What flipped
Two things changed at once, and they reinforce each other.
The first is generation. An AI audio guide doesn't get recorded. Speech is generated from the exhibition data — object texts, curatorial notes, the catalogue essay if you have one — at the moment a visitor asks for it. There's no studio booking, no voice actor schedule, no separate localization budget for each new language. We've covered the mechanics of this in our piece on audio guides for temporary and travelling exhibitions; the short version is that what used to be a production project is now a configuration task.
The second is pricing. Production costs collapse, but the more important shift is structural: usage-based pricing means no minimum spend and no upfront commitment. You pay for the sessions that happen. If 30,000 of your 80,000 visitors use the guide, you pay for 30,000 sessions. If the show under-performs at 20,000 visitors total, you pay less. When the show closes, billing stops. The financial risk of "what if we produce this and visitors don't engage?" is gone, because there's nothing produced sitting on the shelf.
Together these change the question. It used to be "can we justify $35,000 in fixed cost for a three-month show?" — a question with a defensible no. Now it's "do we want our visitors to have guided interpretation of this exhibition?" — a question with a much more obvious yes.
Setup time matches. A focused show with 40 to 60 objects can have a working guide in under a week, often in a couple of days if the exhibition copy is already drafted. We have a longer playbook on this in launching a museum audio guide in 30 days; for a temporary show specifically, you don't need 30. You need maybe 5 to 10 working days of focused effort from someone who knows the content.
Where temporary exhibitions actually have the strongest ROI case
Here's the part that's underappreciated: temporary exhibitions aren't just newly viable for audio guides. They're often where guides earn the most.
Paid shows with separate ticketing. When a temporary exhibition has its own ticket price — $18, $25, $30 — visitors are already in a value-evaluation mindset. "Includes audio guide" reframes the ticket from "is this exhibition worth $25?" to "is this exhibition with a personal guide worth $25?" That perceived value shift consistently improves conversion. We've seen it move the needle hardest on dense, scholarly subject matter where visitors quietly worry they won't "get it" without context.
Marketed-hard blockbusters. When a show has a six-figure marketing campaign behind it, you're attracting visitors who don't otherwise come to the museum. Many of them aren't members. Many haven't been since school. They're showing up because the press told them this was important. The worst outcome is for them to walk through nothing-but-wall-labels and conclude the experience didn't match the hype. A guide is the cheapest insurance you can buy on a $300,000 marketing investment.
International press and travel coverage. Blockbusters draw international visitors, sometimes 30% of the audience or more. Multilingual support that took a separate $8,000 translation budget per language under the old model now comes natively in the platform. A show that gets covered in Le Monde, Frankfurter Allgemeine, and El País will have French, German, and Spanish visitors in the room next month. Having the guide already speak their language without a separate procurement is the kind of thing that converts a one-time visitor into a member.
Exhibitions with complex subject matter. Archaeology shows. Anything pre-modern. Exhibitions that span 400 years of context. Contemporary art that needs unpacking. The denser the material, the bigger the gap between what wall labels can carry and what visitors actually need. Audio fills that gap better than any other medium because visitors can listen while looking, instead of looking down at a label and then back up at the work.
Shows likely to drive memberships. Audio guides are one of the few interventions that increase reported satisfaction across nearly every visitor segment. Higher satisfaction at the temporary exhibition correlates with conversion to membership during the post-visit window. If your blockbuster is doing the heavy lifting on member acquisition this year, every percentage point of satisfaction lift compounds.
The pattern across all of these: the more the exhibition matters strategically, the better the audio guide ROI. Which is the inverse of what the old production model rewarded.
Where it still doesn't make sense
I want to be honest about the cases where I'd tell a director to skip it.
In-house staff-curated micro-shows in a side room. A dozen objects, free with admission, three weeks long, no marketing budget. The audience is members and walk-ins who are already deep in the institution. They'll read the wall texts. The handout is enough. Spend the curatorial energy on the next thing.
Very short pop-ups. Anything under three weeks. Even with same-week setup, the operational overhead — staff briefings, signage, monitoring, on-site QA — eats most of the value. Pop-ups want to be light on infrastructure. Add a QR code linking to the exhibition page if you want digital pickup, but a full guide is overkill.
Exhibitions that are themselves audio-led. Sound art. Listening rooms. Anything where the artist's intent involves the visitor's auditory attention being on the work itself. Layering a guide on top here actively damages the experience. If the show has a spoken element baked into the curatorial concept, trust it.
Shows with a strong exhibition catalogue that visitors heavily use. Some exhibitions, particularly scholarly ones, ship with a 200-page catalogue that becomes the authoritative reference. If your historical track record shows visitors actually buying and reading these (not just collecting them), the catalogue is doing the interpretive work. A guide can still complement, but the urgency drops. The case becomes "nice to have" rather than "leaves money on the table."
No internal owner. This isn't about the show; it's about your team. If nobody on staff is going to own the guide content — checking pronunciation of artist names, drafting per-stop instructions for the highlights, monitoring usage in week one and adjusting — don't ship it. A guide that nobody owns will be mediocre, and a mediocre guide on a high-profile show is worse than no guide at all.
Outside of these cases, my honest answer is to do it. The economics work. The timeline works. The visitor experience case is strong on its own.
How to ship a guide in time for opening
Assuming you're convinced and the show opens in eight to twelve weeks, here's the realistic path.
Now (8-12 weeks out): Decide who owns the guide. One person, ideally from curatorial or interpretation. Loop in your audio guide provider. Share the working checklist even if it's not final. Confirm the budget line is per-session usage, not a fixed production fee, and get sign-off on the projected total based on your expected attendance.
6 weeks out: Draft the per-object texts if they don't already exist. For most temporary shows, this is happening anyway for the catalogue and wall labels — the audio guide can pull from the same source material. Decide the persona: is the guide a curator, an enthusiastic colleague, a docent? Pick one and write a short style guide. This takes an afternoon.
4 weeks out: Load the data into the platform. Configure the tour sequence. Test the floor plan against the install drawings. Listen through the highlights in the languages you actually expect — German if it's a German artist, Mandarin if you've been getting Asian press coverage. Adjust pronunciations and emphasis where the generation needs help.
2 weeks out: Final install begins, and reality intrudes. Loans arrive late, pieces shift to different walls, one work doesn't clear conservation. Update the spatial layer in the system. None of this requires re-recording anything. It's configuration.
Opening week: Brief front-of-house. Add signage at the entrance and on the ticket. Promote the guide on the exhibition page and the confirmation email. Watch the usage data from day three onward. If a particular stop is generating long dwell times or repeated questions, add per-stop depth that week. The guide gets better while the show is open, which is the inverse of the old model where it got worse.
The only thing on this timeline that's hard is owning the content. Everything else is mechanics.
The shift to actually internalize
Most of the audio guide industry — and most of the institutional habits around it — was built around the constraints of recorded production. Long timelines. High fixed costs. Inflexible content. Those constraints shaped which shows got guides (permanent collections) and which didn't (everything temporary).
Those constraints are gone. The institutional habits will take longer to catch up than the technology did. But the practical question for your next 12-week show is straightforward: you have a guide option that costs nothing if visitors don't use it, ships in days, speaks 40+ languages, and matches the strategic ambition of the programming.
Platforms like Musa are built for exactly this — per-interaction or revenue-share pricing with same-week turnaround, so a 12-week show carries no fixed production bill and the guide only costs the museum when a visitor actually uses it. For a touring exhibition or a blockbuster with paid upsell, that flips the guide from a line item the finance team argues about into a margin contributor that scales with the thing the show was already designed to do: sell tickets and convert visitors.
If you're planning a paid show, a blockbuster, or anything with international press attached, we'd be glad to walk through the numbers on your specific exhibition.