A heritage director told us last year: "Visitors are walking around outdoors holding cameras, water bottles, and a toddler. They're squinting at the sun. We have no Wi-Fi, no power at most of the site, and the cell signal cuts out behind the chapel. Why would I spend money on an audio guide?"
The answer is a little uncomfortable for the indoor-museum world: outdoor sites are the strongest case for audio guides, not the weakest. A painting has a label next to it. A ruin doesn't. A field where 4,000 men died in 1644 looks like a field. The objects can't speak, the wall text doesn't exist, and the visitor showed up specifically to understand what they're looking at. That gap is exactly what audio is good at filling.
The infrastructure question — how you actually deliver audio at a site without Wi-Fi or power — is solved, and we've covered the technical side in detail here. This piece is about the prior question. Should you invest at all?
Why outdoor sites are the highest-ROI audio guide category
Indoor museums have alternatives. They have wall labels, room introductions, printed catalogues, docents at the door, and most of all they have objects that physically present themselves with clear edges and known provenance. A visitor in a gallery can absorb a lot just by looking carefully.
Outdoor heritage doesn't get that grace. A medieval castle ruin is half its original height. The great hall is a footprint in the grass. The chapel is two arches and a hint of foundation. Without interpretation, visitors see picturesque rubble. With it, they see where the lord ate, where the garrison slept, where the well was poisoned during the 1216 siege. The delta between "nice old stones" and "I now understand a piece of history" is enormous, and audio is the cheapest way to deliver it.
The same applies across outdoor categories. Archaeological sites where the excavation has been backfilled and only post-holes remain. Historic gardens where the planting changes by season and the design intent isn't visible from any single viewpoint. Battlefields. Industrial heritage with rusted machinery whose function isn't obvious. Coastal heritage with tide-dependent features. Open-air museums where the buildings are reconstructions of vernacular forms that look like ordinary barns. In all of these cases, the visual evidence under-delivers and context does the heavy lifting.
There's a second factor most people underestimate. Walking outdoors, visitors are receptive to audio in a way they aren't in a gallery. In a room of paintings, audio competes with the paintings for attention. Outdoors, you're walking from one stop to the next, you can look around freely, and your ears are doing nothing useful. The guide fills dead time rather than displacing visual focus. That's why outdoor visitors finish tours at higher rates than indoor ones at most sites we've looked at.
What used to make it impractical
Until recently, the case against outdoor audio guides was real and reasonable.
Hardware devices weren't built for weather. The chunky handsets museums hand out at the front desk get destroyed in rain. They get sand in them at coastal sites. They get dropped on stone floors. The replacement cost across a season is brutal, and the rental staff time is nontrivial — someone has to be at the entrance handing them out, taking them back, charging them overnight, sanitising them between visitors.
Connectivity was the second wall. Streaming-based guides assumed Wi-Fi or solid 4G. Most heritage sites in rural locations have neither. Beacon-based positioning systems assumed you could install and power Bluetooth beacons across the site, which works in a corridor and falls apart across ten acres of open ground exposed to weather.
Content production used to be specialist work too. Recording a professional voice actor in studio, paying a writer for the script, mastering and editing the audio — small heritage trusts couldn't justify that for a site with 30,000 annual visitors.
All three of those problems are now mostly gone. BYOD audio guides on the visitor's phone remove the hardware question entirely. Offline-capable architectures download the tour once at the entrance and run locally for the rest of the visit, signal or no signal. AI-generated narration with curatorial review has cut content production cost by something like an order of magnitude — a small site that couldn't afford a guide five years ago can now build one in a few weeks of curator time. We won't rehearse the technical detail here; the companion article goes into how the offline and GPS pieces actually work.
The honest position now is that the old reasons not to invest don't hold up. The remaining decision is editorial and curatorial.
Where it absolutely pays off
Some site types are so well-suited to audio that not having a guide feels like a missed obligation to the visitor.
Ruins where the structure is mostly gone. Roman villas, monastic ruins, Iron Age forts, abandoned mining settlements. The ground-truth archaeology might be excellent, but the visible structure is fragmentary. Audio is the only practical way to reconstruct what stood there, who used it, and how it functioned. A guided tour of a ruin without narration is a walk through some lumpy fields.
Battlefields and conflict sites. This is one of the strongest cases. A battlefield is geographically large, visually featureless, and emotionally significant. Visitors come because the place matters, but the place itself communicates almost nothing without help. A well-made audio guide can stand the visitor at a specific spot and tell them what happened there — the cavalry charge they're looking at the line of, the road the retreating army took, the farmhouse where the commander wrote his last letter. There are ethical considerations (more on this below) but they're solvable, not disqualifying.
Archaeological sites mid-excavation or post-backfill. Active digs are confusing without context. Backfilled sites where the trenches are gone are even worse. Audio gives the curator a way to point at the ground and say "underneath here, two metres down, was a kiln complex from the 4th century." Without that, the visitor sees grass.
Historic gardens with seasonal interest. A garden looks different in April and August. The design intent of an 18th-century landscape garden takes hours to read by walking. A guide that adapts to season, weather, and the visitor's pace is genuinely useful. Wall labels in gardens get weathered, vandalised, or removed for aesthetics. Audio sidesteps all of that.
Open-air museums and reconstructed villages. Reconstructions need framing. Without it, a visitor doesn't know if they're looking at a genuine 17th-century cottage or a sympathetic rebuild from 1962. Audio can signal that confidently in a way signage often dodges.
Multilingual heritage destinations. This is the biggest one for international tourism. Outdoor heritage sites — Pompeii, Stonehenge, Angkor, Petra, the Acropolis — pull visitors from dozens of countries. Printed signs in 12 languages aren't viable. Live guides in 12 languages aren't viable either. Audio is the only realistic way to give a Korean tourist at a Greek site the same depth of interpretation a Greek visitor gets. An AI audio guide that works in 30+ languages effectively closes a gap that's been open for the entire history of heritage tourism.
Where it doesn't
Honesty matters here, because over-claiming is how the audio guide industry burned its credibility in the first place.
Single-marker monuments. A standalone war memorial in a village square, a single standing stone, a plaque on a wall — these don't need a tour. They need a paragraph. A QR code linking to a one-page web article is fine. Building a multi-stop audio tour for a single object is overkill, and visitors won't engage with it.
Sites where live guiding is the product. Some heritage operations are built around their human guides. National Trust properties with knowledgeable volunteers who run hourly tours, archaeological sites with university-trained guides who lead small groups for two hours, battlefield sites where retired soldiers provide commentary that's part of the draw. If your live guide programme is well-staffed, runs in the languages your visitors actually speak, and visitors rate it highly, an audio guide is at most a supplement for off-hours and untranslated languages. Pushing audio into a site whose identity is human-led can dilute what makes the visit special.
Sites with genuine curatorial sensitivity that resists narration. Some places shouldn't be talked over. Holocaust memorial sites, the more solemn battlefields, sites of recent violence or ongoing trauma — there's a real argument for silence. This isn't an absolute rule. Some such sites have done audio extremely well, using primary sources, survivor testimony, and silence between segments to remarkable effect. But it requires an editorial sensibility that goes beyond standard tour-writing, and getting it wrong is worse than not doing it.
Tiny sites with strong existing signage. A single-room interpretation centre with good panels, a 200-metre heritage trail with five well-written signs — adding audio doesn't move the needle much. The marginal cost isn't justified by the marginal improvement.
If your site falls into one of those buckets, save the budget. Spend it on signage, on staff training, or on something else entirely. The point of this piece isn't that every outdoor site needs an audio guide. It's that the default assumption — that outdoor sites are too hard to do audio for — is no longer true.
What to look for if you decide to invest
Assuming the case lands for your site, the procurement question is what to actually buy. A few specifics worth pushing on.
- BYOD with offline support, not hardware rental. This isn't a preference, it's the only sustainable model for an outdoor site. Hardware loses to weather and theft.
- Genuine multilingual coverage. Not "we support Spanish" meaning a one-time human translation of a 2018 script. AI-driven translation that updates when your content updates, in 20+ languages, ideally with regional variants where it matters.
- Curatorial control, not a black-box AI. You should be able to write the canonical narrative and have the system stay faithful to it. Pure generative output without source-of-truth content is risky for heritage. The right model is a knowledge graph the curator owns, with AI doing the linguistic work on top.
- Stops the visitor can skip, return to, and rearrange. Linear tours are a 1990s constraint. Outdoor visitors wander.
- A way to handle ambient noise. Voice work and audio mastering matter more outdoors than indoors. If the demo sounds fine in your office but disappears under wind on site, it's not built for you.
- Honest performance claims about offline mode. Ask what works without signal and what doesn't. Curated tour stops should work. Open-ended Q&A may not. Anyone claiming everything works offline is either confused or selling you something.
If you're operating temporary or travelling exhibitions adjacent to a permanent outdoor site — a touring archaeology display in a marquee for a season, say — the considerations overlap with what we've written about temporary exhibitions. For sites that are specifically archaeological in character, there's also a more focused take in our archaeological sites guide.
A short word on the worry the director started with
Visitors really are walking around outdoors holding cameras and water bottles and toddlers. They really do squint at the sun. The signal really does drop behind the chapel. None of that is fictional, and a vendor who tells you those things don't matter is a vendor not to trust.
What's changed is that those problems are now design constraints, not deal-breakers. Phones handle the heat and the rain better than rented hardware did. Offline-first architectures handle the dead zones. Content can be written for outdoor pacing rather than gallery pacing. The visitor with the toddler will dip in and out of the guide rather than listen straight through, and that's fine — the guide should be built for that.
If you run an outdoor heritage site and you've been putting off the audio guide question because it felt impractical, it's worth a fresh look. Platforms like Musa price on revenue share or per-interaction with no hardware to rent, which is a much better fit for a seasonal outdoor site than the old capex model ever was — the site only pays for visitors who actually engage, and the guide goes from a fixed-cost gamble against weather and footfall to a direct line of margin on the visits that happen. We're happy to talk through what it would take for your specific site — get in touch and we can sketch out whether it's a fit.