How to create an audio guide for a museum

Creating an audio guide for a museum today takes weeks, not years, and you no longer need to buy dedicated devices. Most visitors already carry the tool to listen to it in their pocket: their own smartphone.

This guide sums up the good practice used by museums and providers in Europe and the United States in seven steps. It is written for directors, curators and education managers at small and mid-sized museums who want to offer an audio guide without a large institution's budget.

If you are mainly interested in writing the scripts, there is a dedicated deep dive in How to write an audio guide script.

Why an audio guide, even in a small museum

An audio guide lets the works speak when there are no staff to do it. For a museum with few people, it is often the only way to offer a "guided" visit every day, at any time.

Four benefits come up in every source we consulted:

• Independent visits at the visitor's own pace: visitors stop where they like, skip what doesn't interest them, listen again.
• More languages without more staff: a multilingual guide welcomes foreign tourists and tour groups, who often ask in advance whether translated content is available.
• Accessibility: audio, together with a written transcript, helps people with visual impairments, people who find reading difficult and those who simply prefer to listen.
• One more communication channel: the guide can mention events, temporary exhibitions, the shop or the museum's membership scheme.

There is one more benefit that matters above all to small museums: the technology has changed. With smartphone audio guides started from a QR code, there are no devices to buy, charge, clean and look after.

Step 1: set goals, audience and budget

Before thinking about technology, decide what the audio guide should achieve. The useful questions are few and concrete:

• Why do you want it? To welcome foreign tourists, make the museum more accessible, tell the collection's story better, offer an activity for families, or bring in some revenue.
• Who will use it? Families, enthusiasts, school groups, tourists, visitors with disabilities. Each audience needs a different language and length.
• What experience do you want to offer? A route through the highlights only, a complete visit, or layered content (short for everyone, a deep dive for those who want more).
• Who will look after it once it is live? You need someone in-house to update the content when a work is moved or a temporary exhibition arrives.
• How much can you spend, and how? A one-off purchase, a subscription or pay-as-you-go.

A practical tip: look at the audio guides of museums similar to yours and listen to a few tracks. You will quickly see what you like and what you want to avoid.

If several people are involved in the decision (management, the local authority, the communications office), share these points in writing before going further. For larger projects that go through a tender, this document becomes the basis of the specification.

Step 2: design the route and choose the stops

You don't need to cover everything: an effective audio guide selects. The sources agree on a few useful orders of magnitude to start with:

• 10 to 20 stops for a small or mid-sized museum. You can even start with 10 highlights and add more later.
• Around 15-20 stops add up to half an hour of listening in total.
• The overall length should stay around half of the average visit time: if visitors stay an hour, aim for 30 minutes of audio.

To choose the stops, start from three questions: which works visitors ask the staff about most, which best express the museum's identity, and which have a surprising story worth hearing.

Then arrange the stops along the visitor's physical route:

1. A welcome track that explains how to use the guide and how long it lasts.
2. Short introductory tracks for each room or section, giving context before the works.
3. The stops on individual works, which build the story step by step without repeating the labels.
4. A closing track that picks up the main themes and invites visitors to continue (temporary exhibition, shop, website).

Large museums can create several themed routes, for example a short one through the highlights and one for families. In small museums a single, well-built route is almost always the best choice.

Step 3: prepare the content and write the scripts

You already have the source material: catalogue entries, labels, wall panels, curators' notes. It shouldn't be read aloud as it is, but it is the base to start from.

The key principles, in brief:

• Write for the ear, not the eye. Short sentences, one idea per stop, spoken language.
• Short tracks. Usually between 1 and 2 minutes; rarely over 3.
• Tell stories. Anecdotes, people and questions stick far better than dates.
• Guide the eye. Say where to look: "bottom left", "along the edge of the frame".
• Add authentic voices. Short contributions from the director, the curator or a conservator add credibility.
• Consider several levels: a short version for everyone, optional deep dives and, if you get many families, a version for children.

Have the scripts checked by someone who knows the works well for accuracy, then by someone outside the museum for clarity.

Writing the scripts is what decides whether the audio guide gets listened to all the way through. That's why we gave it an article of its own, with examples and a checklist: How to write an audio guide script.

Step 4: choose the voice and record

Today there are three routes, and the choice depends on budget, timing and how often you will update:

• Professional voice (narrator in a studio): high cost, weeks of lead time. The best quality and expressiveness, but every change needs a new session.
• In-house voice (curator, guide, staff): low cost, days of lead time. Authentic and builds a bond with the museum, but audio quality needs careful attention.
• AI synthetic voice: very low cost, hours of lead time. Instant updates and many languages, but you need to check how it pronounces proper names and technical terms.

Many museums combine them: a synthetic voice for the main narration and short contributions recorded by the curator or the director.

If you record in-house, a few precautions are enough:

• Pick a quiet, "soft" room: rugs, curtains and sofas reduce echo. Avoid galleries with hard walls and floors.
• Even a smartphone is fine to start with. A lapel microphone and a recorder raise the quality a lot for a modest outlay.
• If you make a mistake, pause for a second and repeat the sentence: the cuts come later, with free software such as Audacity.
• In editing, remove mistakes, long pauses and filler words. Use music and ambient sound sparingly.
• Export each track as an MP3 and give the files numbered names (01-welcome, 02-room-1...), so they are easy to upload.

A quick alternative, useful if the museum already runs guided tours: record a tour given by an experienced guide, with a few colleagues as the audience, then split it into tracks. In that case the guide must give very precise directions, for example "with your back to the lift, walk right towards the large statue".

Step 5: think about languages and accessibility from the start

Languages can be added gradually. A sustainable strategy:

• Start with your local language plus the two or three languages most spoken by your foreign visitors (usually English, then German, French or Spanish depending on the area).
• Begin with the highlights route in every language, then add more stops.
• Add further languages based on data, that is, on the ones visitors actually choose.

On cost, professional translation is paid by the word and, multiplied by each language, can become the biggest item in the project. Machine translation and synthetic voices have lowered that threshold a lot; for the main languages a human review is still worth it. The experience we gathered shows that foreign visitors prefer an imperfect translation to no content in their language at all.

The real challenge is not translating, but keeping the languages in step when a text changes. Keep a single reference document for each stop and update every version together.

For accessibility, plan these as standard:

• A written transcript of every track, for people who are hard of hearing or prefer to read.
• Clear pace and diction, suited to crowded rooms.
• Simple controls and readable text on the phone screen.
• Detailed descriptions of the works for blind or partially sighted people, if your audience needs them.

Step 6: choose the technology

For most small and mid-sized museums, the simplest solution today is a smartphone web app started from a QR code. Here are the options side by side:

• Web app with QR codes: the visitor scans a QR code and listens in the browser, with nothing to install. Suits small and mid-sized museums and outdoor sites. Watch out for internet coverage in the galleries.
• Native app to download: installed from the app stores and works offline too. Suits large museums or museum networks with a lot of content. Watch out: many visitors give up on the download.
• Dedicated rental devices: the museum hands out a device at the entrance. Suits museums with large visitor flows and front-desk staff. Watch out for charging, hygiene, theft and maintenance.
• DIY MP3 files with QR codes: each QR code leads to an audio file uploaded online. Suits minimal budgets and pilot projects. Watch out: no statistics and awkward management.
• Automatic triggering (beacons or GPS): content starts when the visitor comes close. Suits large spaces, parks and outdoor trails. Watch out for the cost and upkeep of the sensors.

Whatever you choose, check three things: that visitors can start listening within seconds, with no sign-up or password; that your staff can edit texts and audio themselves; and that the platform provides usage statistics.

For museums with large numbers there are also concession models, in which the provider invests in the service and shares the rental income with the museum. They make sense above a few hundred thousand paying visitors a year, so rarely for a small museum.

Step 7: test, launch, promote and measure

Test before opening to the public. Have two or three people who don't know the museum walk through the guide on their own phones. Watch where they get stuck: a QR code that is hard to find, an unclear direction, a track that is too long. Listen to everything again to check how proper names and technical terms are pronounced.

Make the audio guide visible. Many audio guides go unused because nobody knows they exist. You need:

• a clearly visible sign at the entrance and at the ticket desk, with the QR code to start;
• a recognisable symbol next to every work with an audio track;
• QR codes placed where visitors naturally stop, at eye level;
• one sentence from the ticket desk staff: "Did you know we have a free audio guide on your phone?".

Promote it outside the museum too: website, Google Business Profile, social media, newsletter, tourist offices and operators who bring groups.

Decide how to offer it. It can be included in the ticket or sold separately. The first option maximises use; the second brings in revenue. For small museums, an included audio guide is often one more reason for people to visit.

Measure and improve. A few indicators are enough to make decisions: the most-played stops, the point where visitors stop listening, the most chosen languages, the busiest days and times. Add a simple way to leave feedback. A monthly review, owned by one named person, is enough to keep the guide up to date.

How much it costs and how long it takes

The cost of an audio guide depends mainly on four items, more than on the technology:

1. Scripts: written in-house or by a professional.
2. Voice: professional narrator, in-house voice or speech synthesis.
3. Languages: each extra language multiplies translation and recording costs.
4. Platform: a purchase, an annual subscription or usage-based pricing. With smartphone solutions the cost of devices disappears.

Indicative timings for a small museum with 10-20 stops:

• Goals, route and choice of stops: 1 week.
• Writing and reviewing the scripts: 2-4 weeks.
• Recording or generating the voice: from 1 day (synthetic voice) to 2 weeks (narrator).
• Translations: 1-3 weeks, in parallel.
• Upload, QR codes, signage and testing: 1 week.

In total, from one to two and a half months. The slowest phase is almost always approving the scripts, especially when several departments are involved: agree from the start who has the final say.

To get an idea of the costs with Museoo, you will find plans and commissions on the Pricing page, along with a simulator that estimates your revenue if you choose to sell access to the audio guide.

Common mistakes to avoid

• Trying to cover everything: too many stops and long tracks make people stop listening.
• Reading the labels aloud: the audio should add something, not repeat what the visitor can read.
• Asking an occasional audience to download an app: every extra step loses users.
• Not signposting the audio guide: without visible signs and QR codes, hardly anyone uses it.
• No one in charge after launch: within a few months the guide no longer matches the display.
• Neglecting the pronunciation of artists' names, places and technical terms, especially with synthetic voices.

Frequently asked questions

How many stops should an audio guide have?

For a small museum, between 10 and 20. It is better to start with the highlights and expand later.

How long should each track be?

Usually between 1 and 2 minutes, rarely more than 3. The total length should stay around half of the average visit time.

Can I create an audio guide myself?

Yes: with a smartphone, free editing software and a web platform you can produce a first version. A specialist provider shortens the timeline and ensures quality and maintenance.

Human voice or synthetic voice?

It depends on budget and updates. A synthetic voice is affordable and can be changed in minutes; a human voice is more expressive. Combining them is often the best option.

Do I need an app?

For most museums, no. A web app started from a QR code works on any smartphone with nothing to install.