An audio state of mind: Understanding behaviour around audio guides and visitor media

Shelley Mannion, The British Museum, UK, Amalia Sabiescu, RMIT Europe, Spain, William Robinson, British Museum, UK


It is a transitional moment for one of the British Museum's flagship products: the permanent collection audio guide. Nearly 160,000 visitors take the guide each year. It provides multimedia commentaries for 220 objects on display in museum galleries. In late 2014, the Museum embarked on a project to redesign the guide. This coincided with the launch of a new digital strategy focused on user-centred design. In a six-week Agile project, a cross-disciplinary team interviewed, observed, and tested paper prototypes with more than 250 visitors. This paper highlights the results, which include six key factors that influence take-up rate of audio guides and common patterns of behaviour in their use. It is of interest to anyone reflecting on the role of the traditional audio or multimedia guide alongside smartphones and mobile apps.

Keywords: audio guide, multimedia guide, museum, mobile device, research, strategy

1. Introduction

Thirty years ago, when curator Ian Jenkins recorded the British Museum’s first-ever audio guide for the Parthenon sculptures, it would have been difficult to predict where it would be today. That original guide was delivered on cassette players rented to visitors for fifty pence at a desk refashioned from the director’s discarded dining table. Visitors slung heavy players over their shoulders and flipped the tape to hear the full commentary. In 2015, Ian returned to the studio to record a guide to an exhibition that focuses on the same Parthenon sculptures in a new narrative. There are no clunky players and no cassettes. This guide will be offered on sleek Android smartphones that fit neatly in visitors’ pockets and comfortably hold hours of commentary.


Figure 1: British Museum audio-guide users in the Great Court and in the galleries. Images

The rapid evolution of mobile technologies since the introduction of the iPhone in 2007 has brought about an existential crisis for the conventional audio guide. At the same time, it has opened new opportunities. In 2014, Korean Air renewed its sponsorship of the British Museum’s multimedia guide, a successful product that launched in 2009. The sponsorship offers the chance to rethink and redesign the guide from the ground up, and, in turn has facilitated this research. A six-week agile project kick-started the process by examining two important questions:

  1. How do visitors use tools like the audio guide?
    Many Museum colleagues told us, “I am not an audio-guide type of person; I never take guides.” We wanted to know whether visitors felt that way, too, and if so, why. If visitors could be grouped according to their preferences for different products (e.g., audio guides, maps, guidebooks), then we could investigate the characteristics of those groups to make better and more marketable products. This topic broadened the scope of the research: we were looking at all visitors, not just audio guide users, and trying to understand their behaviour.
  2. Can we increase take-up rate?
    The current audio guide is used by around 3 percent of British Museum visitors. At an institution that receives nearly seven million visitors per year, this means nearly 160,000 people use the guide. We want to do better. If take-up could be improved, how large an increase could we reasonably aim for? In order to set achievable targets, we need to understand what influences visitors’ decision to take an audio guide.

2. Team and method

A team was assembled to carry out the project. It consisted of one full-time staff member who was freed from some daily responsibilities to focus on the project, three freelancers (a creative technologist, ethnographic researcher, and digital learning specialist), and an intern. We recruited native-language speakers from Visitor Services staff and a pool of family facilitators to ensure coverage of two of our most popular foreign languages: Chinese and Spanish. Morning scrums were held to set the goals for each day, and progress was tracked visually by a large paper calendar and sticky notes on the walls of the project room.

We tackled the research questions through several paths:

  • We consulted colleagues from other organisations who worked with audio and multimedia guides.
  • We reviewed existing data. We mined earlier evaluations for useful insights and aggregated historical data. We used Data Hero ( to create a dashboard for visualising the results.
  • We conducted surveys and short interviews with visitors.
  • We performed several rounds of structured observations of audio-guide users, pausing after each round to analyse, discuss findings, and refine our categories.
  • We developed personas that informed the design of a paper prototype and tested it with visitors.

3. What did we discover?

Meaningful segmentation

We wanted to investigate the assumption that there is such a thing as an “audio-guide type.” More broadly, we wanted to know whether visitors divided along lines according to product type: guidebook users, map readers, audio-guide users, etc. Thinking about how to meaningfully group visitors to understand their behaviour brought us into the world of marketing segmentation.

The Culture Segments (, and marketing segmentation in general, are problematic for product design. Segments overgeneralise, creating broad categories of faceless visitor types, rather than an inventory of individuals with specific needs. They have no sense of a user journey that covers time and distance, during which the visitor enacts a whole range of emotions and motivations. John Falk (2009) writes compellingly about identity and the museum experience, pointing out that people are not defined by a single category, but move across them both during and between visits. This personalised, flexible view of visitors as humans whose motivations and identities are not fixed or stable was the starting point for our own research.

As we amassed data from surveys, interviews, and observations, we tried to sort respondents into groups based on their media usage. Our initial categories reflected strategies and tools that people used to navigate the Museum:

  • Wandering around
  • Label and panel readers
  • Guidebook users
  • Map users
  • Audio-guide users

The categories above were helpful, but eventually we set them aside in favour of the Visitor Attribute Scales described below. The Attributes proved more useful in representing the complexity of how visitors perform identity and shape-shift during their visit. It was a useful lesson that segmentation has limited applicability in the design of products.

An outlier

Qualitative research often yields powerful insights through outliers. A Russian visitor employed an ingenious strategy. He knew little of the Museum, but wanted to orient himself. He ignored all the visitor media products on offer—printed maps, map plinths, guidebooks, information points, audio guides—and headed straight for the bookshop. He browsed the covers of the books for things that caught his interest and then asked where he could find them. The bookshop is the only place in the entire Museum that gives a visual overview of the collection at a glance. Books are arranged by culture area—just like the galleries—so if the books on ancient Roman sculptures, clocks, or North American canoes look interesting, it is fair to assume something like them are on display in the galleries. No one else used this strategy, so we do not have a category for “bookshop browsers”—but it got us thinking about the advantages of this approach over the map plinths, which can only show a handful of highlighted objects.

Figure 2: British Museum guidebooks in the museum bookshop

Figure 2: British Museum guidebooks in the bookshop

What influences take-up rate?

Across the sector, a take-up rate of around 3 percent for permanent-collection audio guides is standard. Take-up for the British Museum’s guide is in line with this. To improve this, we needed to identify the factors that influence visitors’ decisions to take audio guides. Our initial solution to this problem was to develop a typology of visitors according to their behaviour, but as things progressed, we hit upon a different method. It was inspired by what people said about their strategies for navigating the Museum. When we paid attention to this, patterns gradually emerged in visitors’ approaches to wayfinding. The vocabulary interviewees used to describe these strategies was not always the same, but the ideas behind it began to leap out at us. Set on scales to reflect how people move along a continuum between their two poles, these became Visitor Attribute Scales. The Attributes are Time, Confidence, Authority, Tools, Movement, and Aim.

Figure 3: Visitor Attribute Scales: Key factors that influence take up of audio guides

Figure 3: Visitor Attribute Scales: key factors that influence take-up of audio guides


Time, or more accurately perception of time, proved the most important factor in whether people took an audio guide. Most visitors imagined that the audio guide would take a long time or that they would have to invest lots of time to get their money’s worth. What visitors meant by “a long time” varied greatly. Here is an exchange with an American tourist:

Interviewer: Will you take an audio guide?

Visitor: No, I don’t have enough time.

Interviewer: How much time do you have?

Visitor: All day.

In reality, even as little even as an hour with the guide would be helpful and productive for many people, but the assumption that the time they had was not enough was pervasive. The two ends of the Time Attribute Scale signify whether visitors believe the time they have is sufficient to achieve their goals. In general, visitors had between thirty minutes and six hours to spend. The majority of visitors in our study had about three hours for their visits, which is ample time to use the audio guide. Even then, many thought it was not enough:

Reading and looking is better. It’s distracting to have someone talking at me the whole time. We only have three hours, which is not enough time. (85)


Confidence indicates whether a visitor believes he or she can successfully explore the Museum unaided and the extent to which he or she feels in control of the visit experience. A high level of confidence comes from quantifiable factors such as prior knowledge of the Museum or the topics represented in the collection, but also from the (sometimes inaccurate) perception that the Museum provides enough information on panels and labels to support them. Language plays an important role here, for example, English-speaking visitors are dismissive of the audio guide or other products because they are convinced that they can get whatever information they want by reading.

Interviewer: Do you regularly take audio guides?
Visitor: Sometimes. If I’m in a foreign country, then it’s a good way to find out what’s going on. But I prefer to read captions on artworks when they are in our language. (9)

Audio guides are not worth paying for as galleries and labels have enough detail. (71)

False confidence can be manifested by visitors who have no background knowledge and poor language skills, yet regard the visit as an event that does not require particular expertise. This is often true of tourists engaged in a box-ticking exercise of London’s top attractions: simply visiting the Museum building and seeing its two or three most-famous objects is not seen as requiring assistance. They are confident of their ability to “do” the Museum without help from any guide or product.


Authority refers to the desire to retain or willingness to relinquish the power to decide what is important to see and how to see it. Some visitors are happy to be guided, but others prefer to determine their own paths. Several crucial decisions are at play in this tug-of-war between institutional authority and self-determination:

  • What not to miss
  • How to get the most out of the visit
  • Which objects, collections, galleries, and spaces are worth visiting
  • What contextual information is best for making sense of them
  • The best routes for exploration, discovery, or just getting around

Visitors who favour self-authority generally have high confidence:

I don’t feel like listening to somebody else talking. I think I already know quite a lot. (111)

I like to explore on my own, decide on my own what’s interesting and not have someone tell me what is interesting. (88)

I like to do it myself instead of having someone say “go here.” (91)

If someone is willing to be guided, whom they allow to guide them is an important question. While many view the Museum as a trustworthy external source, others are less certain and put their trust in guidebook publishers or social media sites that they view as objective. A savvy thirty-something visitor from Germany explained that he relied much more on TripAdvisor reviews to decide what to see, rather than anything the Museum published about itself.

Non-native language speakers have little choice but to invest in external authority and often carry guidebooks in their own language. Unfortunately, these publications contain inaccuracies. For example, Lonely Planet London (2014) says the British Museum has “multimedia iPad tours.” We do not. Eyewitness Travel London (Leapman, 2013) suggests the Reading Room is a space not to miss. It has been closed to the public for two years with no plans to reopen. Regardless, the crucial point is that visitors determined to make their own decisions are unlikely to take an audio guide unless the value proposition is extremely compelling or validation comes from others on social media.


This scale refers to the likelihood that visitors will use materials for assistance. This includes Museum products like audio guides, maps, and guidebooks, as well as tools visitors bring themselves. There is a correlation between Confidence, Authority, and Tools, meaning that visitors who are less confident (Confidence: Low) and willing to surrender decision making to external authority (Authority: External) are more likely to carry tools. That said, we observed all types of visitors using multiple mobile devices, especially phones and tablets, which they primarily used for photography. A surprising number of phone-carrying visitors photographed the floor plan in the Museum’s Great Court. They consulted this digital map on their own device as they moved around the Museum. A small but opinionated group of visitors do not use tools at all. They want nothing to interfere with their emotional or intellectual experience. Visitors who want an intellectual encounter speak disparagingly of audio guides and other products as things used by people less intelligent and less diligent than themselves. Emotionally inclined visitors want the freedom to explore where their interest takes them.

We use our brain and look at labels and panels. (137)

I never take any guides. I prefer to take it all in, absorb it. I like looking at people; I find it interesting. (110)


Movement refers to a visitor’s route through Museum spaces. Some visitors move in a structured, self-contained way if there are specific things they want to see. These visitors may plan ahead from home or use a map on arrival to plot the most direct course. They tend to be single-minded about getting to where they are going and do not get distracted by interesting things they encounter along the way. Audio-guide tours are one way of structuring movement, but they may not be followed conscientiously or without distraction. Activity trails for families are linear itineraries that are often followed in an unstructured way, depending on the age of the children, size of the group, and other factors that necessitate deviating from the prescribed route or stopping before the destination is reached. Many museum visits are unstructured from the beginning or change to become unstructured. Two American friends interviewed in the middle of their visit said that after they had found what they wanted to see, they turned to browsing. They exhibited a Movement pattern we saw repeatedly among tourists—a structured start followed by wandering around:

We knew the Rosetta Stone, so we’ve seen that and now we’re just browsing… Just browsing, kicking around. (146)

I have no idea, just take a look casually. Doing sightseeing. I heard of mummies and samurai. (86)

It is worth noting that unstructured movement does not imply an unsatisfying experience. On the contrary, visitors who wandered around had an enjoyable experience, and those who used both modes said they got exactly what they had hoped for from their visit.


Aim, or agenda, deals with visitors’ motivations. Marketing segmentation models delve into the content of motivations, but we found this an unhelpful distraction. The content of motivation was less meaningful than strength and specificity in predicting whether someone would take an audio guide. Strength is a visitor’s resolve to stick to a predetermined goal; specificity is how narrowly focused that goal is. Aims influenced the take-up of products in a similar way to Movement. If visitors had Strong and Specific aims, they did not consider taking an audio guide. If their aims shifted and became less specific, they were more open to taking one. The key point seems to be that an effective guide helps visitors accomplish their aims quickly and recognises when they change modes and responds accordingly. The Visitor Attribute Scales provide a framework for understanding what visitors are really thinking and doing. They became the basis for creating personas and scenarios (Smith, 2003), which determined the design of our paper prototype. They contain the potential to inform not just product features and user-interface design, but also messaging around audio guides and other products. If we know what visitors are interested in (e.g., a guide that helps them “do” the Museum in an hour), then we can market our products more effectively.

Why do people take audio guides?

The Attributes provide a framework for understanding what drives visitors to take guides. Time was the biggest factor. Visitors perceived they did not have enough time to take a guide. This was combined with misunderstandings about the guide itself: people assumed that the guide only contained tours and would force them to spend more time than they had.

Confidence was the most surprising influencer. Our assumption had been that Museum visitors lack confidence and actively look for help. As it turns out, visitors are far more confident of their ability to navigate and understand the collection than we imagined. This is true even when they do not have any scholarly background or knowledge of the Museum. A retired Australian couple explained that, while they would be likely to take a guide in an art museum, they had no need for one at the British Museum because it was self-explanatory. Further probing exposed that they knew almost nothing about the contents of the Museum, but they remained adamant that additional support was unnecessary.

Language ability is closely linked to confidence. Foreign visitors make up 60 percent of the British Museum’s on-site audience. Some of these visitors are comfortable with English, but many are not. They struggle to understand the written labels and panels that English speakers find helpful. This undermines their confidence to navigate the Museum on their own and makes them more likely to take an audio guide. That non-native speakers are more likely to take guides is perhaps obvious, but it is useful to see language as one of several factors that determine where a visitor is on the confidence scale. A scale acknowledges that language ability is not a Boolean; people may have some English ability, but it is the breadth of their vocabulary and familiarity that determines whether their skill is sufficient to achieve the goals of their visit.

Aim is another key influencer. Many first-time visitors know exactly what they want. They have heard of the most well-known objects and arrive determined to see them. Tourists may have equally strong and specific aims as scholars, but they rarely see an audio guide as being useful in fulfilling that aim.

Although the tourists’ aims may be strong and specific, our observations showed that they are fulfilled quickly. As opposed to scholars who may spend an entire day pursuing their research goals, tourists complete their box-ticking in an hour and then become aimless wanderers. This is a perfect example of how visitor motivations change during the course of a visit. It demonstrates why segmentation at a single point at the start of the journey is problematic. While the strength of their aim when they begin explains why visitors don’t bother with an audio guide, the fact that they become aimless halfway through their journey presents an opportunity. If the Museum can intervene at the moment when aims are becoming weaker and unfocused, visitors will be much more receptive to purchasing an audio guide.

Who are the audio-guide users?

With a better idea of how to increase take-up of the guide, we began to home in on the behaviour of guide users. As observations were completed, we developed a categorisation scheme by reading data sheets aloud to the group at the end of each day. Through this process, four patterns of behaviour emerged. The table below includes a description for each and an explanation of how they relate to tools and social interaction.

Behaviour pattern Description Tools Social interaction
Tour following. About 10% of audio-guide users followed tours. This was fewer than expected. A previous evaluation (British Museum, 2011) reported a higher proportion of tour takers (35%), although this was on asking visitors to self-report their behaviour, rather than direct observation. Listened to stops in linear order, rarely skipped stops, and tended to ignore other objects in their paths, no matter how interesting. Tour takers were a serious bunch and usually had plenty of time, at least three hours or more. Use of maps along with audio guide, but this group is not characterised by use of other devices. Desire to pursue personalised learning by following tours tends to override social agenda, yet if a tour taker is in a group, she may stop to discuss what she and her companions have seen and heard.
Code hunting. In our observations, 75% of visitors exhibited code-hunting behaviour. Two earlier evaluations (British Museum, 2010, 2011) confirmed the random access keypad was the most frequently used feature of the guide: 55% to 70% of people used it. Non-linear journey through the galleries driven by search for audio-guide icons. The visitors’ first action when arriving in a gallery is to see out objects with commentary. There are three variants of this behaviour: Selective: After having an overview, focus on a limited number of objects; Accidental: Choose one or more stops per gallery at random and usually move on after listening to one to two stops; Exhaustive: Go from stop to stop listening to content for most or all objects in a gallery before moving on. Tend not to use other tools such as maps or guidebooks, but rely heavily on the audio guide as main source of information and guidance. No strong social patterns observed with this group; however, as they make up the largest group overall, it is worth noting that the majority (at least 60%) of audio-guide users are accompanied by a companion or group.
Conscientious consumption. A minority (about 3%) of guide users in our observations. This group was not identified in previous studies, so no historical data is available. Focused and thorough exploration of stop objects with the audio guide and other material. Multiple modes of information acquisition: listening, reading, looking. May have a strategy that involves taking a general tour of the room, then homing in on objects of interest for in-depth exploration. Most likely of all four behaviour groups to carry additional tools or materials including maps, guidebooks, camera, etc. Also relies on information ecosystem in the gallery such as panels, labels and map plinths—even related objects. If more than one member of a group exhibit this behaviour, they may decide to split up to allow each to pursue their own interests. They reconvene later to exchange experiences. If is a pair where one partner is less engaged, they may stay together, but have little interaction.
Casual use. Around 15% to 20% of visitors in our observations. This group was not identified in previous studies, so no historical data is available. Sporadic and often disinterested use of the audio guide. Randomly select stops and may not listen all the way through. May give up after a few stops and not listen anymore. There are two variants: Light touch: Browse freely and look for things that attract them. Listen to stop if interest is sparked; Disinterested: May only pretend to listen when authority figure is around; otherwise avoid using it or gradually cease using it. Tend not to use museum-provided materials such as guidebooks, labels, and panels because information does not interest them. May be very involved with use of their own phones, tablets, or cameras. Almost always visiting in group. A common social dynamic within these groups is that one individual determines what the group should do (e.g., takes an audio guide, and insists that everyone use a guide). These are non-guide takers who have been given a guide by someone else.

Table 1: behaviour patterns of audio-guide users

4. Testing paper prototypes

The final stage of the project was to translate our findings into a paper prototype to test with visitors. Personas helped clarify what features we wanted to test. We created several personas inspired by common configurations of the Attributes. For our first round of testing—which is as far as we have reached at the time of writing—we selected three personas that were representative of the majority of visitors.

Persona name and description Visitor Attribute(s) Scenario
Annette is a 60-year-old retired school administrator from Melbourne, Australia. Frequent museumgoer who feels confident in museum environments, especially in English-speaking countries. Time: Enough, needs to “do” the Museum in three to four hours. Confidence: High, based on her ability to read panel and label text in English. Confident in both her own knowledge and in Museum’s written materials. Authority: Self. Prefers to decide on her own. In London for two weeks to visit friends. Her husband, a cycle fan, hopes to see the Tour de France. First time at the British Museum with her husband. They plan to spend half a day (three to four hours).
Leonard is a 45-year-old New Yorker who runs his own home repairs business. He is the father of two teenage children, ages fourteen and seventeen. He visits museums less than once a year at home; he prefers live sporting events. Aim: Strong and specific (to do highlights as suggested in guidebook). Movement: Starts specific, but becomes loose once he has seen what he intended to see. In London for one week on sightseeing trip with his family. It is the children’s first trip to Europe. Has read Rick Steves’s Guidebook and wants to see the Rosetta Stone and Parthenon sculptures.
Ming is a 30-year-old pharmacist from Shanghai, China. He is a Mandarin speaker who has studied English, but rarely has the chance to practise. He rarely visits museums, but loves his phone and uses it for texting and photos. Time: Not enough, only one to two hours between other sightseeing activities. Confidence: Medium to low, Struggles with written English, especially complex text on panels and labels. Tools: All. Likes to take audio guides and use other devices at the same time.  Three days in London as part of a European tour. He is visiting with an extended family group. It is their first time in the United Kingdom and at the Museum. They went to Westminster Abbey in the morning and used the audio guide. He wants to see the Chinese gallery.

Table 2: personas and scenarios for the paper prototype design

Creation of the paper prototype began with a discussion of these three personas. We knew that Time was a major consideration, not just for Annette and May, but for many other visitors, too. We wanted to address that in the interface of the guide. Leonard was also representative of many tourists who have a few highlight objects in mind that they want to see, but generally want to “do” the Museum. These visitors told us they were interested in the Museum building as well as its collection. In order to facilitate Annette’s preference for Self-Authority, we needed to provide an option to help her pursue her own interests. We designed a simple opening screen with four options:

  • About the Museum
  • What you should see/What not to miss
  • How much time do you have?
  • Search/I know what I want

We turned this into a paper prototype and took it to the floor. Visitors queuing to purchase the existing audio guide were invited to test the prototype. They chose an option on the opening screen and told us why they chose it. We then asked them what they expected to find behind each of the four options. Prototype testing revealed that Asian and Western visitors had different expectations about “What you should see.” Europeans and North Americans wanted to be directed to the Museum’s obvious stars—the Rosetta Stone and Egyptian mummies—but Asian visitors wanted first to see things from their own cultures. Accommodating cultural preferences in the audio guide is something we will continue to investigate as the project moves forward, especially for Chinese visitors as, from July 2014, Mandarin is the top-selling language.

Figure 4: Testing the audio guide paper prototype

Figure 4: testing the audio-guide paper prototype. Images

5. Lessons learned

The limits of segmentation in product design

The main outcomes of this project are the Visitor Attribute Scales, which offer a helpful model for product-driven research. It is a departure from conventional marketing segmentation, which focuses on the content of motivation rather than actual behaviour. The Attributes deal with what people do. They acknowledge and accommodate shifts in behaviour that occur during a visit, freeing designers to invent new products and features that serve genuine, but unrecognised, needs.

Key factors that determine audio-guide take-up

The Visitor Attribute Scales identify the primary factors that influence the take-up of audio guides. For British Museum visitors, the greatest influencers are Time, Confidence, and Authority. False confidence, in particular, is an interesting phenomenon. If confident visitors are reluctant to take a guide at the beginning of their journey, perhaps we can design a product that anticipates the moment when their confidence begins to flag. Tailoring products and the messaging around them to key influencers will increase take-up rate.

Letting go of value judgements

As museum professionals, we often work within an implied hierarchy of intentions—learning agendas are seen as more worthy than sightseeing or social ones—which makes it difficult to see visitors objectively. Moving away from value judgements about the substance of visitors’ aims and focusing on describing their behaviour means we can design products that speak to broader audiences and deliver on enjoyment and satisfaction, as well as education. If we do not understand what visitors are doing, we cannot make things they want to use.


Korean Air has generously sponsored the British Museum multimedia guide since 2009, and the renewal of their sponsorship in 2014 made it possible to carry out this project.


British Museum. (2010). Evaluation of the British Museum Multimedia Guide. Internal document.

British Museum. (2011). Evaluation of the British Museum Multimedia Guide. Internal document.

Falk, J. H. (2009). Identity and the museum visitor experience. Chicago: Left Coast Press.

Leapman, M. (2013). DK Eyewitness Travel Guide. London: Dorling Kindersley Limited.

Lonely Planet. (2014). Lonely Planet London. Lonely Planet Publications Ltd.

Smith, R. S. (2003). Using Scenarios to Gather Requirements. Last updated February 19, 2003. Consulted January 2015. Available

Cite as:
. "An audio state of mind: Understanding behaviour around audio guides and visitor media." MW2015: Museums and the Web 2015. Published February 1, 2015. Consulted .