Using audience insights to improve the Science Museum Group collection stories online

Emily Fildes
Science Museum Group Digital Lab
9 min readMar 23, 2020

--

Online storytelling in museums is a growing area of interest for the cultural heritage sector. Organisations are looking beyond digitising their collection to using tools to bring those collections to life, and engage audiences beyond the museum walls.

Storytelling means different things to different people, but for the Science Museum Group it was originally outlined in our 2015–17 Digital Strategy as “narrative content” — a way to tell the stories behind the objects in our collection. In our most recent strategy (2018–21), it was included as part of the objective to increase our online audience reach.

In practical terms, our stories aim to contextualise our diverse collection, which ranges from locomotives to artworks, from images of electrical engineering to computing, from industrial chemistry to domestic appliances, and showcase the many meanings and interpretations behind these representations of science and technology.

This post will describe how we used audience insights to shape and improve this content, and our initial findings on the impact it had.

What was our starting point?

When embarking on delivering on these new stories, we had the potential to build on, learn from and counter existing story content that — particularly the Science Museum — had delivered in the past.

These ranged from a standalone website built in 2004 which outlined some key developments in science and invention that led to a modern industrialised society; a standalone website for students on the history of medicine; and a series of online stories drawn from the content of the permanent gallery Information Age at the Science Museum.

In addition, we tested a content approach — a more esoteric form of collection stories — as part of the Mathematics: The Winton Gallery in 2016.

Some of this content drove high traffic, some very limited, some we had no data on, but they did provide us with an initial starting point to carry out some audience research.

How we approached our audience research?

Over the course of two years we gradually built up and refined the understanding of our online audiences.

We used our legacy content as a starting point to understand our online audiences, discover which audience we wanted to focus on and hone our content to meet their needs.

We used a mixture of different analytical techniques including quantitative analysis using Google Analytics, qualitative analysis using online surveys and focus groups, and usability testing. This was primarily carried out on the Science Museum website.

We were also able to establish some benchmark survey questions which will enable us to continue to monitor and evaluate our content.

What we found out about our audience?

Whilst within our organisation we have a strong distinction between a collection object record: content entered via our collection catalogue, meeting SPECTRUM standards and published onto our collection website; and a story: interpretative content which uses text, image or film to make meaning or connections between one or more objects, actually to our audiences they were interchangeable.

This means there is a very strong overlap to the online audiences for our collection.

Therefore we see similar trends in terms of users coming with a strong desire to look for something specific, rather than to browse our site.

A graph showing the reasons visitors came to our story content. 80% of users were looking for specific information
Source: Frankly, Green + Webb Survey 2016, Q. Why are you visiting our website today?

We also found that users — in the main — had some connection to science and technology; the connections ranged from students to those with a personal interest in the subjects.

A bar graph showing what interest users’ had in science and technology. 45% were studying it; 35% had a personal interest
Source: Frankly, Green + Webb Survey 2018, Q. Do you have a particular interest in science or technology?

The diversity of the audience was also shown by the breadth of ages coming to the content.

A combined column graph showing the diversity of ages visiting our content. The ages range from 18 to 65+
Source: Frankly, Green + Webb Survey 2018

We also tested their satisfaction with the content offering across three key areas: trustworthiness, level of detail, and enjoyment. We determined there was room for improvement, particularly around the level of detail and enjoyment factors.

A graph showing how users responded to the content’s: reliability (70%), level of detail (64%), enjoyment factor (59%)
Source: Frankly, Green + Webb Survey 2018, Q. How would you describe the content you found?

We followed up this research with interviews to dig into what people were looking for from our content.

Our overall finding was that rather than having a passive audience who wanted to come and browse to read stories, our audience was driven by a question on a topic or theme — big or small. They were bringing their own starting point and looking to us — as a trusted source — to provide an answer or open a gateway to more information.

We chose to define the audience by their motivation i.e. they were “enquiry-led”.

How this impacted the approach to our content?

The first thing we had to decide was whether we wanted to pursue this audience — and improve their experience — or try to find a new audience who were interested in more esoteric long-read stories.

Given the resource available to us (in terms of time and budget), as well as appreciating the challenge of developing a whole new audience, we chose to prioritise and enhance the audience that existed: to create content that matched their expectations and encouraged them to return and associate us with high quality, trusted content. We would trial small experiments on the side to deliver collection content in a more serendipitous way to new audiences.

A screenshot of the Museum in a Tab experiment

With this in mind, we went back to the drawing board to brainstorm different content approaches that would meet our audience expectations.

We came up with a list of content principles for our stories which would meet our audience needs and also adhere to our organisational brand. These included:

  • Present our content in a clear and understandable structure — an audience need;
  • Adopt a tone / structure which rewards and encourages curiosity — an organisational goal.

From this we developed three content frameworks to structure our stories.

A photograph of some post-it notes from a workshop on our content princples
A photograph from our collaborative workshops on content principles

We then created some test content which followed these frameworks and tested these with a focus group which met our audience profile. The insights gained here helped to hone our approach for delivering story content.

What were the key findings and how were they implemented?

Well-structured

As this audience was driven by a starting query, it was important for them that our content was well-structured. Sites like Wikipedia were regularly referenced as exemplar to this audience.

We weren’t intending to create encyclopedic articles, as this didn’t play to our strengths, but we decided we could apply some similar techniques to our content: breaking story content into chunks with easy-to-scan and understand sub-headings, and pulling out the key information early so that audiences can find what they are looking for quickly.

Findability

This audience used search engines as their starting point for an online journey, and then filters the results according to what looks relevant and trustworthy. From our focus group, people did not expect our museums to offer this content, but were pleasantly surprised to find it.

This meant that embedding SEO techniques into our content creation was a valuable way to help audiences find our content. In collaboration with our curatorial team, we have been sharing keyword research and honing titles to ensure our content is as findable as possible. This approach is a work in progress to find the balance behind a click-worthy title that isn’t click-bait or misleading.

Media choices

Providing content via a mixture of media — text, different types of images, visualisations, video, audio — was highly valued by this audience, providing it was the most efficient way to showcase information.

This has challenged us to think creatively about our content and (within our budget) to look beyond text and images to tell our stories. We have been able to align with other projects around the organisation to — for example — create videos to show how an object worked or include a 3D scan of an object to enable users to explore it in more detail.

Be enjoyable

Although this audience is motivated by finding something out, the focus group revealed that they still wanted to enjoy the content they found.

For us, this had the potential to align with one of our brand values — to ignite curiosity. This has influenced the way we have chosen to structure our content: to flip users’ expectations from what they might think they know; to tell them something new; to structure stories so that users’ feel one question is answered but it leads to a new question.

Be a trusted starting point

For some of this audience, our content is the starting point in their journey to find out more. From the focus group, they wanted to follow up links and check sources.

This means that for every story, we have included a bibliography which covers our key sources, as well as links to further information to enable onward journeys.

What have been the initial results?

Over the course of 2019, we have published over forty new stories on the Science Museum website, including on the themes of Medicine and Chemistry, following our new content principles.

A screenshot of our Chemistry landing page — bringing together all our stories on Chemistry in one place

We have seen a significant increase in traffic to the Objects and Stories section of the website, with the number of visits doubling over the year.

We have also defined a “read rate” internally to judge page engagement, which will be a useful future metric of success.

A line graph showing the continual increase per month of visitors to the Objects and Stories section
Source: Google Analytics, January 2019 — January 2020

We continue to see this audience has a connection to science and technology.

A bar graph showing users’ interest in science and technology. Most had a personal interest
Science Museum Group survey, February 2020

We also see that users are still coming to find specific information — they are still an “enquiry-led audience” (75%).

We have seen a slight increase in users interested in planning a visit. This is likely to be because the content is now all integrated in one website. However, it challenges us to create links to the physical museum offer in sensible places through this content.

Bar graph showing the reasons users came to this content
Science Museum Group survey, February 2020

Finally, we rated users’ satisfaction. We added a further two measures which meet the new aims of our content — helpful and relevant; and tell me something new.

Although we need to continue to work on the level of detail, and whether this meets users’ expectations, as well as looking for opportunities to tell unique parts of a story using our collection, we have improved how much users find our content to be reliable (82% up from 70%) and how much users enjoy our content (75% up from 59%).

Graph showing how users described the content they found: reliability, level of detail; enjoyment; helpful; something new
Science Museum Group survey, February 2020

The future

By understanding and listening to our target audience, and incorporating this into our content creation, we have had positive successes in our qualitative and quantitative measures for Science Museum Group story content.

Our main areas of focus going forward will be to improve the findability of content within and beyond search engine rankings, as well as to experiment with the level of detail of our stories.

Going forward, having a short survey that we can run regularly will enable us to keep in touch with our audience, their motivations and their satisfaction, alongside any straight number reporting.

Acknowledgements

With thanks to Frankly, Green + Webb who helped us to work through this audience research project, and the Science Museum Group Web Team and Lalita Kaplish, Medicine Content Editor on secondment from the Wellcome Collection, for collaborating on this project.

--

--