Ahead of her talk at UX New Zealand 2016, Lana Gibson from Lanalytics writes about a project she worked on with Te Papa.
Te Papa (a museum in Wellington, New Zealand) created audience personas based on user research, and I used these as a basis to create audience segments in Google Analytics to give us further insight into different groups. By regularly engaging with our audience using both qualitative and quantitative user insight methods, we’re starting to build up a three-dimensional picture of their needs and how Te Papa can serve them.
Personas based on user research
At Te Papa the digital team created six audience personas to inform their site redesign, based on user research:
These formed a good basis for understanding why people are using the site. For example the educator persona wants fodder for lesson plans for her class — trustworthy, subject-based resources that will excite her students. The tourist persona wants practical information — what’s on, how to plan a visit. And they want to get this information quickly and get on with their visit.
We’ll follow the tourist persona through a couple more rounds of user research, to give an example of what you can find out by segmenting your audience.
Interpreting tourist needs with data
Te Papa holds information for the Tourist audience in the Visit and What’s on sections of the site. I created a segment in Google Analytics which filters the reports to show how people who visit pages within these two sections interact with the whole site. For example the keywords they search for in Google before arriving on Te Papa, what they search for when on the site, and how many of them email us.
Deeper digging revealed that the Tourist audience makes up about half of our overall audience. Because the content is useful to everyone wanting to visit the museum, and not just tourists, we broadened the scope of this persona and called the segment ‘Museum visitor’.
Why segment by site category — what if the audience goes beyond these pages?
Google Analytics segments allow you to see all the pages that a particular audience visits, not just the ones you’ve filtered. For example over 2,000 people who visited a page within the Visit and What’s on sections also visited the Kids and families section in July 2016. So, the audience segment allows us to expand our concept of our audiences.
You can segment by a lot of different behaviors. For example you could segment visitors by keyword, isolating people who come to the site from Google after searching for ‘parking’ and ‘opening hours’ and seeing what they do afterwards. But segmenting by site category tests the information architecture of your site, which can be very useful if you’ve got it wrong!
Visit persona wants opening hours information
What did we learn from these personas? One example is that the most searched term on the site for the Visit persona was ‘opening hours’. To help fix this, the team put the opening hours on every page of the redesigned site:
This resulted in a 90% drop in searches that include ‘hours’ (May 2016 compared with May 2015):
Developing personas with Matariki
After the re-design the team ran a project to increase the reach and engagement of the Te Papa Matariki audience. You can read more about this in "Using data to help people celebrate Matariki". Te Papa holds Matariki events in the museum, such as the Kaumātua kapa haka, and this event in particular enhanced and challenged our ideas about this audience.
Experiencing Kaumātua kapa haka performances online
The Kaumātua kapa haka is the biggest Matariki event held at Te Papa, and this year we had 4,000 unique page views to the two Kaumātua kapa haka event pages. Traffic spiked over the event weekend, particularly from Facebook and mobile devices. We assumed the traffic was from people who were planning to come to the event, as they sit in the What’s on section. But further analysis indicates that people were visiting for the live streaming of the event — we included embedded Youtube videos on these pages.
The popularity of the videos suggests that we’re taking events held within the museum walls out to people on the move, or in the comfort of their own homes. Based on this insight we’re looking into live streaming more events.
We’ve taken Te Papa personas through three iterations, based on user research, analytics, then a practical application of these to the Matariki festival. Each user research method has limitations, but by regularly using qualitative and quantitative methods we’re engaging with a three dimensional view of our audience that’s constantly evolving. Each user research piece builds that view, and allows us to plan projects and site changes with greater clarity about what our users need. It means we can plan projects that will have real and measurable impact, and allow people to engage with Te Papa in useful and meaningful ways.
Want to hear more? Come to UX New Zealand!
If you'd like to hear more about how Lana and Ruth redesigned the Te Papa website, plus a bunch of other cool UX-related talks, head along to UX New Zealand 2016 hosted by Optimal Workshop. The conference runs from 12-14 October, 2016, including a day of fantastic workshops, and you can get your tickets here. Got some questions you'd like to ask Lana before the conference? You can Tweet her on @lanalytics00!