This article was originally published on the Webcredible blog.
User research encompasses all kinds of methods: user testing, focus groups, interviews, ethnography, surveys, card sorting, tree testing, to name just a few.
One thing these all have in common is that they also suit us as UX researchers, in that we can conduct them in a lab environment. Whilst this is great in terms of planning and executing controlled experiments to answer specific research objectives, it can introduce things like the observation bias (i.e. that people behave differently when they are being watched).
So what else can you do to get more information on users? While you wouldn’t normally think of Facebook and Twitter as a potential rich source of user data to inform UX requirements, they can in fact be immensely useful.
Pros and Cons
The obvious advantage of social media research is that it’s cheap and quick to do. You don’t need to recruit participants and you don’t have to worry about data protection. You can capture user thoughts in a fairly natural way and you remove the traditional observation bias (albeit whilst introducing new ones).
At the same time it does have some cons. Your sampling frame is limited – not everyone tweets or feels comfortable sharing information on Facebook. The data is self-reported, rather than natural behaviour. And there is a potential risk that users are personifying a certain self-image rather than their real thoughts, a different kind of bias.
With all that said, this is probably worth trying when your research budget is tight and you know lots of people are talking about your brand or product on social media.
Dave Ellender recently gave a great talk at a UX in The City event in Oxford. He pointed out that social media can be a rich source of data on customer needs and pain points. In a recent project he was tasked to find out how digital solutions could better service the Dartford Tunnel. By analysing 1,000 tweets related to the Dartford Tunnel he was able to produce 36 user stories relating to customer pain points that needed solving.
He talked through a process, which you can try on your own projects.
1. Find a Suitable Data Source
Look at different social media platforms and find one where there is plenty of content related to the client/product/subject you are interested in. You can use metadata such as tags so this could be Twitter, Facebook, Instagram, etc. essentially anywhere the conversation is happening and you can collate that data.
2. Define Your Keywords
Create a list in a spreadsheet of all the keywords related to the subject or research objective you have in mind.
3. Use a Monitoring Tool to Collect the Data
For Twitter there are lots of tools out there that can help you automate your data collection for given date range. Try Hootsuite, Sproutsocial and Socialmention. You can use these tools to looks up phrases, hashtags and user accounts related to your keywords.
SproutSocial Smart Inbox allows you to monitor different social media accounts based on tags and filters.
Then you need to get a sample of Tweets into one place. If you have a developer you may be able to get them to scrape the data and export it out into a spreadsheet. Otherwise you may have do it manually.
4. Analyze the Data
Follow a 3-step process for qualitative coding of the data set.
- Run through each Tweet and tag specific events and objects.
- Run through a second time tagging common categories of the subject.
- Take stock and run through a third time looking for implicit meaning from the Tweet – What is the user trying to imply?
If you have lots of data, you could use a team to get through the data quicker, also sharing learnings. This is a qualitiative analysis so sharing interpretations may help balance out the subjectivity.
5. Create Research Insights
Group common themes together and document specific needs and pain points that have arisen from the analysis.
A good format is to document these as user stories e.g. As a user I need to do X so that I can achieve Y.
———————
And there you go. In just a few hours you’ve learnt a load of new things about your users. You can use this data to triangulate with other data sources such as user interviews, come up with design hypotheses and inform your product strategy. The possibilities are endless!
Jack Josephy is a Senior Experience Consultant with Webcredible. Jack is a UCD purist and is as much about the process as he is the execution of projects. He sees UX as a process of aligning user needs with business objectives to inform the right product strategy.
