I had a bit of an eye-opening experience recently – with Word Clouds.
A colleague of mine in Marketing had included some Research slides in a NPD presentation she was giving to management. The Research part included some Word Clouds charts to present open-ended responses.
To to my amazement, the Word Clouds were a hit with the audience, apparently. Folk loved it.
Simply by changing the format from a grid of coded open enders with nets and other mentions to a visualisation of data had really fired the imagination of quite a few in marketing. Duh?
The weird but totally wonderful thing is that creating a simple Word Cloud is relatively easy and open to pretty much anyone. Take Wordle.net – http://bit.ly/WXjW – no registration necessary, it’s easy, you can play around with colours and fonts and it’s free. Plenty of the examples look stunning.
Beyond WordClouds, it seems there’s plenty afoot in the area of infographics – visual.ly http://bit.ly/r0fwOo – caught my attention, it’s in the process of allowing everyone to quickly and easily create professional designs with their own data sets.
Could we researchers learn something from this? Are we underutilising the power of data visualisation? Here’s my take.
1. Visualising data invariably helps an information-heavy presentation. Data that otherwise might seem daunting is suddenly appealing to the eye.
2. Data visualisation – including Word Clouds – when done well can help position a market research study as contemporary, cool. Frippery maybe, but so what – packaging works.
Reserach as an industry has, in my view, a positioning challenge – we’re still the dowdy analytical boffins in the eyes of many. This simple trick helps reposition ourselves.
3. Often we overcomplicate things in research. Wordclouds are simple, easy both to understand and explain (bigger = more important).- and often that makes them more memorable and actually makes the data more easily digested.
4. Care needs to be taken in Word Clouds with phrases, paraphrases or words that are so similar as to be identical in a coding exercise but wouldn’t be seen as such by a machine. It shouldn’t be a substitute for reading as many verbatims as you can fit in.
5. We need to learn to shift modes throughout a research project. The time for detail, precision, rigour is all fine for survey design, choice of methodology, sampling, questionnaire design – and also for the analysis phase of thinking through what the data means, what story it is telling us.
But when it comes to presenting, we need to think as communicators and analysts. And dare I say it, learn to perform, project.
I think there are serious pressures on both supplier and client side that work against crafting a careful, visually compelling final narrative and presentation – cost and deadlines for example. But going the extra visual yard (enjoyed avoiding a reference to a mile there….;) is actually extremely valuable, as it’s often the only part of the iceberg that internal clients and other influencing stakeholders see.
How good are we at bringing our data to life? I see many causes for optimism in my daily work, but data dumping is still alarmingly prevalent. Sometimes it seems to me that we are so focussed on engaging respondents with new playful techniques that we forget that our internal clients also need engaging.
So – here’s raising my glass to the humble Word Cloud. And hoping for more eye-opening, easy, scalable techniques and tricks to help us get our insights across in an engaging fashion.
Curious, as ever, as to others’ views.