…and the dumb cousin of text analytics
…and the dumb cousin of text analytics
First, I want to thank all the Top-NGMR Bloggers who participated in the largest ever spontaneous social media collaboration in market research yesterday. It was an interesting experience for me in so many ways.
I’ve been asked several times if I’m going to summarize the results. I’m actually hoping to leave that to others. I was also hoping that perhaps some of the bloggers at Forrester or one of the mainstream market research publications such as Quirks, IR or RFL would give us their summary and macro analysis. I’m sure many of us would find that very interesting.
I’m hesitant to do the analysis myself as I wouldn’t want to bias the summary (I organized and participated in the event myself, and plan to do so again in the future). Additionally, this week is a busy one for me, and to truly give this the attention it deserves you really do need to commit to more than a cursory glance at the many blogs who put effort into their predictions. I definitely plan to come back to these predictions more than once for inspiration, as I’m sure many others will.
That said, I’m often asked, “Can we use text analytics to analyze and write our report for us” . The answer is that – for many reasons but perhaps especially because this is a sample of n~35 – text analytics isn’t really appropriate (Qual researchers everywhere your jobs are definitely safe for the foreseeable future).
In one of my NOT HOT predictions I mentioned one of my pet peeves, word clouds. Word clouds are like the dumb cousin of real text analytics. However I also said I reserved the right to use them one more time on this blog. So here goes, word clouds of ‘what’s Hot and what’s not’ in the future of market research:
Can you tell which word cloud is which?
It’s harder than you think isn’t it. The NOT HOT word cloud is listed first, and the HOT cloud second. If you spend more time tweaking these wordle word clouds I’m sure you can get something with greater differentiation, and I welcome anyone who wishes to do so. However, I think this also illustrates the problem I have with word clouds and why I think they won’t be HOT too much longer.
If anyone is doing a more thorough summary of yesterdays predictions please ping me, I’d love to link to it from this blog. NGMR Top Bloggers interested in participating in the next event please DM me on Twitter for details.