For the past five years or so, I’ve been watching the social media monitoring space. An early blog piece, with nearly 1,000 reads (my all-time high!), and a follow-up post discuss what I called “discourse analysis” – now more commonly called social media monitoring or listening. While vendors still make a lot of claims for advancement in this area, no one seems to have cracked the nut on how to analyze all the text data accurately, comprehensively, and contextually. I still think doing so is still quite a ways off.
I volunteer with a QRCA group that monitors/evaluates social media tools which may be used by us to help our clients. The other day, I was introduced to a new service, Ditto, whose tagline reads “The visual web, discovered.” This new service turns monitoring on its head because rather than analyzing text, it recognizes logos within images posted on Twitter and Instagram. You can even click on the pictures to see the captions/text to go along with the pictures. For me, a big “Wow, they can do that!” went off in my mind about five minutes into the demo.
Here’s a bit of info from their website:
So there are three types of recognition: facial recognition (what Facebook uses to know when a picture is you – scary), OCR – text recognition (now being used when you take a picture of a check and it’s deposited in your bank or handwriting recognition software), and finally pattern recognition (based on pixels forming shapes, be they large items like bicycles or hair dryers or refined images like company logos).
The service is clearly new and they have some kinks to work out, but in my view this is an important innovation to follow. Sure there are minor improvements (like image geocoding) which would add value. But do these tools have the potential to blend social media and qualitative market research? I think so.
A sidenote on QRCs history in this area: MROCs (market research online communities) were the first methodology which allowed qualitative researchers to bridge or marry social media with qualitative research. These studies, however, remain quite costly, so many QRCs, including myself, continue to look for new tools to serve their clients.
Ditto’s new service opens up new possibilities for more qualitative interaction moving forward. Imagine the following:
- We’re conducting a month-long traditional bulletin board project and for the month only, we ask participants if they’d allow us to see their Facebook, Instagram and Twitter postings. This could provide a lot more information and context to their qualitative research postings.
- We’re recruiting influencers in a particular category – let’s say outdoor apparel. We could use North Face’s or REI’s Facebook fan pages, but there might be very strong bias to those fans. Or we could use Ditto to help identify the biggest influencers via image uploads to validate actual use in daily life coupled with Klout scores and reach out to those individuals to see if they would be willing to participate in a research project.
There are probably many more examples of how a service like this would allow us to get to know our target audiences/participants even better. While I know Ditto is constantly refining the offering, and while it will have its limitations (what tool doesn’t?), in my view, this is a company marketers and qualitative researchers should watch.
I’ve always liked getting my feet in the water at the beginning of something to really understand how it works. I can’t wait to find a good fit for the inclusion of a service like this one. (If you’ve got something which you think this might work for, please let me know!)
Your thoughts on this latest option? Please weigh in!!