Is the result of a Penn State study in the States.
Researchers led by Jim Jansen, associate professor of information science and technology, and Twitter chief scientist Abdur Chowdhury looked at half a million tweets. 20% of them were apparently people 'asking and providing' product information. Assuming three million tweets a day, that would translate into 600k posts daily of direct relevance to brands.
I initially found that % on the high side, though 'providing product information' is a definition that's wide enough to include any tweets about a product or service - I went to this restaurant today, I bought this mobile phone and so on. I guess it is true that as a personal broadcasting system we do use Twitter to talk about stuff we buy or like / dislike a great deal.
Case in point I've - almost unconsciously - made some kind of comment about four different companies since the weekend.
According to the study authors, the large amount of brand data on Twitter can pretty much provide you with a sentiment map if you monitor and analyse tweets over time: What do your customers and non customers think about your product, what features are going down well / not so well, and how are your competitors faring.
One to add to our list for internal clients of 'what is Twitter good for?'
Image - marc.benton
Link to original postLink to original post
Link to original post