The Forrester Wave presents leaders in Social Media Listening. My contention is the market needs social media intelligence but this is not what excites the industry. Rather they just love “clicks on charts”.
Moreover, it’s in their DNA. There’s a need to click on a chart and see the actual tweet or verbatim underneath a trend. It’s from a good place. Let’s see a pattern, such as people talking about price increases, and click on the peaks or troughs and see who wrote what, when and check if they are influential. For all the natural language and machine learning that’s going on, they like a chart and a click through.
It’s second nature.
But it’s not that important in market research. Not unless you are doing influencer marketing. And then, frankly for most walks of life we know who the movers and shakers are, it’s reaching them that’s expensive.
What remains important is knowing your audience. Identifying the segments that exist amongst the audience you know intimately but also the unknown audience. The unknown audience that may well be the future success of your product and if you’re not asking the right questions you’ll never find them
The analysis of audience has a rich tradition in market research but has been evolving dramatically since the growth of social media. Indeed, it’s perhaps the richest promise of social media for marketers. Why? Because Facebook. Billions of users express their interests, and through behaviour show their colours. And while we can critique Facebook for previous failings on “Personally Identifiable Information” (PII), the reality they now create these audiences and open them for analysis anonymously.
This eats social media listening for breakfast. If you rely entirely on the actual posts people make, there are so many quality risks and misrepresentations. If you have access to actual behaviour that is completely anonymised you can dig deep into the segments, build personas based on maths and creativity, the very basics of marketing in a digital world.
E.g. Do women 35-44 that like technology also like travel to Asia. Do people over 65 that like travel also read more sci-fi novels. Is it only the young that are interested in bitcoin? This layering of information provided by Facebook and analysable by companies like SoPRISM answers questions that traditional social media monitoring cannot, at least not with the same accuracy.
I’ve seen and used the tools that take chatter from Twitter for audience analysis. But with respect, the data that gets returned is still at the mercy of numerous erroneous mentions, spam, a struggle that seems directly analogous to the struggles of sentiment analysis. The result being something good in theory, but rarely useful in practice.
So, why are SML companies not getting closer to proper audience analysis?
Firstly, it’s access to data. Facebook can be fussy and with all the shenanigans around Facebook’s connection to DataSift, I can understand it, to a point.
Secondly, there’s the Cambridge Analytica situation. Facebook quite rightly was challenged to explain that and deal with it. It has hurt their reputation and made for genuinely deep questions about who owns our data. But that has created one of those issue conflations that social media has burst forth with alarming rapidity. Being able to understand audience but having controls to protect against PII should be acceptable. Blithely opening your user base to exploitation of unscrupulous politicians and lobbyists is not.
Thirdly, they don’t have the technology. These companies have focussed on collecting data, building or integrating analysis software and making dashboards. This focus has enabled marketers to have access to all kinds of social media chatter when they say data overload is one of the biggest challenges. This approach has created a myriad of dashboards when users pay third parties to analyse the data. It hasn’t opened audience to any better level of insight though. It’s like you know your world is getting bigger and more complicated but you can’t pinpoint where, how or why.
So, going forward? People know that social media platforms aren’t charities. They can, in the main, accept their data will get processed. But, that data is valuable and there will be huge pressure for this analysis to be done with extreme caution, in an anonymised way. This is doable by using tools available today. I can show you if you like. Players in social media intelligence need to work with this (some have already, including one of the leaders in the Forrester Wave), those that don’t will keep making pretty graphs that you can click on and see who posted what. But it won’t be what customers who need insight want.
For more information, email me at simon.mcdermott@soprism.com