Text Analytics - Not Tweet Analytics
[Interview with OdinText founder Tom H. C. Anderson reposted from TotalCIO (IT Knowledge Exchange)]
A social media contrarian speaks out on the limited value of tweet analytics
by Nicole LaskowskiProfile: Nicole Laskowski
A customer loyalty or a customer service department may not be the most “analytically inclined” departments within a business, but they may have some of the best data. That’s according to Tom H.C. Anderson, consumer insights expert and founder of Anderson Analytics - OdinText. In fact, he advises businesses forgo analyzing social media data in favor of milking those internal and potentially lucrative cash data cows.
“Phone calls that come in from customers, for instance, and emails — whenever there’s contact like that, there’s a lot of information being generated,” Anderson said. The information ranges from unstructured text to structured data that can be as specific as an SKU number.
That data, alone, is valuable. Combined with other metrics, such as consumer behavior or ad campaign success rates, it becomes more valuable still. And much more valuable, in his view, than a folder full of tweets.
But he’s not preaching to the choir. Businesses have bought into the notion that social media and customer insights are joined at the hip. Plus, what Anderson is proposing companies do instead is not easy. As CIOs know, accessing and integrating data from different sources — let alone different departments — remains a challenge.
Next week, Anderson will moderate a session on customer analytics at theUseful Business Analytics Summit in Boston. In advance of the summit, I had a chance to sit down with him and talk about why businesses should be wary of social media data.
What is the biggest challenge businesses face in consumer analytics and how can CIOs help business overcome that challenge?
Part of the problem has been and continues to be just the sharing of data — getting access and combining the data that matters most. On the one hand, everyone’s rushing toward social media data because no one owns it — it’s free, available …. Meanwhile, the data companies are already collecting, [they] aren’t merging. So, what I’m talking about, for instance, are things like customer satisfaction survey trackers along with metrics such as actual behavior — how much a customer spent, how successful were certain campaigns, and returning behavior. Giving people access across departments to that data and doing more interesting analysis that involves more than one set of data and using the data that’s most suitable to answering the business problem: That continues to be a challenge.
It sounds like you’re suggesting that companies need better data accessibility as well as user-friendly tools that can provide that access or enable consumption.
Both of those. Tools are becoming more powerful and easier to use, so we have that. Where there’s a shortage is of people who are good at analyzing the data. One of the things I’ve championed for is that the consumer insights/market research folks should play a greater role [here] because they have so much experience analyzing customer behavior already.
Let’s get back to social media data.
They call it ‘social media data,’ but, really, 80% of it is Twitter data … Facebook, LinkedIn — all of that other stuff is off limits.
Why are tweets so attractive? Is it because the business is data hungry and tweets are easy to access?
People think just because data is big it’s got to be important. No. Data is as interesting as what it represents. And in this case, [tweets] represent about 10% of the population. And there’s huge overlap between the 10% who blog and tweet. They’re not your average person. They’re very different from the average LinkedIn or Facebook user. …
Yes, [the attraction] is partly because it is so easy to gain access to and it’s also the way it’s being sold. [The message has] moved, to some degree, away from, ‘you’re going to find the answer to every question in social media monitoring.’ (Not totally, some people are still claiming that.) But [now they're] selling it as a fear factor-type thing, saying, ‘You can’t afford not to listen to Twitter because, God forbid, one person says something bad about your company.’ I’d argue that’s not a good reason to spend a lot of money on Twitter.
The real point here is that businesses [caught] in this social media frenzy are forgetting they have interesting, much richer data that can answer a lot more problems.
What are businesses doing with that social data other than listening?
The short answer to that is, it’s very limited. I’m invited to speak to 20 conferences a year, so I’m exposed to a lot of presentations, and a lot of them are about social media in one respect or another, but I’ve yet to be really impressed.
Now there is talk about doing customer acquisition targeting on social media. So, in other words, somebody who’s selling home insurance, [will use] algorithms to look for people who are tweeting about buying a new home. You find them, classify them in certain income groups and so forth, and [determine if they're] a good target to sell your product to. But that’s still very much vaporware. It’s good in theory; it’s not actually being done.
Do you see this changing anytime soon?
A few years ago, we were talking about walled gardens. Facebook has a walled garden, LinkedIn has a walled garden. In other words, they own the data. We were talking about how all data is going to be free for everyone — like Twitter. Because Twitter [uses] basically the same technology that blogs [use]. You post it out there and it’s pushed to everyone. But that never happened — these walled gardens coming down.
Instead, we see the opposite: Data is being protected more. Facebook has access to obviously large amounts of extremely interesting information on everyone. Google is another one. And what do you think is going to happen? It’s not going to be about sharing that data. People are thinking more about privacy and there’s legislation happening. Some people in my industry for whom data is important seem to think Amazon, Google and Facebook will protect us from that legislation … but the opposite is going to happen.
Facebook, Google — there’s more for them to gain by closing the opportunity — basically a data monopoly. So Facebook and Google can say, ‘You’re right. Privacy is an important issue and nobody else can protect or should have access to this data. Let’s close it up and make sure we don’t have any competition.’ And that’s really what they’re petitioning the politicians for.