Photo: hoyasmeg
Right now, if you’re among the vanguard involved in social media measurement, you’re probably measuring the media – the channel. It’s time to consider measuring the social – the audience. Let me exaplain…
Social media really is different from “old” media, in that the vast majority of it isn’t media at all but parts of it are, which make it all confusing.
By media I mean something like the New York Times or ABC – professionals creating content, content being collected into some outlet (usually but not necessarily any longer) with some editorial oversight, outlet distributing content to audience. Blogs on NY Times are still media even though you can comment and “share”, and NBC shows on hulu are still media even though they have user reviews and discussions. The only difference (“only” not making it any less significant) is that now the audience gets to participate more than they ever could in the past.
The point is – it’s still a channel, and marketers realized pretty quickly that they can use it as just another channel – a way to deliver their message to an audience. Here’s a good example – Audi delivering an R8 to a high profile blogger to drive for a week, with the expectation (which he didn’t fail to meet), that he’ll write a positive authentic (buzzword du-jour) review. Don’t get me wrong now – I think it’s good marketing – I read Guy Kawasaki regularly, but not any car magazines, and his authentic enthusiasm and great photos resonate with me, probably better than a professional review would have. I had an Audi years ago, and this blog sure managed to re-kindle a certain yearning.
Back to measurement. The sort of results you’d want to measure after this publicity stunt is just what you’d measure for “old” media – something like what you can find here:
“Dunkin’s SMI rose five points or 20% — from 42.6 to 47.6 – between the day before the Inauguration (Jan. 19) and Jan. 26.”
In other words, there are 20% more mentions of Dunkin on social media. Clearly their promotional activities had an effect.
Now finally my point: By measuring social media this way you’re missing all the really important stuff:
- Is it the same people mentioning Dunkin 20% more, or 20% more people mentioning Dunkin?
- Who are these people? Do they belong in the market segment that Dunkin targeted with its promotional activities? Did they manage to reach out (maybe inadvertently) to a new segment?
Both cases – looking at “lift” alone, and not segmenting – are classic old-media bias – where you had a few news sources and only cared about how many times you’re mentioned. That’s because you see media not as your customers, but as a channel through which to reach those customers. But that’s just not true with social media! A better name for it is consumer generated content. Once you realize that what you’re measuring is not the channel but rather the target audience itself, it becomes clear that buzz measurement alone isn’t what you’re looking for.
I’ll repeat this last message again: 99.9% of the people you find on social networks, blogs, message boards etc. aren’t high-profile bloggers like Guy Kawasaki. They’re read by only a handful of friends and thus are hardly worth the expense of giving them a sports car for a week – you ain’t gonna get much publicity that way. But they’re worth measuring, and in a way that’s different from media measurement.
Social media isn’t the media speaking at your customers.
It’s the customers speaking at one another.
So the takeouts are:
- You shouldn’t count mentions. You should count people.
- People are different. Learn about them. Segment.
