Measuring Twitter

Barely a day goes past in social media without another survey or statistics being thrown around. 47% of Instagram users have taken a picture of their pet. Just 14% of tried to engage with a QR code on TV. And so on. There's so much of this stuff flying around - much of it interesting, or a good starting point for discussion - and easy to digest that it's difficult to sort the relevant from the noise. And yes, I'm equally guilty of Tweeting out links with links to these surveys.

Which is why the BBC College Of Journalism's post on How To Sell A Tweet is so refreshing. It looks at the numbers and asks questions that any ood journalist should be asking (but sometimes get lost among the churn of social media news and research).

It's worth asking questions about data and engagement surveys rather than just accept the results blindly. So 47% of people behave this way with social media - what about the other 53%? And, fine, only seven per cent of people may engage with a social advert - but seven per cent of what?

For example, I've seen the stat about weekends being the best time to Tweet. Yes, but as Charles Miller says, has anybody asked if this is due to an under-populated area.

Whether you’re a journalist or a brand, to get attention you’re trying to find an under-served audience. If your competitors all start tweeting at the weekend, attention per tweet will inevitably fall.

And if you schedule some Tweets for the weekend, that may not take into account other events such as (don't laugh) the weather, big news stories, events that will invariably catch the imagination of Twitter, and definitely big sporting events (although this also shows that you should have a good understanding of the demographic you're aiming to reach with your Tweet).

In some respects this is no different from traditional PR and event planning, just with a slightly different mindset. Journalists and news organisations, as Adam Tinworth and others have noticed, need to start thinking more in the mindset of individual marketeers and acting as their own PR. Or, at the very least, learning more from content aggregators.

My day job involves communicating to several different groups, the majority of whom who tend not to be online overly early in the morning and are more active in the evening, as well as a smaller influential grouping, who are often (but not always) more active during the day. Even just having an idea of who I'm trying to reach at any given time helps.

Currently, I use SocialBro for analysis of the best times to tweet around these very different yet equally important groups who we want to reach, although there are other tools out there.

But getting a picture of when the majority of followers are online is invaluable. Four months ago, it was 2pm on a Thursday afternoon. Now it's 5pm on a Friday. Next analysis may well show a different time altogether (weekends have never featured prominently though).

Again, it comes down to knowing and profiling your audience and acting accordingly. If, say, you're targeting sports fans, you may see that many of them are online during the weekend.

But if you Tweet at the weekend, you're probably catching them when they're watching a match and their attention is elsewhere. It may seem like common sense, but research and planning into social campaigns and when to roll them out are essential.

Just turning up with statistics from one piece of research that says "Weekends are the best time to Tweet," is an interesting starting point but not to be followed blindly. Drilling down into this data is likely to produce far more useful results (and may, indeed, indicate that for your professional accounts, you need to do the opposite), if you're prepared to spend a bit of time on more detailed analysis.

Which, in all honesty, is largely true for most data and statistics in any walk of life.