Just how do you measure how effective Twitter is?
One day w
e will all click a button and some kind of advanced free analytic will do it all.
There is of course twitter.grader.com.
But a score out of 100 isn’t really going to cut the mustard with the chief executive.
What are we doing? One way is to keep a log of the traditional opportunities to view figure. In other words the number of times a tweet has been put in front of people.
Until something better comes along we’re keeping tally ourselves with little more than a word document and a pocket calculator.
I’M AN OPERATOR OF A POCKET CALCULATOR
We do it monthly. First, we keep a tab on the number of followers on the first day of the month.
1. At the START of the month log the current number of followers
2. At the END of the month log the new total of followers
3. Work out the average number of followers that month.
4. Work out the number of times you’ve tweeted that month.
5. Then its MONTHLY AVERAGE FOLLOWER score multiplied by MONTHLY TWEETS. You are left with opportunities to view.
OPPORTUNITIES TO VIEW: THE MATHS
This is the formula we use. It’s a formula that is designed to show the impact of our Twitter use. It’s not neccesarily the one for everyone but it’ll do until a smart app designer comes up with something that gets traction industry wide. It works like this: If there are 100 followers at the start of the month and 200 at the end the average follower score is 150. Yes?
Let’s pretent the tweets we sent that month was 50.
If that was the case our opportunities to view score would be 150 x 50 = 7,500.
In other words, there were 7,500 opportunities to read your organisations tweets.
As a swift number crunch, work out how many months you’ve been tweeting, your followers today, divide your followers today by the number of months and you can come up with a rough figure without having to put in months of investment.
A CASE STUDY: @WALSALLCOUNCIL ON TWITTER
From April to November ’09 our tweets could have been read more than 700,000 times. This sounds a compelling score – and is – but is by no means unique.
April 09 — 59 tweets x 77 average monthly followers = 4,543
May 09 — 172 x 175 = 30,100
June 09 — 251 x 166 = 46,646
July 09 — 450 x 288 = 129,600
August 09 — 540 x 215 = 116,100
September 09 — 642 x 244 = 157,136
October 09 — 758 x 293 = 222,094
Opps to view total April to November ’09: 706,319
WHAT ABOUT MEASURE LINK CLICKS?
If you are really keen you can use a link shortening website like bit.ly.
From that you can also get data for the number of clicks. However, this is only collected link by link so you can’t bring them all together. It’s also time consuming going through each click. Mashable reckons an average is about 3 per cent click through.
SO, GO AND TELL PEOPLE
Now you’ve got your stats what are you going to do with them?
You could leave them on your hard drive but isn’t it better to spread the word?
Stick it on the intranet. Tell all your friends.
With thousands of organisations on Twitter I’m amazed a free killer app hasn’t been designed already to properly measure.
Until then, I’d be genuinely interested to know what others do…
Three basic things organisations should be doing when they use social media:
1. Measure. Whatever the way you want to measure – followers, friends, opps to view or views – keep a log. Yes, it’ll take time. Yes, you will come up with some compelling figures that paint a picture of what you are doing.
2. Broadcast. Tell people in your organisation what the statistics are. Don’t keep them to yourself.
3. Circulate case studies. Turned around an inaccurate Chinese whisper using Twitter? Take a screen shot of each tweet. Put together a mini Power Point and circulate. Let non-adoptors know what you can achieve.
4. Put your social media stats with your Press monitoring stats. Don’t keep them in the box bedroom. Let them breathe. It’s also a good way of getting the message over to people that the weekly paper that has had the monopoly for 100 years is not the only game in town.
LINKS
E-Consultancy debate on measuring social media success http://bit.ly/Qohd3
This is a really useful post Dan. Thanks for writing out your thinking.
Definitely agree with the value of taking screenshots, that you mentioned in 3). Screenshots of a discussion, say in a Twitter client, have helped me demonstrate the conversational nature of Twitter – to counter the ‘it’s people saying what they’re having for lunch’ position.
I’ve been looking at how to measure link clicks in tweets and wondering if you and others might have some more ideas to try out?
I’m using bit.ly for links in @thewmro’s tweets but, frankly, not getting much out of the link click stats. The content of the organisation’s tweets is specific to a geographical area in England – as are most of the followers – yet approximately 70% of the link clicks reported by bit.ly are from outside the UK. I’m putting this down mostly to bots rather than real people interested in the link.
Also, it’d be interesting to hear how others track referrals from links in Twitter clients to their sites. I might be missing something in Google Analytics but don’t see referrals from bit.ly. Referrals from twitter.com, while genuine, are only a tiny proportion of actual referrals – as most visitors are arriving from clicking a link in a Twitter client, not the Twitter website itself.
Cheers, Gavin.
THis whole measuring social media impact lark is very much in its infancy. Frankly, I’m amazed that someone bright has not come up withsomething that quantifies things better.
It’s no good telling your boss it’s groovy and cool. You have to deal with real world figures. I’m positive of that.
I don’t think measuring followers is quite it. This leaves me to think that some kind of overall measuring is the answer.
In many ways this approach is a bit like http://www.tweetreach.com in counting total impressions. (For an example see: http://bit.ly/bLIzPw.)
Yes, bit.ly is there to link shorten and count clicks. There is some merit in it but it tells part of the picture. An unclicked link which tells a positive story may still be read as it floats through the stream. Only looking at bit.ly means you miss this.
One apprtoach at quantifying things that I quite like came via Steph Jennings of the Light House arts centre in Wolverhampton (she is @essitam on Twitter). Her approach was suggested at a session she attended.
Basically, you count tweets over a give period and then work out how much it would cost to send a fraction of those by post as a direct mail shot. Even at 30p a pop and disregarding print costs that still leaves you with a fair pile of cash.
From research published on Mashable the average click through was around 6 per cent. We’ve started to use this as a measurement at Walsall too.
Mind you, if someone talented like @pezholio would just get a wriggle on and invent something we’d all be much better off…
Thanks for the reply Dan. I haven’t seen tweetreach before.
Like the idea of Steph’s approach. I spoke to a Councillor who looked at visits to his blog in a similar way – he compared visits to his blog with delivering the same information door-to-door with printed leaflets.
A tweet’s reach compared to traditional method of sharing the same information sounds interesting.
While this area is new, I guess it’s down to individuals to think about the metrics that are important to them, and start measuring informally, rather than going for an all-encompassing strategy.
JFDI again!