I just received a note from a colleague about the “Twitter Influence Calculator.” Not being able to resist, I immediately put in my name and out popped more statistics than you can shake a stick at…
I chopped out the actual numbers because they are beside the point (and tiny).
Is this a good thing?
There are so many ways to respond (I’m of two minds about my ambivalence).
Is this site and others like it, a reflection of our endless self-absorption or is this another useful tool for measuring our interaction and effectiveness with the developer community?
More generally, do sites provide these numbers because they are meaningful or because they can? Are they meaningful? I’m not enough of a statistician to know whether the values make sense, though the person who created this site clearly has a great deal of expertise
But we do this a lot. There is also Twitter Grader which happily assures me that although my overall rank is 17,705, that is out of 1.4 million Twitter users, giving me a rank of 98.8 which they pleasantly round up to 99. 99 is good. I like Twitter Grader better, cause I have better numbers there.
Side note: my one good number on the Influence calculator is the Twitter Ratio, where I scored 11.3. The explanation you receive when you click on the term (see the last line in the figure above) is quite detailed but it includes this: “A …ratio of 10 or higher indicates that you’re either a Rock Star in your field or you are an elitist… You like to hear yourself talk. Luckily others like to hear you talk, too. You may be an ass.”
Figures Lie and Liars Figure
Microsoft is big on numbers, we spend a lot of time measuring influence and customer happiness (remember to fill out your eval forms after each talk) and I think that is great, because what you measure you’ll get more of, but sometimes we split things a little fine.
We measure everything, and then we set our pay and rewards based on those decimals and infinitesimals. Overall, who can argue? It makes for a highly motivated group, and weeds out the dead weight, but one does worry a bit about how stochastic some of the results are.
Qualitative != Quantitative
At another company I worked for, we used to hold focus groups with 20 people, ask their opinions, and then come back and say “83.7% of the people we talked to said…” It made my head hurt. When n = 20 (or 50!) you really don’t want to be using quantitative values finer than, “some” or a lot” – and even then with a shaker of salt.
Don’t mind me, I’m just jealous that Scott Hanselman and I set out to Twitter around the same time and a week or so later he has, according to the Twitter Influence Calculator, over 8 thousand followers with a visibility that they characterize as “Astonishingly High.” What can I say? He’s a rock star. I’m a roadie. Life is good.
Next: Using Statistics Well