Is It Just Me, or Are There a Lot of #measure Tweets These Days?

By on in , , with 5 Comments

<Standard “good golly I haven’t been blogging with my planned weekly frequency / been busy / try to get back on track in 2011” disclaimer omitted>

Update: This update almost warrants deleting this entire post…but I’m going to leave it up, anyway. See Michele Hinojosa’s link in the comment for a link to an Archivist archive of #measure tweets that goes back to May 2010 and doesn’t show anything like the spike the data below shows, and also shows an average monthly tweet volume of roughly 3X what the November spike below shows. Kevin Hillstrom also created a Twapper Keeper archive back in early November 2010, and the count of tweets in that archive to date looks to be in line with what the Archivist archive is showing. So…wholly invalid data and conclusion below!!!

Corry Prohens’s holiday e-greeting email included a list of hist “best of” for web analytics for 2010, and he really nailed it. That just further validates what all web analysts know: Corry is, indeed “Recruiter Man” for our profession. He’s planning to turn the email into a blog post, so, I’ll sit back and wait for that. But, I did suggest that the #measure hashtag probably deserved some sort of shout out (I actually dubbed #measure my “web analytics superhero-sans-cape” in my interview as part of Emer Kirrane‘s “silly series”).

That got me to thinking: how much, really, has the #measure community grown since it’s formal rollout in late July 2009 via an Eric Peterson blog post?

10 minutes in my handy-dandy online listening platform, and I had a nice plot of messages by month:

Yowza! My immediate speculation is that the jump that started in October was directly related to the Washington, D.C. eMetrics conference in the first week of October — the in-person discussions of social media, combined with the continuing adoption of smartphones, combined with the live tweeting that occurred at the conference itself (non-Twitter users at the conference picking up on how Twitter was being effectively used by their peers). That’s certainly a testable hypothesis…but it’s not one I’m going to test right now (add a comment if you’ve got a competing hypothesis or two — maybe I will dive a little deeper if we get some nice competing theories to try out; this will definitely — the horror! — fall in the “interesting but not actionable” category, so, shhhh!!!, don’t point your business users to this post!).

It’s also possible that the data is not totally valid — gotta love the messiness of social media! I’d love to have someone else do a quick “conversation volume” analysis of #measure tweets to see if similar results crop up. Unfortunately, Twitter doesn’t make that sort of historical data available, I shut off my #measure RSS feed archive a few months ago, and, apparently, no one (myself included) ever set up a TwapperKeeper archive for it. So, I can’t immediately think of an alternative source to use to check the data.

Thoughts? Observations? Harsh criticisms? Comment spammers (I know I can always count on you to chime in, you automated, Akismet-busting robots, you!)?


  1. I wonder about not just the raw number of tweets for #measure but also the individual accounts that use it. There are plenty of “(super high number)% of tweets come from (super low number)% of accounts” articles floating around right now.

    Is it mostly just the #measure superstars tweeting more or is there more of a community being generated here as Twitter becomes a preferred platform for the discipline?

  2. I have an archive of #measure tweets using The Archivist that doesn’t seem to show such a dramatic spike:

    However, they do caution that they may not archive every tweet. The FAQs suggest that if an archive has less than 1000 tweets/day, they should mostly all be captured, but at higher levels, may miss some.

    I would be surprised if it was eMetrics DC. From most of the tweets I sent, read, etc, it seems the #eMetrics hashtag was mostly commonly used, and #measure was often left out if the tweeter needed characters for their actual tweet.

    Is it possible that the listening platform you’re using is more accurate re: recent history, perhaps because Twitter isn’t very helpful too far back? (Can you tell that when things shift dramatically, I first assume the data is wrong until proven right?)

    Assuming the increase is legit, it would certainly be interesting to see whether it’s tweets per user that are increasing, or number of users contributing to the hashtag. Also, I won’t lie: I’m curious whether use of the #measure hashtag has increased around Jason Thompson’s charity:water campaign (

  3. Ty – that’s a good point. And, man, it would be super-simple to answer if I just had all of those tweets and the usernames associated with them. Dangit.

    Michele – I’m definitely now questioning the integrity of the data I used. It’s probably a whole blog post to work through the caveats of trying to use an enterprise online listening platform for some types of analysis (or even for performance measurement). This may wind up as a springboard for that!

  4. It’s frustrating trying to properly analyse social data, huh? To be honest I don’t necessarily trust any of the data (not even the one I posted a link to), but if we have corroborating data from kevin at least that’s good.

    FYI I think it’s good to have kept the post up with the disclaimer.

  5. Pingback Gilligan on Data by Tim Wilson » Blog Archive » If the Data Looks too Amazing to Be True…

Leave your Comment

« »