Adopting An Analytics State Of Mind

Last weekend’s 2013 Manning Networking Conference had a number of interesting speakers and topics. One session in particular, The Data Revolution (summary here) featured Twitter rockstar Minister Tony Clement, Victory Lab author Sasha Issenberg and social media pro Stephen Taylor. During the discussion, the panel discussed how important data-driven organizations are to the future of campaigning for votes.

Unfortunately, the practical reality is that most political organizations in Canada are far from data-centric campaign machines; the thought of achieving a fully integrated campaign infrastructure like the 2012 Obama outfit is a daunting one. But I assert that Canadian politicos can take a huge step forward by adopting an “analytics state of mind”: striving to measure as many campaign activities as possible and using that feedback to inform future decisions.

Canadian politicos must adopt an “analytics state of mind”, measuring as many activities as possible.


The Obama campaign didn’t necesarily set out to create a highly advanced digital operation. Rather, they adopted a philosophy that instead of relying on instinct or convention, they would measure everything – using that data to determine what worked and what didn’t: “We were going to demand data on everything, we were going to measure everything,” said Obama Campaign Manager Jim Messina. And they sure did. In fact, they tested 13 different email subject lines before settling on the one that would go on to raise over $2.6 million.

While Obama ran statistical models to predict turnout in swing states, your campaign doesn’t need to have it’s own CTO to start adapting an analytics state of mind. The great part about this approach is that it is one you can adopt right now. Further, you already have access to the (mostly free) tools to begin measuring and analyzing campaign activity. What kind of activities am I talking about? Here are just a few:

  • Open rates on outbound email blasts
  • Most popular content on your website
  • Facebook content that gets the most likes
  • Subject of tweets that get the most retweets
  • Which links are clicked on
  • Ads that are clicked on the most

These examples may be limited in their individual impact, but you need to walk before you can run. In my experience, most political organization in Canada are taking a “fire and forget” approach: get stuff out the door and hope for the best. Ask yourself: how often do you look at your web analytics? What do you do with that data? Does it inform your decision process? Probably not.

But by taking the view that you need to measure outcomes to truly understand what works is the best way to begin refining your methods. This is referred to as “A/B testing”, an important method of comparison. After all, if you send out an email that no one opens, did you actually send it? By trying out different approaches and measuring them against each other, you can separate the wheat from the chaff, discarding that which doesn’t produce results.

Is a photo more likely get likes and shares on Facebook than posting a link? Which social media channel drives the most traffic to the website? Does a humourous video have more impact than one with a serious tone? Which email subject line will get your supporters to open it? Do visitors respond to a certain word or phrase? The analytical approach can go quite deep.

So before you turn your attention to hiring data scientists, it is important you first adopt the habit of measuring any activity that can be measured. Make collection and analysis of this data a central part of every campaign effort. You will then be on a solid path to becoming a data-driven organization.

Scroll to Top