Democratic Institutions Minister Maryam Monsef invoked Twitter during Question Period yesterday (May 12). She declared, “Yesterday [May 11], the first day that we brought forward the conversation on electoral reform, the hashtag #electoralreform on Twitter alone garnered nearly 12 million impressions in one day.”

Her inference is there is a plurality of online interest in electoral reform, particularly on Twitter.

I decided to fact check Minister Monsef’s numbers using Sysomos MAP. Is she right? Or, is interest confined to us political geeks?

Definition: Impressions

Minister Monsef’s statement was so confident and assertive, many people hearing her remarks on the news might think that Canadians had issued 12 million tweets including the #electoralreform hashtag. The benefit of insider language in public messaging is the public is likely to accept your implied meaning.

A Twitter impression is actually just an opportunity to see a tweet (for example, the Tweet appeared on someone’s screen). It doesn’t mean people saw or read the content. It’s the same as saying there will be 100 impressions of the front porch of my house as people drive to work this morning. While it’s true 100 people may drive past my house, few if any of the drivers will notice our front porch, much less the fact that it doesn’t have our house number on it, even though it’s in their clear field of view. The drivers have to decide to direct their gaze directly at the porch over other visual cues to actually see it.

More importantly, an impression is a passive metric. It doesn’t necessarily reflect interest or concern. For that you need an actual relevant tweet or a completed action based on a tweet.

As a result, I’m not a sold on impressions as a meaningful measure in isolation. At best, it has a supporting role in measurement and analysis.

Minister Monsef’s numbers are incorrect

On May 11, there were just 2.5 million estimated impressions on 672 #electoralreform tweets issued from 423 unique Twitter accounts. That’s 80% fewer impressions than Minister Monsef declared. [CORRECTED 9:40amET, May 13]

Her numbers remain incorrect even if we expand our analysis to cover the week spanning May 6 through May 12 inclusive. In that case, 1,188 unique Twitter users issued 1,896 tweets tagged with #electoralreform, accounting for an estimated 8.4 million impressions. That’s still 30% fewer impressions than the suggested 12 million on May 11 alone.


Daily #electoralreform Twitter activity. Analysis using Sysomos MAP.

It’s also worth noting that Canadians account for 78.7% of discovered #electoralreform Twitter activity, 15.6% was from the UK and 3.2% from the US.


Distribution of #electoralreform tweets by country. Analysis using Sysomos MAP.

If we focus on Canadian #electoralreform Twitter activity for May 6 through 12 (inclusive), the numbers get even smaller. In this case, there was an estimated 4.4 million impressions on 998 tweets from 549 Canadian Twitter users.

Basically, #electoralreform participation and quantity of mentions (and impressions) were decidedly low. And, that doesn’t even consider the nature of the mentions or the sentiment/affinity regarding #electoralreform, its various options, committee structure, etc…

Perhaps this is what Minister Monsef was thinking about

[APPENDED MAY 14, 2016, 11AMET] Changing the search criteria to include mentions of #electoralreform, electoralreform, electoral reform, first past the post, ranked ballot and proportional representation for May 11 reveals that 1,163 Canadians issued 2,234 relevant tweets with an estimated 25.1 million impressions. Maybe that’s what Minister Monsef was thinking about, though she is is still incorrect. Those numbers increase to 3,859 relevant tweets from 2,233 participants if we consider global Twitter traffic for the day. In that case, estimated impressions increase to 33.3 million.

Not that most of that traffic (59%) was retweets to tweets issued by federal politicians (of all parties), journalists and media reports.

For reference, expanding the analysis period to May 6 through 12 reveals there were 14,837 relevant tweets (4,923 from Canadians) issued by 7,665 unique Twitter handles (2,086 Canadian) with an estimated 114 million impressions (54.1 million based on the Canadian tweets).

Parting thought

Twitter is an incredibly powerful platform for meaningful communication, conversation, engagement and opinion measurement — among many other uses (e.g. humanitarian relief coordination, diplomacy, etc..). However, people and organizations continue who to demonstrate weak understanding of how to use Twitter and measure its impact.

Twitter has a role in our political system and in productive democratic discourse. Let’s be careful how we use it and talk about it lest we create the conditions where its actual value and impact can be dismissed.