How 3,000 can be bigger than 2.5 Million

For a while now I've been interested in Social Media marketing and as a nosy beggar, some of the data behind it. I recently read an excellent piece in the Financial Times (may require subscription/registration) from Simon Kuper talking about the change in journalism and the rise of 'content'.

This post isn't about that piece itself, but about some of the data around it, I saw about it from a tweet from the Deputy Editor of FT's weekend magazine Esther Bintliff that was retweeted on to my timeline:

Esther has around 3,000 followers but this tweet really took off probably due to a mix of the strong endorsement and also having an extract from the article included as a picture.

Compare that with the activity of Simon Kuper himself who despite himself having 46.8k followers (who you'd presume by following were interested in him) had fewer Retweets/Likes for his tweet.

There could be a whole range of reasons why Simon's tweet gets less activity such as:

  • Esther's tweet came out first (9:19am on Thursday 6th Oct, Simon's at 11:05am), although as well as who came first you'd also have to factor in when is the 'best time to tweet'
  • Esther's tweet contains two paragraphs from the article compared to Simon's one
  • Esther's endorsement of the piece 'Brilliant'

There was also a tweet about the article from the FT's main account which has 2.5m followers:

As the FT use Bitly as their link shortener, it's possible to publicly see activity from the link (that said I can't vouch 100% for the accuracy and how it would compare with their internal metrics).

 

As you can see from the above volume of clicks is pretty small (as was the Retweet/Like activity).

Is this due to timing (this came out at 4pm so several hours after the other tweets) or something else, such as the use of an illustration rather than an extract or the fact that it mentions Simon Kuper by name rather than his Twitter username.

Obviously retweet figures and Bitly links only give you part of the answer, for the full impact of the article you'd need the FT's own traffic data to try and untangle who/what was responsible for each spike in traffic for the piece along with a better view of any possible retweet activity of any of the above tweets from big influencers.

The whole area of what makes 'good' Social Media content is a huge one but things that need to be considered include:

  • Time/Day of tweet
  • 'Shareability': even if content itself is good, if the tweet itself doesn't sell it, people will be less likely to retweet
  • Tone of voice/Endorsement: You don't want to be saying 'THIS IS THE FUNNIEST TWEET EVER', but a bit of energy or enthusiasm for what you're linking to wouldn't hurt
  • Repeat tweeting (both from same and other accounts in the same 'family'): There's a balance between expanding your reach and spamming those who follow you. Most tweets unless they get picked up will effectively disappear with 10-15 minutes but on the other side, tweeting the same thing every hour will annoy those who do regularly check in to Twitter. The main FT account for example tweeted about Simon Kuper's piece a 2-3 times a day for a couple of days which I'd imagine is an attempt to balance reach vs spam.

All of the above is where data comes in, with so many variables creating the 'perfect' tweet will be a matter of trial and error, testing and refinement (even unwaveringly sticking to a 'perfect' format is arguably not an optimal solution).

The same applies to other situations such as when to send email newsletters or even this post 

If you're interested in seeing how people linked to the article enter https://www.ft.com/content/ed208574-8b4d-11e6-8cb7-e7ada1d123b1 into the search bar on Twitter and you'll find various types of links/promotions of the piece.