Londoners rejoice! After months of campaigning, no longer will our twitter feeds be clogged with awkward candidate selfies and #labourdoorstep. (Note to politicians: not all photos of you are selfies. If you didn’t take it yourself, it’s simply called a photo.)

Over the past two months, over 600,000 tweets have been sent mentioning the candidates. Some were positive, but according to analysis by CASM at Demos, most tweets have been negative.

Of course, if tweets translated into votes, Ed Miliband would be prime minister by now. In fact, the relationship between Twitter support and support in the ballot box is a complicated one. With the most recent polls showing Sadiq is holding onto his 12 point lead, in this, the last in our series on the role of social media in ​the mayoral ​elections, we take a quick look at the extent to which Sadiq’s lead translates online.

Sadiq vs Zac: head-to-head on Twitter

Let’s start by analysing the language Londoners use when they mention a candidate.

Overall, we can see that Sadiq has provoked more reactions on twitter than his Conservative rival. However, if we compare the share of tweets that have been positive and negative, the two are more or less neck and neck.

By looking at specific hashtags, we can see how the two main candidates have performed at the biggest hustings and debates of the campaign, including the City AM debate (Tuesday 12 April); Debate Tech (Tuesday 9 February); The Evening Standard hustings (Wednesday 3 February) and head-to-head debate (Thursday 21 April); the Greener London hustings (Friday 4 March); and ITV London’s broadcast debate (Sunday 24 April).

With the exception of the two ​Evening Standard ​events, Sadiq has ​secured more positive over negative twitter coverage than ​Zac on every occasion. This includes the Greener London hustings, where you would have thought the Conservative candidate ​– a prominent green campaigner and previous editor of the Ecologist – ​might have done better.

Zac’s biggest losses were at the City AM debate (85 per cent boos) and on ITV’s debate (78 per cent boos). Sadiq’s was at the Evening Standard hustings, where 91 per cent of tweets mentioning his account were boos.

Tweets and turn out

Despite thousands of tweets and trending hashtags, the truth is that the people who have been tweeting throughout election are a minority: the majority of Londoners have not paid much attention to the candidate’s campaigns. The 600,000 tweets sent during the election campaign were sent by just 100,000 users.  

Nevertheless social media has played a larger role in this election than in previous years. This means that, when it comes to understanding the relationship between Twitter and politics, we need to think beyond translating likes and retweets into turnout. Looking at how politics and Twitter interact is vital to our understanding of the how political engagement is changing.

At the end of this election campaign the new mayor should continue to use digital tools to sustain and further develop engagement. This digital engagement is best seen as an extension of, rather than a replacement of, conventional political engagement; it could include using online tools for participatory budgeting at a local level, for example. If the potential for digital engagement is to really be harnessed, it is imperative that online campaigning extends beyond polling day.

The nature of engagement between politicians and the public is changing. Whether or not the new mayor will maximise the benefits, and indeed negotiate the downsides of this relationship, remains to be seen.

Kat Hanna is research manager at the Centre for London. Alex Krasodomski-Jones is a researcher at the Centre for Social Media Analysis at Demos. 

Want more of this stuff? Follow CityMetric on Twitter or Facebook.