2 years ago
1 4
Structural Analysis
  Structural and Behavioural Analysis of #CameronMustGo

The Awedience data team were mulling over some ideas this morning whilst drawing the perfect espresso. Time for a quick check on Twitter’s current trending topics. Whilst Awedience in no way espouses any political affiliation, a hashtag did pique our interest from a data discovery perspective. So, like the Eye of Sauron we decided to turn Awedience towards the hashtag #CameronMustGo.

The sentiment and meaning behind the hashtag were clear, but we wanted to understand if there was any structure or coherence to those creating the content. We ignored anyone who was simply passing along existing content (retweets) and focused on those creating the messages. In order to do this we took a random sample of 1000 content creators from Twitter and passed them through various algorithms on our platform.

Even in this relatively small sample we found that people were still clustering into distinct sub-communities. Our platform identified 7 groups of any meaningful size. In the following sections we’ll categorise them.


Awedience has the ability to automatically identify clustering communities of people within groups of, sometimes seemingly random, connected Twitter users. It also uses algorithms to determine the importance of individual accounts within these communities and their behaviours based on past activities. The coloured graph images that follow show how people within a community are connected to one another and the relative size of each circle is proportional to its important within the community.

Communities Overview

Community Name Percentage Size Connection Density
Political Press & Action Groups 37% 3%
Broader Socialists 21% 21%
Labour Supporters 20% 11%
Wildlife and the Environment 6% 27%
TV & Politics 6% 15%
Twitter Trends 4% 21%
Peter George Owen hub 3% 80%

Political Press & Action Groups

Political Press and Action Groups Graph

The largest of the communities, but also the least connected making only 3% of all possible connections with each other. This lack of interconnectivity makes it difficult for this group to act together as their channels of communication are minimal. Their world view stretches outside the UK with topics such as Ferguson, Syria and Aleppo. The community is largely undemocratic with only 5 Twitter accounts holding the majority of the penetrative ability within the group:

Avatar Name ScreenName Followers
UK Uncut @ukuncut 77.5k
The Guardian @guardian 2.93M
ARTIST TAXI DRIVER @chunkymark 38.5k
James Walsh @jamesofwalsh 1,486
People’s Assembly @pplsassembly 18.3k

Key account: UK Uncut

Proportion of Replies: 2%
Proportion of Retweets: 56%
Proportion of Links: 47%
Proportion of Images: 23%
Proportion of Geo-tagged tweets: 2%

Most mentioned: ARTIST TAXI DRIVER

Favourite way to post: Twitter Web Client

Other trending hashtags: #ferguson #c4news #nhs

Most read news source: The Trust 24

Broader Socialists

Broader Socialists

Of all the communities, this one hits the sweet-spot of optimising both size and connection density. Not only do they constitute 21% of our sample set, but their connection density of 20% is far higher than would normally be seen in sparse social networks. It is a very democratic community with nearly everyone possessing an equal ability to spread ideas to everyone else. However, only 4% of their messages form part of a conversation. They have the potential for communal discussion, but they’re not using it. Largely focused on the UK with no significant international view, they also talk about health workers, banking and the Scottish Independence Referendum.

Key account: Amy

Proportion of Replies: 4%
Proportion of Retweets: 54%
Proportion of Links: 46%
Proportion of Images: 21%
Proportion of Geo-tagged tweets: 0%

Most mentioned: The National

Favourite way to post: Twitter.com

Other trending hashtags: #c4news

Most read news sources: The Mirror | The BBC

Labour Supporters

Labour Supporters Graph

Very similar to the Broader Socialists community in the way they describe themselves on Twitter. However, their structure and behaviours are different. With only half the connection density of the socialists, they have a far less democratic makeup with, similar to the Politic Press community, a handful of people holding the power:

Avatar Name ScreenName Followers
Dr Eion Clarke @laboureoin 36.4k
Labour Left @labourleft 20.4k
Jon Swindon @jon_swindon 2,908
Grahame Morris MP @grahamemorris 11.2k
Rosie R. @yorkierosie 2,697

Key account: Dr Eion Clarke

Proportion of Replies: 6%
Proportion of Retweets: 60%
Proportion of Links: 29%
Proportion of Images: 19%
Proportion of Geo-tagged tweets: 0%

Most mentioned: Jon Swindon

Favourite way to post: Twitter for Android

Other trending hashtags: #artravelivestream

Most read news source: The Trust 24

Wildlife and the Environment

Wildlife and the Environment Graph

A slightly unusual community in that 50% of people actively support wildlife and environmental issues, but they have ‘adopted’ a spoof David Cameron account. This is meant that non-wildlife accounts have also been pulled into this community. Again, for a very well connected community only 3% of their activity is conversational. When not engaging in the main hashtag of this article they are concerned about badger culling, hunting and Gadhimai, the world’s largest ritual slaughter ‘festival’ in Nepal. They take their news from The Guardian and The BBC, but also love videos on YouTube.

Key account: David Camoron

Proportion of Replies: 3%
Proportion of Retweets: 60%
Proportion of Links: 35%
Proportion of Images: 30%
Proportion of Geo-tagged tweets: 0%

Most mentioned: dominic dyer

Favourite way to post: Twitter Web Client

Other trending hashtags: #gadhimai #badgermonday

Most read news source: The Guardian

TV and Politics

TV and Politics Graph

The only community to contain a significant proportion of explicit Conservatives. They have tried to engage with the hashtag and invert it back towards Labour. They mainly use Twitter.com to publish, but they also have a large number of mobile usages (iPhone, iPad, Android and Windows Phone). This group has a broad world view including US, Ukraine, Russia and Iran. Their news sources are diverse (Guardian, BBC, Daily Mail, Huffington Post, Independent and The Mirror) and they also love YouTube.

Key account: Ban the BBC

Proportion of Replies: 7%
Proportion of Retweets: 67%
Proportion of Links: 30%
Proportion of Images: 2%
Proportion of Geo-tagged tweets: 0%

Most mentioned: Jack Monroe

Favourite way to post: Twitter Web Client

Other trending hashtags: #us #c4news

Most read news source: The Guardian

Twitter Trends

This community made up 4% of the overall sample but merely reported on the current trends automatically.

Peter George Owen hub

Peter George Owen hub Graph

A community dominated by Peter George Owen with a little backup from Aaron D of Unite. Less than 1% of all communications are conversations, yet 78% mention somebody. Nearly 4% of tweets are geotagged – the highest of any of the communities. The mainly use Twitter.com but, like TV and Politics, they are also very mobile oriented. There is good inter-community references as they frequently mention Dr Eoin Clarke (Labour Supporters) and ARTIST TAXI DRIVER (Political Press and Action Groups). As well as health worker issues there is also a strong Gaza, Israel and Palestine seam running through this community. Their news exposure comes via the Guardian, Independent and the BBC.

Key account: Peter George Owen

Proportion of Replies: 1%
Proportion of Retweets: 73%
Proportion of Links: 38%
Proportion of Images: 33%
Proportion of Geo-tagged tweets: 4%

Most mentioned: Jack Monroe

Favourite way to post: Twitter Web Client

Other trending hashtags: #nhsstrike #nhspay #gaza

Most read news source: The Guardian

Follow up

Got an interesting hashtag or niche market that you’d like our data team to help you explore? Get in touch with us here.

  1. Pete B

    3 years ago

    Interesting (didn’t know you existed :) ). Any data on where and when this hashtag originated – and its early growth rate? It seems to me to have been triggered by reporting on the Thornberry tweet. Just as the (perhaps related?) WeBackEd tag seemed to be a more-or-less instant and spontaneous response to the Guardian’s ‘plot to oust Ed’ article and follow-ups in social and mainstream media.

    • Chris Arnold

      3 years ago

      Hi Pete, thanks for the comment. We just took a quick snapshot of the data from a structural and behavioural standpoint. We don’t normally look at trends as we focus on BI for niche markets, but this looked interesting :). Personally, I think the Jack Monroe (@msJackMonroe) blow up may have been one of the key triggers for its virality. My hunch is that the Thornberry tweet isn’t connected to this, but I don’t have any data to support that.

  2. Gracie

    3 years ago

    The Emily Thornberry incident had nothing at all to do with the #CameronMusGo hashtag. We had two very successful hashtags before this one. #webackEd and #6monthstowin, each drawing big support and a high volume of tweets.

    The Telegraph cover the CameronMustGo hashtag and more or less accused all using the tag of being left wing bullies! It seems that if you are left leaning then you are not allowed to have an opinion let alone voice it. The Telegraph absurd analysis!

  3. Ed Fondue

    3 years ago

    The #CameronMustGo hashtag of November 2014 was instigated by @jon_swindon and @CharlieWoof81. It was trending well before the Jack Monroe “incident” and started 6pm 22/11/2014


Leave a comment

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>