CHAPTER
5

The Online Campaign

CHAPTER
5

The Online Campaign

  • Online and social media battles were critical in this election
  • The Conservatives’ campaign was a dramatic improvement on 2017
  • Labour’s online campaign reflected a general failure of integration and innovation
  • Labour’s online campaign was weakened by a lack of professionalism and capacity
  • Labour’s online campaigning suffered from poor internal coordination
  • Labour’s online supporters did not reach or win over a wide enough range of voters

Online and social media battles were critical in this election

The importance of new online and social media in influencing voters’ perceptions and behaviour has risen rapidly over recent election cycles. Social media now plays a key role in amplifying messages placed in traditional media and in influencing what traditional broadcast and print media covers during election campaigns. The potential for narrowly targeted advertising, as well as the “organic reach” achieved when content is voluntarily shared, can make it a highly cost-effective form of political communication. 

The level of online campaigning and advertising in this campaign was unprecedented. Labour declared its central focus to be “making viral persuasive content – bypassing the media and breaking out of the bubble”.[109] Significant resource was dedicated to this strategy:

  • Labour employed digital media teams in both Head Office and the Leaders’ office, as well as consultants and groups across the Party.
  • Labour spent heavily on paid online communications – including £1.4m on Facebook advertising – more than the Conservatives’ £900,000 (though this figure does not include spending from outrider pages or on sponsored posts)
  • Labour invested in “Promote” software designed to enable narrow targeting of individuals’ social media timelines
  • Labour could call on the support of numerous allied organisations and individuals active on social media, many with very large followings

Successes were claimed by Labour’s digital teams, including:

  • Over six million views for two videos
  • 1.5 million likes for Jeremy Corbyn on Facebook and 2.2 million followers on Twitter
  • More shares and retweets for Jeremy Corbyn than Boris Johnson or the Conservatives
  • A large scale mobilisation of members and supporters on the ground in which digital communications clearly played a role

However, the evidence we have received shows that overall the Conservatives were far more effective in this arena.

In preparing this chapter, we were able to draw on

  • a review of Labour’s campaign, based on an investigation of Party structures and research interviews with key staff and campaigners commissioned from campaign cooperative, Common Knowledge[110]
  • expert assessment and comparison of the Conservative and Labour digital campaigns commissioned from digital communications agency Valent Projects[111]
  • analysis of the role of online outriders in the campaign commissioned from the Centre for Countering Digital Hate (CCDH)[112]
  • insights of Labour Together’s 11,060 survey respondents and 30 interviews with defeated Parliamentary candidates

The Conservatives’ campaign was a dramatic improvement on 2017

Research conducted for this Commission by Valent Projects found the Conservatives were determined to improve on their 2017 campaign, and invested in infrastructure, skills and expertise, developing a comprehensive and multi-faceted strategy to beat us next time.

Key elements of this included:

  • a strategic focus on digital media as central to engaging and persuading voters outside their traditional supporter groups
  • a clear senior role for the digital team, based alongside the head of the campaign in the election ‘war room’
  • a professional team who knew how to create quality content most likely to gain reach through algorithms and organic shares, as opposed to paid advertising
  • less emphasis on narrow targeting of individuals, more on reach among broader groups of voters – including through banner ads on newspaper websites or Google and YouTube, where they spent significantly more than Labour[113] 
  • a creative focus on finding fresh and controversial ways of reiterating and reinforcing the same core messages 
  • deliberate exploitation of the effect social media can have in increasing the “fluency” of talking points repeated on print or broadcast media or enabling “distributed spin” on running stories – made possible by the close coordination of digital and news functions at the centre of the campaign
  • a willingness to allow creative experimentation with content, without slow or cumbersome sign-off procedures
  • intensive testing of content for impact, that began when the new Leader took office in July 2019 and continued throughout the campaign; we were told the Conservatives tested 11,000 different ads compared to 2,500 tests by Labour in the same period
  • a system of support for candidates in 50 target constituencies, provided by consultancy firm Westminster Digital, that allowed them to create their own, personalised, localised content to a professional standard that was reinforced by national messaging and themes.[114] Analysis by Valent Projects suggests that, since August, the company succeeded in raising their clients’ social media followings by between ten and twenty per cent
  • a network of supportive outriders that succeeded in drawing voters away from rival parties – from abrasive groups like the Campaign Against Corbynism Rebel Media, which was seen as drawing Labour Leave voters into their orbit, to less obviously political groups such as Working4UK or Parents’ Choice. According to Valent Projects’ research, as a rule, the groups did not acknowledge a link to the Conservative Party, although investigation often revealed a Conservative Councillor[115], official or lobbyist[116] behind each.

Over the summer, and well ahead of the election, the Conservative Party was running hundreds of ads with slight differences in colour, wording and even using different emojis. Although Facebook’s portal for political advertising doesn’t show what criteria advertisers are using to identify those they want to reach, the content of the ads and relatively small budgets (often under £100) suggests they were being targeted at individual constituencies.

By using the statistical feedback provided to ad buyers on the performance of ads, the near £100,000 that the Conservative Party spent on the ads in the summer of 2019 would have bought immense amounts of constituency-level data on what messages work best with different groups of people, based on their ages, gender, area of work and political affiliation.[117]

By the time the election campaign began, the Conservatives were able to capitalise on the information they had accumulated. Their ads have been described as “laser-guided” – using neon graphics and up-tempo music to push a “Get Brexit Done” message to 300,000 men under 34; classical music, softer colours and additional pledges on the NHS and crime for 350,000 women over 55.[118] Towards the end of the campaign, the Conservatives ramped up its Facebook advertising, with 7,000 ads in early December – 90 per cent of which contained misleading claims, according to analysis by First Draft.

Clearly this approach proved highly effective, particularly in exploiting negative perceptions of Labour’s Leader, with many candidates and campaigners reporting the impact of a surge in negative online campaigning in the final week of the campaign.


This was reinforced by “outrider” activities. Monitoring group “Who Targets Me” identified nine non-party groups that spent around £300,000 on Facebook ads in the month before the poll.[119]


Local groups may also have had an important influence. For example, research conducted for this Commission by Centre for Countering Digital Hate identified a Facebook Group in Dudley which built followers by posting local news which hosted a large amount of anti-Labour and anti-Jeremy Corbyn content, with “comments” being used to organise protests against Jeremy Corbyn’s visit to a local pensioners’ club during the campaign; this story later appeared in The Sun, with the headline “Jeremy Corbyn heckled as ‘dirty IRA scum’ when he arrives in key Dudley marginal.”[120] The account hasn’t posted since 12 December 2019.[121]

Labour’s online campaign reflected a general failure of integration and innovation

Reviews we commissioned from Valent Projects, CCDH and Common Knowledge show that, despite all the activity and resources invested, Labour’s online campaign fell well short of the Conservatives in this election. The primary source of this shortcoming was the absence of strategic integration and structured innovation, identified as a fundamental organisational and cultural weakness in the previous chapter.  

Labour lacked an imaginative strategy for digital campaigning. Online output was siloed off from broader strategy and communications, instead of being centrally integrated. The Party’s social media channels simply became an additional broadcast platform, rather than a dynamic and responsive tool for targeting, engaging and persuading key groups of voters. Our communications with the voters remained one rather than two way, seriously limiting their effectiveness.

Valent Projects and CCDH found a lack of focus and – reflecting broader criticisms of Labour’s communications effort – no consistent messaging strategy. The Conservatives’ core message “Get Brexit Done” lent itself easily to the “distributed spin” approach, whereby supporters can pick up and disseminate frames and messages and help to shape audiences’ views of the campaign and related news stories. By contrast, the proliferation of different messages and policies in Labour’s online campaign made it harder for online supporters to find a core message to reinforce, and made it easier for them to get diverted into less helpful activities such as criticising the BBC’s election coverage.[122]

As in other areas, Labour was not well-prepared for the online campaign when it began. Valent Projects’ research shows that the Conservatives began using Facebook to stress test messaging and imagery in the summer, as soon as Boris Johnson was elected as their Leader. Labour only began similar testing once the election had been called. In August, independent online monitoring group “Who Targets Me” noted that, while the Conservatives were investing heavily in data collection and attack ads, Labour’s strategy was “unclear” and “notably less disciplined”. [123]


Labour’s online campaign was weakened by a lack of professionalism and capacity

According to the findings of the reviews we commissioned, the fast-moving nature of online campaigning saw Labour’s collective levels of skill and understanding fall well behind the curve. Labour does not have enough politicians, staff or activists up-to-speed on digital campaign techniques – something the Conservatives addressed by handing most of it over to a commercial consultancy.

Symptomatic weaknesses included:

  • a senior level focus on limited metrics such as number of shares, rather than more meaningful measures of impact and quality of interactions
  • a continued focus of attention by MPs and parliamentary candidates on Twitter, despite the far greater importance of other platforms (such as Facebook or Instagram) for the voters they need to reach
  • a failure to understand how effective content and ongoing optimisation for engagement, in relation to algorithms operated by platforms like Facebook, can result in far greater views and reach than static and costly paid ads 
  • an insufficient commitment to message experimentation and testing; available online evidence suggests Labour conducted far fewer content tests than the Conservatives, and tended to settle on content, rather than testing continuously
  • overemphasis on Twitter and likes for Facebook pages, both of which focus on one-way communication, and the failure to use forums like Whatsapp groups and Facebook groups more effectively. For example, Labour support among Hindu voters fell significantly in this election, due to the extensive sharing of anti-Labour content across a network of Whatsapp groups. Labour has yet to work out how to respond to this more invisible and engagement-driven form of social media campaigning[124]

Labour’s online campaigning suffered from poor internal coordination 

Labour’s confused messaging was compounded by the number of people involved in producing or approving content without any shared framework or strategy. It’s clear from all submissions that little thought was given as to how different platforms, channels and messengers across the Party could be used effectively to reach different audiences, both members and voters. 

An organogram prepared for this Review by Common Knowledge shows multiple social media teams across the Party, without clear lines of communication or coordination with each other. The Leader’s Office established a separate social media team of eight, which had no clear relationships to the team in HQ. This led to internal battles, leaked reports and frustration on both sides.[125] The involvement of an agency in message testing further complicated the split responsibilities.

The Commission heard that the Party’s digital campaign team were held back by a lack of creative freedom, and that cumbersome sign-off processes were a problem, despite it being public knowledge that the Conservatives knew a key weakness of their 2017 digital campaign had been slow and unresponsive decision-making.[126] We were told there was a significant time lag between social media content being produced and it then being shared on the Party’s social media channels, reducing Labour’s ability to manoeuvre in a fast-moving online environment.

Problems of leadership and coordination at the centre were compounded by insufficient guidance, support and leadership given to Parliamentary Candidates and CLPs, who were left to devise their own strategies, without professional support or a framework of consistent messaging.

Labour Party staff certainly did their best – for example, the Campaign Office for Rosie Duffield’s successful campaign in Canterbury “welcomed the exceptional support with social media from the Regional Office team”.[127]

However, lack of coordination from the centre meant the level of social media support available to candidates varied, depending on the resource available to regions and nations. The digital campaign was highly centralised, with little resource given to Regional Offices for content production and minimal training or support for candidates to produce content. Many Labour candidates and activists did not know how to use the Promote tool, or found it difficult and cumbersome. Candidates often did not understand how to use Facebook effectively, let alone shoot quality videos or create shareable content. One defeated MP complained that the Party did not “inform or consult us on what Facebook/social media they were directing and to which voters, so our own could complement these”.[128] Another told us “the internal mechanics of getting social media to work was terrible in 2017 and 2019… Real difficulties with Promote – we weren’t able to use it in either election”.

This contrasts strongly with the effective use by the Conservatives of a professional consultancy to support candidates’ social media campaigning in key constituencies, which the review we commissioned from Valent Projects showed to be a key ingredient in the Conservatives’ online success.

Labour’s online supporters did not reach or win over a wide enough range of voters

Labour-friendly organisations and individuals with large social media followings clearly played a role in engaging and mobilising Labour activists and supporters.

Momentum, for example, believe they “significantly outperformed the Party in terms of member engagement and mobilisation while also filling substantial gaps in Labour’s campaign”. This included what they saw as a “dramatically” improved social media performance, including:

  • a targeted online advertising campaign that resulted in the registration of approximately 201,000 voters in key marginals
  • 106 million social media video views, more than twice the level of their 2017 campaign

However, it is unclear what effect many Labour-friendly organisations and individuals had in broadening Labour’s support. One digital consultant commented: “While organically Labour did really well…most of the views and shares came from people who were already going to vote Labour.”[129]

The review conducted for us by CCDH showed that many alternative media sources have declined in reach since the 2017 election, partly due to changes in Facebook’s newsfeed algorithms, which gave greater prominence to content posted by friends and family and discussions taking place within Facebook groups or forums, and less to “followed” pages.[130]

These outlets also tend to operate in silos, focused on internal Labour Party processes and debates taking place within “the Left”. They were not focused on reaching across political divides, to target particular demographic or interest groups, or undecided voters. 

We analysed ten of the most active Labour supporting Twitter accounts: in total, these accounts collectively have 667,600 followers, but their unique reach is half of this, as they collectively only have 373,800 unique followers. Of those 373,800 unique followers: 257,000 follow just one of the ten accounts, while 115,800 follow multiple accounts. This shows that the large following of many “left-wing” Twitter accounts could actually be the result of a much smaller number of accounts just following each other, resulting in little reach beyond those who are already supportive of Labour.[131]

Figure 42: Twitter audiences of “Left outrider” accounts

A screenshot of a cell phoneDescription automatically generated

Source: Correct as of 30th March 2020, data extracted from Twitter


As indicated above, there was not a strong strategy for creating content for supporters that would likely be picked up and shared further to broader audiences. Occasions when Labour’s Twitter supporters reached the widest audiences were often in the context of issues unlikely to win over new supporters, such as disagreements over the Party’s handing of anti-Semitism or criticism of the BBC’s reporting of the election.[132]  

OUR COMMISSION CONCLUSIONS AND KEY FINDINGS

  • Despite all the activity and resources invested, Labour lost the online campaign in an election where it was more important than ever before. 
  • Whilst the Tories learnt from their failure online in 2017, Labour did not invest and strengthen its online capacity, making use of the brightest and the best available. The groundwork was not properly laid to test strategy, tactics or messages ahead of the 2019 election campaign.
  • Online campaigning was hamstrung because it was siloed off from broader strategy and communications rather than centrally integrated. Poor internal coordination, exacerbated by factional tensions, resulted in slow, inconsistent decision making and an inability to act quickly.
  • Our digital infrastructure was underfunded and inadequate. Candidates and local party campaigners found it very difficult to access and use the tools or support necessary to wage the campaign online consistently enough. Some of these systems were creaking in 2017, but the lack of internal reflection meant that issues went unresolved.
  • Not enough was done to rebut attacks in digital spaces or elsewhere against the leadership, our Brexit position, or to reassure people about our policies and plans for the country. The Party’s social media channels became simply an additional broadcast platform rather than a dynamic and responsive tool for targeting, engaging and persuading groups of voters.
  • Labour’s supporters online spent too much of the campaign talking to themselves rather than reaching out to convince swing voters to support Labour. In contrast the Tory online presence was vastly improved from 2017, at national and local level, using proxies to attack Labour and build support for the Conservative campaign in key seats. The Conservatives central message of “Get Brexit done” lent itself to their very effective approach to organic shares and “distributed spin” online.
  1. ‘Labour runs most successful social media campaign ever seen’, Labour Party Press Release, 11 December 2019: https://labour.org.uk/press/labour-runs-most-successful-election-social-media-campaign-ever-seen/
  2. Common Knowledge: https://commonknowledge.coop/
  3. Valent Projects: https://www.valent-projects.com/
  4. Centre for Countering Digital Hate: https://www.counterhate.co.uk/
  5. Mark Bridge and Billy Kenber, ‘Election 2019 result: how social media gave Labour false hope’, The Times, 14 December 2019: https://www.thetimes.co.uk/edition/news/election-2019-result-how-social-media-gave-labour-false-hope-5whzhnt3s
  6. ‘Under the radar: how do politicians use social media?’, Sky News, 8 December 2019: https://youtu.be/eKIGj5tHpew
  7. Will Hayward, ‘How a Conservative politician uses Facebook to disguise party political smears on Jeremy Corbyn”, Wales Online, 28 November 2019: https://www.walesonline.co.uk/news/politics/conservatives-facebook-corbyn-general-election-17324805
  8. Rowland Manthorpe, ‘General election: women accused after posting ads costing £17,000 on Facebook attack adds’, Sky News, 7 December 2019: https://news.sky.com/story/general-election-woman-accused-after-posting-ads-costing-17-000-on-facebook-attacking-labour-11880162
  9. Who Targets Me, ‘What do Facebook Ads tell us about Parties’ UK General Election Strategy?’, Medium, 15 August 2019: https://medium.com/@WhoTargetsMe/what-do-facebook-ads-tell-us-about-parties-uk-general-election-strategy-af84dddb675: Who Targets Me, ‘Conservatives’ Boris ads primarily targeting men over 45’, Medium, 5 August 2019: https://medium.com/@WhoTargetsMe/conservatives-boris-ads-primarily-targeting-men-over-45-1b3a6511d8fe
  10. Beth John and Carlotta Dotto, “UK Election: how political parties are targeting voters on Facebook, Google and Snapchat ads”, First Draft, 14 November 2019: https://firstdraftnews.org/latest/uk-election-how-political-parties-are-targeting-voters-on-facebook-google-and-snapchat-ads/
  11. Mark Bridge and Billy Kenber, ‘Election 2019 result: how social media gave Labour false hope’, The Times, 14 December 2019: https://www.thetimes.co.uk/edition/news/election-2019-result-how-social-media-gave-labour-false-hope-5whzhnt3s
  12. Holly Christodoulou, ‘“SCUM!” Jeremy Corbyn heckled as “dirty IRA scum” when he arrives in key Dudley marginal’, The Sun, 22 November 2019: https://www.thesun.co.uk/news/politics/10396246/jeremy-corbyn-ira-scum-dudley-votes/
  13. Campaign for Countering Digital Hate research for Labour Together
  14. Campaign for Countering Digital Hate research for Labour Together
  15. Who Targets Me, ‘What do Facebook Ads tell us about Parties’ UK General Election Strategy?’, Medium, 15 August 2019: https://medium.com/@WhoTargetsMe/what-do-facebook-ads-tell-us-about-parties-uk-general-election-strategy-af84dddb675
  16. Rowland Manthorpe, ‘General election: Whatsapp messages urge British Hindus to vote against Labour’, Sky News, 8 November 2019: https://news.sky.com/story/general-election-whatsapp-messages-urge-british-hindus-to-vote-against-labour-11856692
  17. Ned Simons, ‘Jeremy Corbyn’s digital team hits out at Labour Party’s Online Election Performance’, Huffington Post, 21 January 2020: https://www.huffingtonpost.co.uk/entry/jeremy-corbyns-digital-operation-labour-election-performance_uk_5e309224c5b6e8375f641e75
  18. Tim Ross and Tom McTague, Betting the House: The Inside Story of the 2017 Election, Chapter 9, Biteback, 2017.
  19. Pete Smith, written submission to Labour Together
  20. Helen Goodman, written submission to Labour Together
  21. Mark Bridge and Billy Kenber, ‘Election 2019 result: how social media gave Labour false hope’, The Times, 14 December 2019: https://www.thetimes.co.uk/edition/news/election-2019-result-how-social-media-gave-labour-false-hope-5whzhnt3s
  22. Rowland Manthorpe, ‘The UK’s left is scrambling to adapt to Facebook’s algorithm change’, Wired, 20 March 2018: https://www.wired.co.uk/article/facebook-algorithm-changes-engagement-labour
  23. Labour Together analysis, correct as of 30 March 2020, data extracted from Twitter.
  24. Campaign for Countering Digital Hate research for Labour Together.
FOREWORD
PART ONE: A HISTORIC DEFEAT
part two: LABOUR’S CAMPAIGN
PART THREE: THE WAY AHEAD
ACKNOWLEDGEMENTS
APPENDIX