The Internet, Policy & Politics Conferences

Oxford Internet Institute, University of Oxford

Neil Bendle, Joseph Ryoo: Understanding Candidates’ Positioning Strategies Through Twitter

Neil Bendle, Ivey Business School, Western University, London, Ontario, N6G 0N1, Canada

Joseph (Jun Hyun) Ryoo

We examine the strategies of candidates seeking their party’s nomination for the 2016 U.S. presidential election. These candidates are particularly interesting as, unlike candidates in U.K. politics, the candidates have their own strategies which can be quite distinct from that of the party. Indeed some candidates are thought to damage the party brand. We approach the study from a political marketing perspective seeking to understand the positions that the candidates take and the messages that they send.

Our focus is on the use of Twitter, the microblogging site (Bode and Dalrymple 2015; Jackson and Lilleker 2011; Aharony 2012; Conway, Kenski, and Wang 2013). This site is useful for political candidates as it disseminates messages cheaply and directly to voters. Done especially well, or especially badly, Twitter also earns traditional media mentions. For example, Donald Trump’s controversial tweets about the other candidates, debate moderators, and even the Pope garnered widespread television and newspaper coverage in the U.S and beyond. 

Our data is all the tweets, retweets, and favorites of the major party candidates ‘3 Democrats (Clinton, Sanders and O’Malley) and 12 Republicans (Bush, Carson, Christie, Cruz, Fiorina, Gilmore, Huckabee, Kasich, Paul, Rubio, Santorum, and Trump). We have captured all this Twitter activity from January 2016 and will do so until the end of each candidate’s campaign. We will provide full descriptive statistics (Vergeer, Hermans, and Sams 2011) to show the activity, e.g., number of tweets, and the achievements of each candidate, e.g., number of followers, and visualize the data to aid comprehension.

Traditionally political marketing research uses case studies (Miller 2013), interviews (Parsons and Rowling 2015), and manual coding (Graham et al. 2013). We will employ text mining, (Bendle and Wang 2016), specifically using R and its associated lexicons. This will allow us to look at themes, such as healthcare and immigration, in all the messages from each candidate. We will assess which terms are relatively popular in each candidate’s tweets by comparing the popularity of terms with a given candidate against the popularity of terms in the entire corpus of all the candidates’tweets. This will show how themes highlighted differ within, and between, parties (Enli and Skogerb° 2013). 

We will use sentiment mining on each candidates’tweets. For example, we will compare Hillary Clinton’s and Bernie Sander’s positivity. Thus our sentiment analysis does not consider, as might be more common, the mood of the general Twitter users. Instead we will uncover the mood displayed by each candidate through their messages.

Our theoretical approach will seek to understand the positioning and messaging of the candidates in primaries (Bendle and Nastasoiu 2014; Bendle 2014). We will examine the suggestion made by the media that the other Republicans somewhat avoided discussing Trump in the early part of the campaign. It was suggested that the other Republicans focused more on criticizing each other than Trump. The strategic thinking underlying this focus being the belief that there was an anti-Trump majority to be inherited when only one non-Trump candidate remained viable. Candidates thus targeted rivals further down the polls who seemed vulnerable to being driven out of the race. At time of writing (early March) this appears to have significantly changed. We will confirm if this borne out by the data.

We are compiling a complete database of the Iowa Electronic Nomination Market prediction prices from the market’s start, January 2016. This gives daily ‘wisdom of the crowd’assessments of the likelihood of major candidates winning the eventual nomination. This gives us a measure of changes in perceived performance. When compared to the tweets we shall see how perceived progress, or setbacks, impact messaging. (We will do a similar task in comparing Twitter strategies to polling data from Real Clear Politics).

Overall, we can assess the focus of candidates’Twitter strategies and help understand how consistent, even well-managed, the strategies are. From this we can theorize about the attention that candidates are paying to Twitter and how much the platform has impacted each candidate’s marketing strategy.


Aharony, Noa. 2012. ‘Twitter Use by Three Political Leaders: An Exploratory Analysis.’Online Information Review 36 (4): 587’603.

Bendle, Neil T, and Mihaela-Alina Nastasoiu. 2014. ‘Primary Elections And US Political Marketing.’In Political Marketing In The United States, ed. Jennifer Lees-Marshment, Brian Conley, and Kenneth Cosgrove, 85’111. New York, NY: Routledge USA.

Bendle, Neil T. 2014. ‘Reference Dependence in Political Primaries.’Journal of Political Marketing 13 (4): 307’33. Bendle, Neil T., and Xin Wang. 2016. ‘Uncovering The Message From The Mess Of Big Data.’Business Horizons 59 (1): 115’24.

Bode, Leticia, and Kajsa E. Dalrymple. 2015. ‘Politics in 140 Characters or Less: Campaign Communication, Network Interaction, and Political Participation on Twitter,.’Journal of Political Marketing.

Conway, Bethany Anne, Kate Kenski, and Di Wang. 2013. ‘Twitter Use by Presidential Primary Candidates During the 2012 Campaign.’American Behavioral Scientist 57 (1596-1610).

Enli, Gunn Sara, and Eli Skogerb°. 2013. ‘Personalized Campaigns In Party-Centred Politics Twitter And Facebook As Arenas For Political Communication.’Information, Communication & Society 16 (5): 757’74.

Graham, Todd, Marcel Broersma, Karin Hazelhoff, and Guido van ‘t Haar. 2013. ‘Between Broadcasting Political Messages And Interacting With Voters.’Information, Communication & Society 16 (5): 692’716. 

Jackson, Nigel, and Darren Lilleker. 2011. ‘Microblogging, Constituency Service and Impression Management: UK MPs and the Use of Twitter.’The Journal of Legislative Studies 17 (1): 86’105.

Miller, William. 2013. ‘We Can’t All Be Obama: The Use of New Media in Modern Political Campaigns.’Journal of Political Marketing 12 (4): 326’47.

Parsons, Michael, and Martyn Rowling. 2015. ‘Social Media and the Paradox of Fear: An Exploratory Study of Political Relationship Marketing Within South Wales.’Journal of Political Marketing.

Vergeer, Maurice, Liesbeth Hermans, and Steven Sams. 2011. ‘Online Social Networks and Micro-Blogging in Political Campaigning: The Exploration of a New Campaign Tool and a New Campaign Style.’Party Politics 19 (3): 477’501.

Neil Bendle, Joseph (Jun Hyun) Ryoo