Cambridge Analytica: The future of political data is in the enterprise
Image: iStock / cofotoisme
TechRepublic recently reported on controversial tweets by former White House technology staffer Gerrit Lansing and former Trump campaign social media strategist Gary Coby alleging Cambridge Analytica overstated its role in the campaign and the capabilities of the firm’s technology. Specifically the company was accused of taking credit for Trump’s Facebook advertising strategy and his victories in Michigan and Florida.
More about Big Data
The tweets sparked a fiery debate about the role of big data in politics and vertical industries. Big data undeniably played a huge role during the 2016 presidential campaign, and after big elections political data innovations are often adopted by enterprise companies and SMBs. Some political technology firms—particularly partisan startups like NGP VAN and Targeted Victory —are now focused on local and regional races in the US. Other firms, like L2 and Cambridge Analytica, deploy their analytics product across enterprise verticals such as media, finance, and health care.
Last year in an interview with TechRepublic CEO Alexander Nix reiterated the company’s non-partisan status. “We are fundamentally politically agnostic and an apolitical organization,” he said. “The high volume of Republican primary candidates this cycle allowed us to enter a competitive market.”
Cambridge Analytica famously states its database contains over 5,000 data points on nearly every American consumer. The company is hardly the only big data company to make grand assertions about the power of analytics. Nearly every firm that TechRepublic spoke with while reporting this story agreed that big data is undeniably powerful and expressed concern that “magical” claims undermine the true value of analytics.
“Faulty products that exaggerate results can mismanage client expectations,” said a person familiar with the political technology industry. “Innovation only really succeeds if the product works. In any [business] sector, if one company exaggerates it harms the entire ecosystem. It erodes trust in the market.”
Cambridge Analytica refuted the allegations and clarified its role on the campaign in mid-February during a 30-minute phone conversation. Portions of the audio interview, however, were off record so we asked Cambridge Analytica’s head of product, Matthew Oczkowski, to respond to the allegations and clarify how their technology works.
Can you contextualize the controversy regarding the tweets from Lansing and Coby?
At CA, we break up our data into three buckets: political, commercial, and 1st party. We work with several of the main voter file providers, depending on the preference of our clients, to access vote history and voter profiles. We also access many of the top commercial data providers on a licensing basis; this data includes things like general demographics, geographics, purchase history, interests, etc. Finally, we collect data internally from R D projects like internal surveys and research, exclusive data relationships, and data collection through direct response projects.
Can you articulate what Cambridge Analytica’s technology is, how it works, and how you applied your tech for the Trump campaign?
At our core, we are a data and behavioral science company. Simply put, we help organizations figure out who to talk to and what to say to them. We believe that demographic-based marketing has become relatively obsolete and that understanding the underlying motivations of an individual are a far more effective way to communicate with someone. We deploy this research through both digital and television, which serve as the vehicles to reach these individuals.
For the Trump campaign, we served as the data agency of record, but our role quickly evolved as the cycle progressed. Our three core pillars of execution were data science and analytics, digital marketing (mostly persuasion and GOTV), and polling/research. Having a large amount of control and input into each of these three areas allowed us to be extremely efficient and reactive. It also allowed us to easily integrate with the staff at Giles-Parscale and the RNC.
Our approach allows our clients, like the Trump campaign, to more efficiently spend their resources and better persuade and mobilize their advocates.
Since data is the underpinning of every actionable insight that an organization would want to make, the applications of what we do are endless. With the Trump campaign, we assisted with everything from resource allocation for media buys, calculating the most efficient candidate travel stops, influencing the content in surrogate speeches, and personalizing messaging to the individual voter.
We learned that no two campaigns are alike and that there is no ‘playbook’ for the standard way to run a data, digital, or technology program on a campaign. Developing a custom program five months out from election day proved to have its unique challenges. Also, the style of President Trump as a candidate was far different from what we have seen with other candidates. This required us to changes some of our approaches to more quickly measure the attitudes of the electorate.
Hindsight is always 20/20, but going back and doing it all over again I would have traded money for time. I say this on most campaigns that I work on, but a bit more maturity in a few of the programs we set up may have yielded even better results. Given the outcome of the election, we are more than happy with the result.
At the end of the day, people are people. Convincing someone to show up and vote or donate for a political candidate is correlative with many issues businesses face everyday. Whether it be finding new customers, or improving brand loyalty, many of our techniques are suited to do just that.
Regarding your tech, does the firm build tech in-house, acquire tech from startups and other companies, or a mixture?
We acquire data in a number of ways – everything from commercial licensing, political exchange agreements, internal investment (R D projects), and proprietary relationships with partner companies.
We most certainly do build tech in house. We recently hired our CTO, Darren Bolding, who served as CTO of the RNC this past cycle. Darren and I manage, and are currently expanding, our in-house team of engineers to productize our offering a bit more. We aim to move more towards general automation of what we do in the form of a SaaS product.
Big data is a huge business trend. What’s your take on the current state of political big data? And what does the future look like?
We don’t consider ourselves a ‘big data’ company, we are a data analytics company. Many organization out there want to serve as the database of record for their clients. We aim to be the layer on top of that, which provides our clients with actionable insights – essentially acting as the brain behind the decision-making process.
We’ve come a long way, but there is still a lot of work to be done on the side of automation. Campaigns move so quickly that often it’s difficult to keep up. The more automation we can bring to the process outside of the campaign HQ will greatly increase our speed in delivering insights to the decision makers.