Mining for personal data has been going on for a while
The Cambridge Analytica scandal has opened a can of worms. But let’s start with Blue State Digital1. The company was founded by former staffers of Howard Dean’s 2004 US presidential campaign and subsequently provided “digital services” for the 2008 and 2012 US presidential campaign for Barack Obama.
So, what does Blue State Digital’s software do for a political campaign? Well, it’s sort of like a large-scale CRM system with programmable automation rules. For instance, if someone donates $5 to the campaign, this person will then be automatically targeted with a string of messages that is different from the string of messages that you would be targeted for if you donated, say, $50 instead.
Donating any amount of money is, of course, one way to interact digitally with a political campaign. But there are many other triggers, too. You could interact with the candidate on social media, sign up for his or her newsletter, or be affiliated with the party or the president in some other shape or form2. I first came in contact with this immense power in 2007 when I was responsible for launching a Swedish online service to verify broadband speeds (Bredbandskollen) and later in 2009 when my employer at the time, Springtime, acquired Early October to use their proprietary technology Social Media Lounge for online monitoring.
A few years down the line, around 2012, everyone was talking about the potential of “big data” — but not many people truly understood the implications.
The power of going beyond basic demographics
Cambridge Analytica was founded in 2013 and their business model was quite similar to that of Blue State Digital with a focus on data mining, data brokerage, and data analysis — and some high-end consulting on top. (This is also similar to social media intelligence agency Whispr Group, founded in New York by Joakim Leijon in 2010 where I served as the COO between 2010-2013.)
Segmenting people on basis of their communicative behaviour is extremely powerful compared to traditional demographic segmentation3. But it wasn’t only the data mining companies that could sense potential. So did the social networking sites. As Facebook did their IPO in 2012, they they took the lead and embarked on an aggressive journey to monetise social media usage.
In short: Facebook went hard for the advertising dollar.
As such, advertising is hardly a new monetisation model, but Facebook took the route of modernising the self-servicing targeting functions instead4. With Facebook moving forward aggressively, Cambridge Analytica decided to take a shortcut.
The epic proportions of the Cambridge Analytica scandal
An external researcher approached Facebook and told them he was collecting data for academic purposes. For his “study”, he launched a Facebook app called Your Digital Life. The app was used by some 270,000 users who gave the app permission to collect data on their friends which allowed the app to mine data on some 87M users. And then, somehow, this data ended up in Cambridge Analytica’s systems.
Still, there’s more to the Cambridge Analytica story:
One of Cambridge Analytica’s founders and investors were conservative Steve Bannon. Bannon divested his holdings in the company in April 2017 when he was appointed White House Chief Strategist, but at that point, he had already been working as the CEO for Donald Trump’s presidential bid sin August 2016. And what software did they use to persuade the American opinion to vote for Trump?
Cambridge Analytica, of course.
Social media users are beginning to catch on
To run a US presidential campaign on the back of illegally (and unethically) acquired data was, of course, a scandal in its own right. However, what really got under people’s skin was the actual data analysis:
With so many data points on US citizens, Cambridge Analytica were able to match behavioural data with psychographic characteristics. From psychographic studies and test results (such as Myers-Briggs and The Big Five Aspects Scale), it’s possible to assign people to groups not just based on how they have behaved in the past, but also to how they’re predicted to behave in the future. It gave them answers on who to target and how to trigger them psychologically.
The whole scandal surrounding Cambridge Analytica brought the general public and politicians one step closer to understanding the immense power of big data. Before Cambridge Analytica, many people believed that the power of big social networking companies had more to do with their direct access to people’s attention.
Suddenly, more people started to grasp the real power of the “Silent Miners”.
The techlash has gained momentum and it won’t stop
Expect a power struggle between the big tech companies on the on side — and legislators and the news media on the other. In the wake of the scandal, Cambridge Analytica closed their operations in 20185. But the techlash 6 have already gained too much momentum to be stopped.
This is a war for our minds:
Shall we be social media optimists or shall we be social media pessimists?
- “We move people to elect presidents, change laws, fall in love with brands, donate millions, and more.”
- A number of Scandinavian political organizations became rather interested in Blue State Digital and their type of technological advancements. However, these types of services will work best in democracies with a large number of people. My guess is that many nations much smaller than the US did test runs with similar softwares, but decided to go back to more traditional campaign tactics.
- For a more in-depth explanation of behavioural segmentation, see The Public in Publics Relations).
- Anyone involved in programmatic advertising should already be more than familiar with Facebook’s level of sophistication when it comes to Business Manager targeting.
- Several executives moved to Emerdata, owned by the same parent company and residing in the same building in London.
- According to Financial Times: “(noun) The growing public animosity towards large Silicon Valley platform technology companies and their Chinese equivalents.”