Emerging Technology

Obama’s analytics chief speaks to SmartCompany: How to use big data to supercharge your business

Patrick Stafford /

Both of Barack Obama’s presidential campaigns in 2008 and 2012 have been widely praised for their use of data and analytics – technology is a huge part of why he was able to win the White House.

So how has a political campaign been able to master the effective use of data, while many businesses are still figuring out how to create decent websites?

Rayid Ghani, the research and analytics expert who ran the data science for Obama’s campaign in 2012, spoke to SmartCompany about this very problem – and says small business needs to lift its game.

“The core of what we did in the campaign was looking at all the data we had about the voters, and then used that data to predict future behaviour.

“So, for example, who [a person] might vote for, whether they would vote at all, and so on.”

Ghani, who is in Australia to speak at the Association for Data-driven Marketing and Advertising, says businesses haven’t yet cracked the code for taking consumer data and applying it to real problems.

“What you really want to do is look at your customer base, and use that to figure out who is likely to become a future customer, who is likely to switch from a competitor, and who is likely to increase their engagement.”

The use of technology within the Obama campaign is well-known within the analytics industry. The campaign’s analysis was so detailed, it could determine how long email subject lines should be to each individual recipient – and changed the length on each message accordingly.

A voter who didn’t open emails with long-winded subject lines would be sent emails with shorter subject lines instead, for instance.

Ghani says small businesses need to do the exact same thing with their customers.

Gaining the data is not the problem, he says. It’s taking that data and then applying it to practical solutions.

For that, he says, businesses simply need to experiment.

“It’s critical to build a culture and infrastructure for doing these types of experiments, and that is going to change over time as well.

“You can’t assume what you knew three days ago is the same thing that will work today. So you constantly have to be running experiments to know what you next move is.”

But in order to experiment, Ghani says businesses need to start hiring people dedicated to analysing data and targeting it in a more effective way – which will ultimately lead to more sales.

“It’s not as simple as just getting the data and then running some software. You need to know the people who are doing this actually know what they’re doing.

“There’s also an over-reliance on data without knowing where it came from. You need to know the limits of your data.”

Sometimes, he says, gathering data about your customers can lead to you to make assumptions.

“Getting the data and then knowing not to make a prediction about something is just as important as actually making it,” he says.

Finally, he says, businesses need to start turning their followers on social media and their customers into enthusiastic brand advocates.

“One of the things we did well in the campaign was being able to take our enthusiastic supporters, and turn them into volunteers.

“That may be something I haven’t yet seen organisations do. You need to take your best customers and get them to upsell for you.”

Advertisement
Patrick Stafford

Patrick Stafford is a freelance journalist and a former deputy editor of SmartCompany.

We Recommend

FROM AROUND THE WEB