Just Giving has raised over £4 billion for good causes and has been used by 25 million people across 160 countries. Chief analytics officer Mike Bugembe puts this success down to digitising the simple paper form to make fundraising a richer, more shareable experience. In 2008, Just Giving began investing in big data and data science with a mission to “grow the world of giving”.
The first thing Bugembe realised was that the “big boys” are in a league of their own. He said that Google, Facebook and Amazon are always mentioned in the same breath as big data because they are “great examples of data science”. However, it would be difficult for a company with a much smaller budget to emulate them.
So Bugembe looked to other industries for inspiration - smaller players in sport and law enforcement. Catapult uses data science to reduce the number of injuries sustained by sportsmen and women and Mark43 is a cloud software system that helped to reduced crime in New Jersey by 40%.
The Just Giving data team then decided to look at the digitised fundraising page in the same way a machine does. “That meant seeing everything on our page as some form of data. We were capturing the number of fundraisers and supporters, the amount they gave, the charity they were raising for and the target amount,” said Bugembe.
The team found that some users would continually refresh a fundraising page to see what percentage of the target had been reached. Of those, a few would want to be one of the first givers and be less likely to donate once 5% of the target had been reached. Others would wait until the target had risen to 95% or more so they could sweep in as the “hero giver” and donate an amount to get the campaign to hit its target.
“Using data to understand each of those individuals allows us to give them a really personal experience that increases their propensity to give,” said the CAO. Bugembe explained they collected observational and experimental data as well and realised that connections were the “real gold”.
“Generosity, influence and opinions all flow through connections,” he said. The team was able to run calculations on the information that flows through networks and find out who is likely to be an influencer. And by using machine learning and graph theory, Just Giving was able to find out what was interesting and personal to the givers. “The machine now does a lot of work around consideration, around making the person want to give, fundraise or engage. It figures out what you care about and understands how to keep you engaged.”
As a result of the work done by Bugembe and his team of analysts, statisticians and data scientists, Just Giving is able to be much more helpful to its fundraisers. “We know when is the right time to send email, what to send and who to send it to,” he said and calculated that up 60% of customer relationship management is powered by the machine.
And the work to uncover more insights continues. Bugembe said Just Giving is working with Pivigo on a machine to identify which images with which features are most useful to display at which time.
Setting out on the data journey has allowed Just Giving to help its users with their giving and raising journeys. Said Bugembe: “It’s about finding out what people care about and taking them along a journey so that they better understand what it is you’re doing.”