Last October at our annual product and customer experience conference Canvas we had the pleasure of having Mike Bugembe, Chief Analytics Officer at JustGiving, to talk to us about how big data can work for a variety of organisations, big and small. When we think about big data, either Facebook, Google or Amazon typically come to mind; rarely do we hear about or recall the successes of other companies that may be either smaller, less technical or purely traditional.
The good news is that success in big data is not restricted to large, purely technical organisations. Mike gave the example of Black Dog Institute, who have been using big data and analytics to great effect in their work around early detection of depression in war veterans.
But 70% of investments in data science are failing.
Getting value from data is time consuming and difficult, with many obstacles and pitfalls that organisations face along the way. A big issue is the lack of resources around an end-to-end process that leaves teams not knowing how to go about working with big data or how to collaborate effectively.
Mike’s talk gave us an inside view on how JustGiving, a small technical company, managed to harness the power of big data and apply it to successfully begin to ‘grow the world of giving’.
They could use data to understand what was happening, but they wanted to understand why it was happening.
JustGiving started by digitising the traditional charity sponsorship form through the power of the internet, helping people to fundraise for causes they cared about. In order to work towards their ambition of ‘grow the world of giving’, the first thing they had to do was understand the psychological mechanics behind ‘giving’.
They observed, captured and analysed huge amounts of behavioural, experimental and transactional data to better understand how people make decisions around ‘giving’. Uncovering user motivations and anxieties were key to understanding the ‘why’ - whilst some wanted to be the first to donate, some wanted to be the hero that reached the fundraising target, and others didn’t want to be perceived as ‘that person who donated less than everyone else’. Whilst some just love cats, some want to raise money towards a cancer charity in memory of a loved one.
This data provided the ammo to build algorithms that could predict what would happen and later, more excitingly, start to influence what was going to happen. In fact, tests showed that their algorithms were 90% more effective than traditional CRM methods in getting their users to do something.
But this success brought issues around internal processes and collaboration. When data conflicted with ideas from the product teams or stepped on the toes of traditional departments, their challenge was to negotiate towards a single collaborative process and solution. In the end, they used 14 years’ worth of online ‘giving’ behaviour (engagement, personal and social data), web traffic, machine learning and graph theory to build the ‘Graph’; a big model of constantly moving causes and nodes that gives a highly granular, constantly evolving view on how people are connected and why. This model powers the 'Give Graph', which knows:
1) How and when you want to give - do you want to give time or money?
2) What you care about - do you care about cancer or one-legged cats?
3) How to keep you engaged - is this the right time for you to give?
This has helped JustGiving get closer to their goal, growing into an organisation with 23 million users across 165 countries that has raised over $4 billion towards charitable causes.
Big data can work for organisations of all sizes. It seems the big question that Mike is trying to answer is, how can we codify ‘getting value from big data’? How can we create a formal end-to-end process that teams around the world can follow to get more out of data?
You can watch Mike’s talk below for full details on how JustGiving managed to harness the power of big data and apply it to begin to grow the world of generosity: