Case Study: Comic Relief - Raising a laugh, raising funds, raising data quality

David Reed, director of research and editor-in-chief, DataIQ

Comic Relief is unusual in the charity world in that it raises the money from which it provides grants to good causes through two bi-annual events - Red Nose Day and Sport Relief - rather than via ongoing fundraising and donation drives. As a consequence, it has until now had no requirement to build a single supporter view in which all of its data sources would be integrated.

As Yemi Okunade, head of data, says: “That is the nature of running a campaign once a year - data was treated like the campaign and managed once a year as well.” But that did not mean the charity had no issues with the data it used in fundraising - in fact, quite the opposite was true. “There was no free-flowing data lifecycle with a limited amount of data available in our 12-week campaign to fuel decisions that really make a difference,” says Okunade. “There were two main reasons for the problems. The first was that the charity had many different suppliers of data which changed year-on-year. The second was that we were trying to get all data sources to conform to a single file layout which complicated matters not just for the supplier, but also for us when it came to sense-checking data supplied.”

Red Nose DayHis was a baptism of fire, joining Comic Relief in January 2015 ahead of last year’s Red Nose Day in March. When reporting on that fundraising effort, management information was limited by a lack of context. With no linking via a single supporter view, it was not possible to identify if fundraisers were old or new, high or low-value because there was no historical, transactional data to add in to the communications data set.

Teams within the charity had also grown to accept the situation. “When we were ouputting data for email or direct mail, the organisation had adapted to the issue of duplicates. Part of the team’s job was to dedupe those output files manually, so I saw a real opportunity to add value - not least because I didn’t want highly-skilled developers and analysts doing that over and again,” recalls Okunade.

Comic Relief has an in-house database development team with a lot of new faces, so much of the knowledge about the systems it uses - and from which it sources data - had been lost. Okunade also found that some of the infrastructure was not fit for purpose, like its home-grown matching software which had 3,000 lines of SQL script. “That was limiting our ability to make changes and adapt things like matching rules,” he says.

As delegates at last year’s DataIQ Link event will know, Comic Relief undertook a sophisticated review of its data needs in the context of marketing communications. The issue under consideration was whether to invest in a full-blown single supporter view (SSV) or if matched data could be created and delivered into a new customer relationship management (CRM) solution. Knowing this decision was impending, Okunade raised the data quality issue to make sure it would not dilute the benefits which would result from the choice ultimately made.

“I thought it would be best to find a solution for the data quality element, mainly as I wanted to separate the management of data from the use of data to an extent, to create an almost plug-and-play architecture for reporting and marketing platforms in the future, so I looked around the market,” he says. This led him towards Trillium Software, not least because of the vendor’s attitude towards working with the charity. “They really wanted to engage with us, not just helping us to improve our match rates, but looking at householding, how we collect data, what our data silos were, understanding the key drivers for internal customers for our data.”

The first stage of the partnership was a 15-day collaborative “rapid data eengagement” on an extract sample of records, followed by a data improvement exercise. The outputs from this provided the charity with some eye-opening insights into where its data quality level really was.

“That gave us some interesting statistics back, like having nearly 10% duplicates and 402 different formats for telephone numbers, 64 different formats for mobile numbers,” says Okunade. “There had been an acceptance that we didn’t need to validate telephone numbers because Comic Relief doesn’t do outbound telemarketing so we had no intention to use them. But it is a useful element to verify and match duplicates. The more data you have got, the better the match rate - but make sure you do have a legitimate reason for collecting the data!” 

With that knowledge of the underlying state of its data, Comic Relief was able to model the value it might derive from investing in a data quality solution. That value comes from two sides: reporting and analytics, and outbound marketing communications. “With reporting and analytics, if you don’t pull all of your interactions and transactions into a single view, you get two different views on one person. You might think they belong in segment A, but when you combine that data, they should really be in segment B. Our outbound marketing could be confusing for supporters, because they might be treated as if they are two different types of supporter,” says Okunade.

Comic Relief is very forward-looking in its use of data, but it still had to work on generating the financial case for making the investment. The key was a marketing communications test compared to a cell that had been cleaned and deduplicated - that showed 25% of people in the test were getting the wrong journey or getting elements of two different journeys. 

“We were able to prove that rolling that out when developing the single supporter view would unlock the potential of a return on investment of between £600,000 and £800,000 per campaign. We also demonstrated productivity and efficiency gains in how we work with data, saving 65 days that had been spent on manual deduplication,” he says.

Releasing human resources is particularly beneficial as these are an expensive overhead. The time made available could be applied to value-adding activities which would accelerate Comic Relief’s progress up the data and analytics maturity curve. Okunade places the charity between stages one (reactive) and two (proactive). He notes: “With the introduction of the data quality solution and the single supporter view, we will be able to use the skills of the team to drive significant improvements.”

With a five-year investment into the cloud-based Trillium Software data quality solution, the pathway to creating that single supporter view has now been cleared. It is a testament to Comic Relief that it sees the power of data, not just in terms of making information more effective in the organisation and getting it right, but also because of the effect it has on the charity’s supporters. As Okunade says: “Great data can help drive great Comic Relief experiences for supporters.”

Knowledge and strategy director, DataIQ
David is developing the framework for soft skills and career development among data and analytics practitioners. He continues to be editor-in-chief and research director for DataIQ.