DataKind UK helps charities to use data to transform their impact, and most of their programmes are provided for free. In this blog, Executive Director Giselle Cory shares their tips for organisations wanting to become more data-driven.
If you’re not sure how to make effective, responsible use of data, take heart, you are not alone. Many of the organisations that come to DataKind UK aren’t sure where to start. They are faced with a shortage of time, dedicated roles, and funding for data work, and feel that they are miles behind other organisations. Based on DataKind UK’s years of experience helping organisations get started with data, here are my top tips.
Forget about data
Firstly, don’t focus on your data! Instead, think about your organisation’s challenges.
You’re unlikely to find high-impact uses of your data just by looking at it. Instead, make a list of your organisation’s challenges and needs. Think about what would make your job easier or your organisation more effective. Once you’ve articulated those challenges, you can begin to identify which of them would benefit from data-driven techniques.
Start small, start useful
Don’t worry about getting it all right all at once!
Instead, start small, with one key insight that is useful, relevant, and will spark that lightbulb moment for your colleagues – particularly those whose buy-in you need.
Look back at the list of challenges you made. They might be: getting a better understanding of the people your organisation works with; evaluating the impact of your programmes; or improving operational efficiency. Where could using data add the most value? And, crucially, which information will be most useful to decision-makers?
Whatever it is, work towards getting to one answer and then flaunt it! Showing your colleagues that something useful can come from data is the best way to get traction for your work. You can worry about the bigger systems (the “plumbing”) and your wider data strategy later on once you’ve got some interest and buy-in.
Encouraging staff to grow their skills in the tools they already know and use can be a powerful step.
Too often, organisations interpret being data-driven as a procurement process. As you do more with data, your organisation might need new tools and eventually maybe even new recruits – but you are unlikely to need them right now. When starting out, you can get a lot done with the ‘basic’ tools that you are probably already using – like Excel and Google Sheets.
For those who do want to go beyond tools like Excel, learning a ‘language’ is the best next step. The good news is
that the most common languages in the data science community, Python and R, are open source (free!) and have
great online support communities. Staff can learn for no or little cost, without the need for expensive
platforms or apps. Explore these options before buying licenses for proprietary platforms. Procurement might
be what’s needed at some point – but it’s unlikely to be what’s needed first.
As well as technical upskilling, using data is likely to involve a shift in mindset too. The most effective uses come about when people are encouraged to really interrogate the data and test hunches rather than the more usual reporting, or worse, finding evidence to reinforce pre-existing beliefs.
Build in responsible data use
Ethical considerations must underlie any and all data collection and use.
Many organisations will avoid having conversations about ethics because it seems a bit thorny and complex. In reality, common sense and talking to your colleagues and beneficiaries about what is responsible and appropriate will get you most of the way. Ask questions like:
- Would our beneficiaries be happy with us using their data in this way? (Ask them!)
- What would happen if our computer-aided decision-making was wrong?
- Could our analysis be using one piece of information (e.g. location) as a proxy for something else (e.g. ethnicity) without us realising?
There are lots of free tools out there to help you structure these discussions. We recommend Sense About Science’s “Data Science: A guide for society”; Dot Everyone’s Consequence Scanning and the ODI’s Data Ethics Canvas. Note that the law, especially GDPR, represents the baseline. Being compliant is the first step in responsible data use, not the last one.
If you’re looking to become more data-driven, you’re not alone! There’s lots of support available. Here are some highlights:
- DataKind UK offers free data and data science support sessions in partnership with SCVO. You can sign up for ‘Office Hours’ here
- If you are a data practitioner working in a social sector organisation, you are invited to join the Social Data Society, a peer network that comes together regularly to share and learn
- If your organisation would like more intensive support, we offer longer-term programmes, like our (free) DataDives, to help you explore your data and solve some of your organisation’s key challenges
- For other types of support, browse our recommendations here