Build a training approach to suit your business needs, your calendar and your skill gaps

Filter By

Show All

Connect to


5 Tips for Starting a Data for Good Project


As part of DMA Scotland's regional content, guest contributor Alistair Adam, Head of Analytics, Optima Connect provides guidance and advice on starting a ‘Data for Good’ project.

Data for Good (DfG) is about data scientists giving back to society. But it’s not a one-way value exchange. If you choose the right project and create the right conditions, DfG projects can also be a valuable learning experience for your data science team.

What is the data for good movement?

A not-for-profit organisation called DataKind inspired the data for good movement in 2012. They explain its purpose like this:

'The same algorithms and techniques that companies use to boost profits can be leveraged by mission-driven organizations to improve the world, from battling hunger to advocating for child well-being and more. However, most social change organizations don’t have the budget or staff to take full advantage of this data revolution and most data scientists don't realize just how valuable their skills can be.'

So, DfG projects are about data practitioners donating their time to good causes. We decided to launch our own DfG project just over 12 months ago when it was becoming clear that the pandemic was going to have a real impact on the UK and the charity sector.

It’s important to be transparent here. This wasn’t a solely altruistic decision. We recognised that a DfG project could also give our data science team valuable real-world learning opportunities. Particularly for the more junior members of the team, DfG presents a real win-win opportunity for businesses and charities alike.

With that in mind, we made the decision to commit our data science team’s weekly training and development time, found the ideal partner, and we’ve never looked back. The last 12 months have flown by, all things considered, and we’ve learned a lot along the way.

In this article, we’ve captured five of the big things we’ve learned. Our hope is that by sharing these tips it’ll make it easier for other companies to get involved and kick off their own DfG projects.

If your experience is anything like ours, you won’t regret it.

5 tips for starting a Data for Good project:

  1. Find a cause your whole team is passionate about
  2. Find the right partner
  3. Select a project that will be valuable for your partner and for the skills development of your team
  4. Make sure there’s a project sponsor ‘client-side’
  5. Think long-term

  1. Find a cause your whole team is passionate about

This might sound obvious, but first pick a cause, a movement or a social change initiative that your whole team really cares about.

There is no shortage of good causes out there, especially with the pandemic hitting charities the way it has over the last twelve months. What’s really important is to choose something that excites your team, something that makes them eager to put in the effort, the hours, and that lights a fire to bring about positive results.

For us, that was social care.

  1. Find the right partner

Once you’ve narrowed it down to a particular area, you need to find a partner. With so many great causes out there, at first, this can feel a bit like looking for a needle in a haystack.

Thankfully, we had some help.

We started by talking to the Data Lab in Scotland, who put us in touch with the Coalition of Care and Support Providers in Scotland (CCPS) and the Scottish Council for Voluntary Organisations (SCVO).

The CCPS connected us with Real Life Options, a UK-wide charity that supports people with learning disabilities, autism and age-related needs to lead more independent lives.

From the first time we spoke to Real Life Options, we felt a strong cultural fit and they had a very specific, very costly problem, but they lacked the analytical resources to address it.

A few conversations later, and we were delighted when the team at Real Life Options agreed to partner with us.

  1. Choose a project that will be valuable for your partner and for the skills development of your team

One of the reasons Real Life Options felt like such a good fit for us was because we could see how our project would help address a big financial issue for them and let us apply our skills in analytics and data science to a new area.

Real Life Options is a charity that supports people living with disabilities. The charity has a large network of employees providing hands-on support to people living all around the UK. Like many organisations in the care sector, employee retention is an important topic. And one that’s become even more critical as a result of the pandemic.

Training their staff is a big investment and, where they have staff leaving unexpectedly or after a short tenure, they often must use expensive short-term contract staff and invest more in recruiting permanent replacements.

The goal of our DfG project with Real Life Options was to use analytics and predictive modelling to reduce employee attrition rates. By giving the charity insights around why employees are likely to leave, and a consistent method of identifying at-risk employees, they could then develop preventative strategies and targeted support frameworks.

We liked the project right away. The data modelling principles are the same as we use in the corporate world, only here we’re focusing on employee retention instead of customer retention.

  1. Make sure there’s a project sponsor at your chosen partner

This is especially important if your chosen partner doesn’t have in-house analytics or data science team. Whether it’s access to the right people, the right data, or someone to help champion the work you’re doing and make sure it gets used in the right way, we were extremely lucky to have Paul Cusworth, Director of Digital and Enablement, fully involved from day one.

Without a project sponsor that was as committed to our work, we wouldn’t have made the progress we have.

  1. Be realistic about how quickly you can deliver results

This might not be true for every company, but for a company like Optima, we had to be realistic about how much time we could commit to the project, and what that meant for how quickly we could make progress. We put four data scientists on the project, each allocating their full training and development time to Real Life Options. Knowing this up-front meant we could give the charity a realistic idea of when they would start seeing results.

Even with multiple data scientists working on it, half a day each week still doesn’t allow you to make as much progress as you might like. Things will take longer to complete compared to paid-for client work where more resource is allocated. And again, that feeds into what makes the ideal project – you want to find projects that deliver a lot of value, but equally where the client is happy to work to flexible timescales.

And bear in mind that just because you’re working on a pro bono project it doesn’t mean there’s any less rigour or security required around data handling. DfG projects still require contracts, NDAs, data security protocols and so on, just as you would for any other client. All this setup including getting access to the right data takes time.

Closing Thoughts

We’ve learned a lot in the last twelve months, and we’ve come a long way. We completed Phase 1 just before Christmas, and the question was never - will we do Phase 2? - it was just a question of when we could get back to it.

Real Life Options has been the perfect partner, and we’re grateful to them for their continued support. And it’s rewarding to see our work bearing fruit and informing their strategies and decisions around such a central issue to their financial wellbeing.

Thanks go to Steph Wright at the Data Lab and Nancy Fancott at CCPS for being so helpful in guiding us on the all-important first steps in our Data for Good journey.

- Alistair Adam, Head of Analytics, Optima Connect

Alistair Adam is Optima Connect’s Head of Analytics. He has over 20 years of experience in data science and analytics within major enterprises and analytics consultancies. His specialisms include predictive modelling and customer segmentation.

Find out more about Data for Good by contacting Alistair on

Hear more from the DMA

Please login to comment.