Never heard of a Data Sprint? It’s our hybrid approach, inspired by all the best bits of Design Sprints, to deliver the value of Data Science and AI at lightning speed. Here's our Head of Data & Analytics, Ryan Moore, to explain how Data Sprints work.

What is a Data Sprint?

A Data Sprint takes a design sprint approach and merges it with what we do really well in Aiimi – data hackathons – and combines those two to create a methodology which is based on discovering, developing, and delivering a data solution.

So, we have a 2-day Data Dash. At the end of this, you will get a feasibility score. You get enough information to make a go or no-go decision whether you take this forward or not. If you do decide to go ahead with it, you go into our 10-day Data Sprint. What happens in those 10 days is that we focus on building a prototype and we also focus on how you're going to take that prototype into production.

The key benefit of a Data Sprint is that you've got a focused group of professionals who are working on a single problem, and you've got all the skills in one room that you need to solve that problem – without having to deal with the bureaucracy that comes with larger-scale projects or programmes.

Virtual Data Sprints

Virtual Data Sprints are very much the norm these days. They work really well for us, because all of these exercises are a combination of online and on-camera, with people talking and interacting with each other. I don't think there are any elements which are lost by us virtualising the Sprint.

Traditionally, Design Sprints have always been associated with digital products – that's something which has a UI or a UX. What we've done is taken the context of that and applied it to data, because the outputs from a Data Sprint are very similar – you will have an interface where people experience data.

A focus on data experience is one of the factors that differentiates or sets us apart from other offerings.

Who should participate in a Data Sprint? How can you work with other teams?

From the customer side, it's always good to have end users who are facing that problem.

The second persona is a business SME (Subject Matter Expert) who is familiar with the processes that relate to that problem.

Then, a data SME who knows the data landscape and can point you in the right direction and say "here's where you go and find that data", or break down some of those blockers we have in accessing data.

As always, we have a discovery aspect to it. In our Data Dash, what we do is produce that feasibility score.

As an information management specialist, we've been involved with large-scale digital and data transformation programmes. We've had a lot of success embedding smaller, focused teams doing data science, AI, and digital development within these organisations. We looked at what was happening really well with those teams – they were able to produce results very quickly, they were able to deliver value right into the heart of the business.

We work very closely with the UI and UX team in terms of facilitating these workshops. They've got the experience of being able to make this exercise user-centric, and there's this kind of stigma around data that it's very complicated.

Another way in which we work with other teams is in the Data Discovery stage, right at the beginning of our exercise. And we've got our own platform, InsightMaker, which really speeds up that process of finding the data you need to be able to solve this problem.

What we produce from data science models and analytics models, and basically the output of a full Data Sprint, is that we're able to put a data product in front of the customer so they can say "Yes, that solves my business problem".

You might also want to check out:

Our Design Sprints & Data Sprints

Data Sprints - why run a data project when you can sprint

What we learned from 2 virtual hacks, 3 weeks, and 150 people

How to make your Remote Design Sprint a success