Data Strategy #1: Introducing the Loft Data Canvas

In this first part of our new series exploring how to develop a data strategy and roadmap for your product, we introduce a process for brainstorming the data in your business and how you can create value from it.

We know that it can be really hard for business leaders to understand their data – what it is, how it’s organised, what the opportunities are, and so on. Too often data is stuck in the realm of techies, whereas we believe it should be something everyone in a business can engage with.

To enable this, we’ve been looking at ways to demystify data, and we’ve come up with Loft Data Canvas – a tool anyone can use to brainstorm and start to map out their data strategy.

Download the Loft Data Canvas

It is of course inspired by the famous Lean Startup Canvas created by Steve Blank, which has led to many other sorts of ‘canvases’ helping businesses to think about different aspects of their strategy.

Each section of the Loft Data Canvas looks at a key pillar of your data process, working backwards from the potential value/revenue streams and business model, to the data sources, and the plumbing you need to put in place in between them.

You can do this on your own, or collaboratively with your team – or even do it separately and then share findings to see where the overlaps are. You can fill it in online, or print it out on an A3 page and stick Post-Its on it.

Here’s how to work through the different sections:

Data outputs

Starting on the far right, brainstorm first the different types of data your business could provide to others that would be of value to them. Think about customers, partners, investors, internal stakeholders, policy makers and other groups – what valuable data and insights can you offer them?

Think also about how they would like to receive and consume it. For example, consider dashboards and other user interfaces and experiences, APIs and other integrations, as well as reports.

Consider – what is the most valuable data that you can build? And what industry standards and other efficiencies can you apply to it to maximise its value?

Inputs

Now on the far left, brainstorm the types and sources of data your business can gather, and where does this data come from?

Think about customers, partners, public data sets, integrations with other tools and systems, operational data, analytics. Think about the data people enter into you web interfaces, the data you can create by analysing it, and also data you can get from your partners and the public domain.

The more you brainstorm, the more ideas you will come up with.

And then, building the processes for your business between these, from left to right:

Collecting

How and where will you get the data from your inputs? For example, user data might come through web and mobile apps, surveys and other forms; operational data from internal systems; while other data may come through APIs and public sources.

Extract, Transform and Load is the data science term (or, in cloud systems increasingly Extract, Load, Transform) for getting large amounts of data from one system, preparing it then loading it into storage.

What connections and transformations do you need to apply to your data at this point to make it more useful? From this you can start to work out what tools and systems will be needed.

Storage

Where is your data kept and in what formats? You may have one or two main databases or data warehouses, but there are probably many other locations your business keeps data – from MS365 or Google Workspace to various CRM and other systems.

It’s really helpful to brainstorm all of these here so you can see the whole picture of where your data is. Many small businesses have 20-30 different systems, so be prepared for this list to get quite long. Mark on it where you have personal data, IP and other confidential or particularly sensitive/critical data.

You’ll probably start to uncover new potential data sources to go in the lefthand column as you work through this too.

In this column, think about what analysis will you need to do on your data to create the valuable outputs and insights that your customers and other data users want.

Analysis

How will you need to clean, link and process data? How much of this can you do with relatively simple existing technology, and what will need research and development?

Can it be accomplished with simple business logic, or is there a role for AI? Do you know if the algorithms you need already exist, or will you need to potentially create them? And what other problems will need to be solved?

Be prepared to have lots of unknowns. The important thing is to get the unknowns and assumptions down, so you can start to prioritise them.

Governance

Across the whole pipeline, you should also think about governance factors. For example:

Consents, agreements, compliance and security – what consent do you need, what GDPR and other regulatory compliance do you need for your project, and what security standards need to be in place for you to operate legally and safely. You might have lots of unknowns here – write them down so you can work with your professional advisors to answer these questions and understand what’s involved.

And of course KPIs – how will you measure success, and monitor how well you are moving towards your goal?

Finally…

Please do share any feedback you have on the Loft Data Canvas. How well does it work for you? How could it be improved?

And of course, if you would like us to take you through it in a workshop, or help you to plan the next steps after your initial work, we’re here to help. Please do reach out any time.