Where Are We Now? First Steps to a Solid Data Strategy

Read this article
⸻ Beyond Data

Getting a broad understanding of how your organisation uses data

Prompt: Compass with many points on the needle, digital art (made with Midjourney)

About a month ago I published a quick overview of my favourite approach to delivering a data strategy. This four question approach I’ve used time and time again with great success. If you missed the post, you can read it here.

Following that post, I’ve had lots of people reach out and ask for more details. This, therefore, is a first attempt in what I guess will become a bit of a data strategy series.

When starting out with a data strategy, most organisations get entirely lost. The word strategy is difficult enough for some – business speak that gets banded about until it’s meaningless. Why confuse things with a data strategy too?

To me, data strategy is about lining up our platforms, systems, and initiatives to maximise the positive and minimise the negative impact data can have on our overarching business goals.

To do that, we need to understand what we’re working with. So in this post, I’m going to talk about question one:

Where are we now?

This part of the framework is all about understanding the current state of play within an organisation: what’s good, what’s bad, and what we are missing.

A word on language

All this business speak can get confusing, so I want just to introduce two terms I use to help shape our thinking:

  • Concerns or dimensions – these are orthogonal, often competing considerations that the data strategy addresses; for example, security and ease of development are often competing concerns.
  • Enablers or elements – these are the things that you will use to achieve your goals. The most common are people, processes, and technology. I have seen some cases where patterns, platforms, and partners (lots of p’s for some reason) also make sense. I won’t talk much more on these here – assume we’re just using people, processes, and technology.

I’ll use these throughout the rest of my discussion.

What are your priority concerns?

You’re about to embark on a journey of discovery, but it helps to know what it is that matters most to you. There’s plenty of advice out there about what you need to cover, and your situation will have specific needs that dictate what’s most important.

For most organisations, the following makes a strong set: security, governance, discoverability, ease of development, ease of use, availability, risk, and cost. You could include many more, but don’t go overboard – more than nine starts to get unwieldy and complicates the whole process. You’re actually better off taking smaller subsets and repeating the whole thing multiple times.

Whatever you do, I’d also recommend picking three that are your absolute primary concerns and getting agreement from the organisation’s leadership. This will make decision-making later much more straightforward.

Start broad

Reading “The McKinsey Way” recently the author talks about how doctors that self-diagnose often make terrible patients and how that’s doubly true for businesses. In this context you’re often met with three types of users:

  • the technologists and data professionals that “know” what the data strategy should be
  • the interested parties that changes might be beneficial but don’t know what
  • the folks that don’t know or care and just want to get on with their job

Now the first group will usually want to take up a lot of your time and lead you through their vision. That’s really good – let them. But be wary too, you can’t be completely sure they’ve dug into the detail enough without actually checking.

You’ll never understand someone’s role and context better than they do so ensure you’re speaking to them. Make sure you engage with all three parties as best you can at the start; you can shrink the group down later.

Cast the net

How do you engage the widest group possible?


Unless you’re in a small organisation or have a lot of time, getting guidance and feedback from people is challenging. That’s why I love surveys. It’s easy to share and asynchronous, meaning it’s less of a burden on the whole process.

I’ll not go into detail here about crafting a good survey (maybe a future post). I will say you want to cover all of your elements 2-3 times in each dimension. Ask questions on a scale (1-5, Bad-Good etc.) and give people the options to answer both “I don’t know” and “This isn’t relevant to me”. You want to get people’s thoughts on the current and potential future states; strengths, weaknesses, risks, inefficiencies etc.

If you have many dimensions, you can break the surveys up. You can also allow anonymity and/or collect some metadata about the respondents.

Finally, if your organisation is larger than 150-200 people it might make sense to do this on a functional or regional basis. It can get difficult and blur the meaning when you’re looking at groups larger than 150.

Collate and group the results

The results from the survey will give you a good idea of how people across the business think the people, processes, and technology either support or hinder their efforts. What things they like and what needs improved. Where there’s potential opportunity missed and money wasted.

You now want to look for patterns within the data.

Take these results and map them against your elements and dimensions. Group them by teams, functions, regions, seniority levels. Try to find regions in the responses where there’s consensus but also look for where two groupings give different opinions.

The objective at this point isn’t statistical significance, we’re still talking small numbers right? This is all people stuff – the individuals that work in the organisation have graced you with their view of things. You want to use these responses to shape the next steps.

One of the best insights we came across was a huge disconnect between management and the people in teams. Almost everyone above a certain seniority throughout this organisation had a positive opinion about the items surveyed and those below that level, quite the opposite. It was like an invisible wall that hid the reality of working with the data in that organisation. This led to some incredibly productive conversations in the steps that followed.

Nothing is set in stone at this point, it’s all malleable. We’re still in the learning phase.

What next

There’s still more to do. We’ll cover this next week. Now that you have a broad view the next step is to uncover which parts are most interesting and what are the highest priority areas to work on. We’ll then look at how to deep-dive and bring this all together.

Thanks for reading and please let me know if you have any thoughts. Also, I’d greatly appreciate any support in sharing and growing this newsletter (and you can earn free gifts from me if you use the referral link at the bottom).

If you enjoyed this, consider sharing this newsletter with others that would find value from it or follow what we’re doing at Hypercube


When you’re ready, there are a few ways I can help you or your organisation:

What do you think of this issue?

I’d love your feedback to ensure I’m writing about the topics you want to read, let me know if this one hits the mark

Want to write for Beyond Data?

Have something you want to share? Interested in writing an issue or working with our team?