Data Interoperability – The missing ingredient in exceptional enterprise design and delivery?

Is our drive for Artificial Intelligence (AI) skipping the basics, and trying to put a sophisticated heating system into a house, before the floor, walls and roof are constructed?

Data interoperability is probably as talked about as AI, but arguably, it is even more important. AI creates value and truly comes into its own in organisations where we already understand how they operate, where the pain points are, how we want them to change, and how data flows between the different services. But when we can’t answer those questions, AI, while a worthy ambition, is not an easy win and it won’t create the full benefits that are anticipated.

To be clear, this blog is not advocating one activity in exclusion of the other; moreover, it is trying to convey that there are stepping stones of stability that should (or must?) be laid down, before taking the bold leap into the world of AI and its many flavours.

Do we want data interoperability?

Whilst AI is being (rightly) scrutinised in terms of its responsible application, governance, source reliability, and recommended use cases, most people would agree that it is wise to share data with other Government partners who need it. Furthermore, I haven’t encountered any resistance to the assertion that organisations should make best use of their data – people generally accept that it is good sense to avoid collecting and storing information multiple times, which in addition to being inefficient, also inconveniences the citizen with repeated and unnecessary questions.

Similarly, I believe that there is also consensus that data should be collected directly from source, so that there is an increased likelihood of accuracy and veracity. The sentiment, passion and understanding seems to exist already, but what is lacking, is business environments that enable this, e.g., common language, consistent data standards etc.

So clearly, we all agree – but now what?

Wanting data interoperability and being able to successfully line up the business in a way that enables it to happen, are two very different things. Success can be constrained or facilitated by several factors that are explored below.

1. Bravery 

For me, the first step in simplifying the data landscape and exploring opportunities for consistency. This is less about tackling technical and/or policy complexity, and more about a willingness and determination to be brave and take the first steps. By this, I mean making a clear decision [and a public declaration] that data standardisation and join-up are a business priority.

2. Understanding

To simplify and standardise the data landscape from collection, through to consistent reporting, there needs to be a shared understanding of what will be common, and then a determination to re-shape the business to make that happen. For me, data interoperability can’t happen if business stakeholders and operational areas don’t identify and align to a common implementation of data. The same can be said of common technology and re-usable components. If they are not designed with the business, and if they are not going to be deployed for use in the same way, then the value of that consistency is lost, or at the very least, delayed and heavily diluted.

3. Consensus

I think this term can be confusing and is often unintentionally mis-used. Arguably, consensus doesn’t have to mean that everyone agrees exactly with a proposal. In my experience, it can simply be a group of stakeholders coming together, who have a chance to explore an idea with each other, be heard, and examine the pros and cons. The important aspect is everyone having a chance to weigh in and shape next steps. With multiple people in a room, it’s highly unlikely that every single one of them will agree one hundred per cent with everything, but responsible leaders understand that progress can’t be made until they move forward as a unified team with shared goals. I believe that this is consensus and if we follow an inclusive, engaging approach from the outset, this is achievable.

4. Commitment

Being brave enough to take the first step, building understanding across the business, and achieving consensus about the way forward, are not enough if there is not the organisational commitment to make it happen. Senior leaders across business boundaries need to communicate their intent and plan to deliver against it, with clear roadmaps and designs that align to enterprise outcomes around data interoperability.

Potential approach to your data transformation journey

I propose some logical stepping stones of stability (a phrase our team likes to use!) that will help you reach your destination:

• Understand what your organisation is trying to achieve: in Government, this links to your policy intent and business strategy. You can’t reach your destination if you don’t know where that is, or if everyone thinks they are going to different places!
• Co-create meaningful and measurable outcomes that break data transformation down into chunks, highlighting what good looks like for each of them. Avoid ambiguity but establish these goals together with the people who will lead the charge across teams.
• Taking the strategic direction described by strategy teams, explore what needs to change in the business environment to deliver these goals, objectives and outcomes. This is broader than the data: what are the wider impacts to people, processes, organisation, policy and technology? These elements need to be considered in parallel.
• Understand what data is required to support your business processes and customer journeys. Technology should align to these, with data flows mapped between systems and components across the enterprise. It’s important to ask and answer critical questions like: where is the best place to get this data? Do I need to retain a copy of it? Do I need all the data, or just a single item/attribute? Will anyone else in the business have a justified need to use this data after me? How will business data be governed and audited?

Success requires enterprise leadership and business areas need to go ‘all in’ in terms of consistent design, but also in their commitment to a single implementation of the solutions. Divergent delivery approaches create walls and reinforce silos, at a time where the ambition is to move seamlessly across boundaries.

To advance service delivery and exploit the emerging opportunities around AI, I propose that we need to build the basics first, and that is as much about establishing baselines and lining the people up behind the vision and shared goals, as it is about understanding what could be common in an organisation.

Please feel free to reach out to me if you'd like to discuss this topic further.