Managing the Data Interface
One of the main challenges in the data industry, in my experience, is generating an appropriate degree of communication among key stakeholders — in particular, managing the translation from raw numbers to end decisions.
In fact, I’d go so far as to say that the failure to communicate timescales, workloads, requirements and other professionally necessary concepts between ‘business’ and ‘data’ teams is probably the number one issue facing the data industry today.
As businesses rapidly scale their data footprint upwards and outwards, challenges around effective strategy and talent availability have been commonplace. As we’ve made progress in understanding an effectively ‘new’ industry, data engineers have become the fashionable must-have resource; with businesses seeking individuals capable of managing the broad range of sources and architecting the ‘technical’ side of information.
Now — hiring a data engineer as the “first hire” is a great decision, and absolutely the first decision businesses need to make to build an effective data function, but even after that piece of the jigsaw is in place, it’s still remarkably common for organisations to struggle to complete the end-to-end with the outputs provided. We still seem unable to identify how to serve the information to end users in a way that both provides effective information and informs on the appropriate solution.
This gap is the ‘interface’. The rarely used line of communication between business professionals and data professionals.
It’s very, very easy to dismiss this step as a foregone conclusion. ‘Data exists. Analyst does thing. Output complete.’ However, it’s just not that simple — and I’ve seen a near endless cycle of opinions on how to make it simpler… for some, data functions should be responsible for providing the business with the information it needs. For others, business functions should be able to self-serve and collate their own information, but data teams are struggling to meet those demands. Ultimately, it all comes back to the same thing:
Managing the interface between the technical and the operational.
Data analysts, broadly, are not subject matter experts when it comes to the content that they’re working with. Sure, they may have a passing understanding of the data they’re reporting on, but it will mostly be a by-product (ie, they learned about it as a result of building reports on it). On the other hand, businesses typically rely on a range of individuals to understand the wider business context. Process owners, business analysts, operations managers — the group of individuals who effectively ‘know’ the business.
As businesses become thirstier for rich and detailed information, data teams become more and more competent at providing these outputs — but it’s in the interface between the two that we’ll find the best measure of a data-driven culture. The ability to communicate effectively between the various groups.
So, as data professionals and business professionals, the next time you begin a project, take the time to ensure that you have effective lines of communication. Take the time to make sure that you’re articulating clearly all those simple things that will ensure that clear expectations are set between groups.
It doesn’t matter how you do it. It doesn’t matter how difficult it is. It’s an absolute necessity for businesses to ensure that both sides are in regular, effective communication. Because ultimately, the management of this interface will determine the success, or lack thereof, of any data function.
