The old debate
The build-vs-buy conundrum is a time-honored business hot topic. As the pendulum has come and gone, the evidence can be made to look strong for both the do and the don’t sides of the argument. The divide is particularly visible in IT. Managers who think of IT as a strategic resource lean more heavily towards employing their own people. But then, are all parts of IT equally strategic? As is often the case, this is a decision that requires careful deliberation. For IT managers it is itself a decision of strategic importance.
The apparent economic advantages of farming out work are straightforward. Managed Services companies’ business models are predicated on scale; given a big enough chunk of work, they are expected to easily beat the price. Things are less clear when the work to be performed is highly specialized or tightly connected to a firm’s decision-making process. Firms will pay a premium to have these functions performed by people with a deeper understanding of their business model and a stronger commitment to the company’s success.
Of a kind
Of all types of information systems, those whose goal is to support decision-making (data applications) form a unique category. They carry the input that managers need to operate at the most basic level; they are the key to describing the outcome of past decisions while informing future ones. It is then no surprise that data applications are more frequently owned and operated by internal teams. But knowing that a lower price is one buying process away, is that premium worth paying?
As upper management’s main navigation tool, data applications require from their owners the ability to be prepared to respond to a much more unusual, if not unpredictable, demand pattern. To make the situation more dramatic, what used to be simple reporting systems—a visual representation of data sitting on some core system—has evolved into a complex, multi-tiered architecture, where each component is critical for the well-functioning of the entire data chain.
And yet, the size and complexity of modern data architectures open up an entirely new set of opportunities for IT managers. Where once all-around, multi-disciplinarians were an absolute requirement for work in data systems, today’s data architectures demand deeper and specialized knowledge of constantly emerging new sub-disciplines. It has become unrealistic, if not unreasonable, to post requirements in a single job description that are as far apart as data ops and observability, data governance, SQL development, and data visualization.
The modern enterprise data application owner must now play a role that’s similar to that of a traditional software product manager. While holding the deepest understanding of their users’ needs and expectations, they ensure plans and resources align well towards delivery that is consistent with organizational goals. It’s the resource’s knot that Managed Services can make a promise to untie for data application owners. Managed Services firms can pool and activate resources much more efficiently than any average-sized company. That allows them to move resources around more flexibly to meet irregular demand spikes, a maneuver that is much harder and expensive to execute when you need to hire and maintain staff for an entire data architecture.
As user expectations and system complexity escalate, IT leaders and data application owners seeing their projects impacted by a lack of resources should respond by opening up the parts of their apps that require the most specialized technical knowledge to external teams capable of absorbing the load while maintaining high delivery standards. The shift will unlock value that can be brought back to management and serve as economic validation for more and better data applications.