Once upon a time, a long time ago, I was a technology strategy consultant. What that really boiled down to was, I translated IT to business executives, and translated business to IT people. Way back then, it was pretty easy to justify infrastructure projects. The IT department’s budget included provisions for things like telecom, network, data center, PC’s, etc. etc.

Even at that time (like, 15-20 years ago), control over IT spending was shifting from IT’s direct control to the business, but what happened was this: for every new project, the technology impacts and costs were assessed, and IT worked to provision against those anticipated impacts, whether through rejigging existing infrastructure, or through the purchase of new hardware or bandwidth, etc. etc.

That worked really well in an environment where technology was fairly stable. Sure, Moore’s law and all that, but the PC hadn’t changed in years, and that was the most change-able technology under your typical retailer’s roof. But then the internet came along. And smart phones. And somewhere along the way, the business side of the house completely wrested control over IT spending away from most IT departments.

Now, on the one hand, some of this was deserved. IT departments would mandate things like “we will only use DB2 “, and reject business leaders’ needs out of hand because the solution they wanted was only available on Oracle (a thing I witnessed at one company). Or (another thing I witnessed, though I’m pretty sure the company has reneged on this policy by now), one company completely blocked corporate access to the internet, out of fear that someone would send an email or access an internet site that was inappropriate.

When you’re just a marketer or merchandiser (or pick your favorite line of business organization within retail), just trying to get a job done, at some point you’re going to throw your hands up and decide IT just isn’t worth it. When no explanation or justification is offered for No and there is little or no help in getting to some kind of yes or compromise, then when IT wants to tack 30% or more of your project cost onto your project for infrastructure, at some point you’re just going to find another way to get it done.

To be fair, this is a gross generalization. Not all companies have operated like this. But I’ve certainly seen enough bad behavior — in both the business side and the IT side — to know that my own limited experience is probably an understatement of the extent of the gap between the organizations.

In the meantime, once business leaders got control of the IT budget, they immediately cut off that infrastructure allocation, and started paying only for what they perceived as their own needs. The problem, as the last 5+ years of marketing investment has proved, is that while parts of IT may be specific to specific uses, there is a lot more of it that is a shared resource than not. Like, for example, the network. And there may be projects that need specific hardware that on their own cannot come up with a business case big enough to justify a large hardware investment.

But if that hardware could be considered as infrastructure for multiple projects that could be built on that hardware platform… well, now you’re talking about something that doesn’t just enable one business case, in enables several. And taken together, those projects deliver more than enough value to pay for the hardware needed. A case in point: mobile devices for in-store employees.

IoT is another case where infrastructure costs may get in the way of implementation, without looking at those infrastructure costs in the context of multiple current and many more future projects. If multiple groups within a company each independently decide to take on an IoT project, they could run into a lot of infrastructure issues if those projects are not coordinated — network bandwidth issues, if the solution is not architected properly. Or incompatible hardware or software decisions, or decisions that then require a steep integration cost when they (inevitably) later need to be integrated.

One solution to this problem is to make sure you have an IT department that is an integrated part of the business, and not the step-child. That requires an IT department that is adequately funded (and I don’t think 1-2% of revenue is a realistic ceiling in our technology-driven age), and it requires collaborative relationships so that “No” only means “Let’s find a different way to yes “, and business leaders understand the long-term impact of immediate expedient decisions about the technology that supports the solutions they need right now.

But even when you have that kind of relationship, infrastructure investments can still be hard to justify and can result in chronically underfunding infrastructure to support the current and future needs of the business. However, there is a way to manage that: portfolio management of technology-driven projects.

Rather than adding an infrastructure surcharge to projects, the technology infrastructure spend is pulled completely from the project and established as its own project. If one project, for example, needs digital signs in stores, and the retailer doesn’t have that already, the digital signage part of the project becomes its own stand-alone project. Both projects are either held, or enter a limited pilot phase, part of the goal of which is to make sure the infrastructure requirements are well-understood — things like, how much bandwidth to stores do we really need to keep a digital signage network running?

Other groups can see that there is this infrastructure project out there in a holding pattern, and can evaluate whether any of their existing needs would be served by that infrastructure. They attach their projects to that infrastructure project. As each new project is attached, the bar for the business case for that infrastructure investment is lowered, in part through accruing benefits across multiple projects, and in part because the more that can be powered by that investment, the more strategic that investment is to the whole company.

I worked with one retailer to develop this kind of investment management of IT projects. I never got a chance to take it to another client — the stars just did not align. But as I see retailers grappling — again — with things like IoT and in-store technology, I find myself thinking that this kind of portfolio management of investments, recognizing dependencies between projects, should once again be considered.

If you look at infrastructure on its own, you’ll never justify it on its own. In the plumbing analogy, it’s not the pipes that are valuable, it’s what the pipes carry. But if you never invest in pipes, you’ll never have running water. So it goes for retail technology.