A note before we begin: I’m arguing that technology ROI discussions are broken, not that ROI as a decision-making tool is broken. A solid understanding of how to calculate and use ROI is an essential skill for any tech executive, and when done right, it’s a powerful decision-making tool. This post is about how technology discussions that exclusively look at ROI often result in a one-eyed analysis that lacks depth.
Technical leaders need a wider range of tools for communicating the value of technology, and especially technology innovation. Communicating the value of technology is not a trivial task—and the point of this post is that exclusive reliance on the most commonly used tool for communicating value—Return on Investment (ROI)—will lead to broken discussions.
One solution to this is to start discussing technology innovation in terms of how it will break the existing ROI model and whether that thesis is believable.
It’s important to say straight off the bat that ROI itself is in no way broken! It should be part of every investment discussion. But ROI does fail to capture some very important elements of evaluating technology investments.
Two of the primary assumptions that feed the raw data that goes into ROI calculations —namely the operational model of the business and strategic tradeoffs that underpin that model—are sometimes the very things that will change with a successful technology innovation.
I believe the approach I’ll discuss actually requires a much higher standard of excellence from technologists. This is a good thing: as the wave of internet technologies are hitting full maturity, its necessary to become increasingly discerning—skeptical, even—of overly simplistic, ROI-based business-case arguments for certain types of technology investments.
What’s Wrong with ROI?
Almost all budget processes require new investments from technology to be framed in terms of the return those investments will produce.
How to portray ROI is very nuanced1Typically, you use Internal Rate of Return, factoring in net present value of all cash flows and whatever discount rate you apply to those cash flows. Sensitivity analysis is also helpful for generating discussions about how robust the ROI thesis is and what value drivers actually matter., but a simplified explanation of how it works will suffice for our purposes. Say you’re trying to decide whether to migrate infrastructure to the cloud. You’d tabulate yearly costs on current infrastructure (summed over a 2-5 year period, depending on how finance wants it), and include an estimate of soft-costs to maintain and complete activities on that infrastructure. Then you calculate the costs to migrate, and the migration-complete costs. If the business will break even and begin saving money within an acceptable window of time, say three years, then the business case is viable.
It’ll end up looking something like this:
If technology was all about operational efficiency, this would be great. But the problem with ROI is that sometimes technology can also deliver value that can’t be described in terms of your company’s current ROI models.
For example, how do you know if a new technology that enables a completely new way of engaging with customers will generate more revenue? You don’t (if you’re honest), and even if you do, how do you quantify that in a financial model?
“Does it Change the Math?”
One of my former bosses used to ask this question whenever we’d propose a new technology: “yes, but does it change the math?”
What he meant was, is this something that changes how we can operate in some fundamental way, such that the equations and models are actually different now? If the answer is no, then the discourse on the change should be based on efficiency, so we’d go ahead and talk about ROI.
Things that “change the math” are almost inherently problematic for ROI discussions: if you’re arguing the model itself is going to change, representing the change in one, consistent model is very difficult. You may have new variables that didn’t even exist in the old model, and those new variables (if you’re being honest) have no historical precedent.
But the part that you shouldn’t miss is that when my boss asked the question this way, it allowed for a much richer (and more interesting) discussion about the value of a given technology.
How to Properly Discuss Technological Value
So, if ROI is sometimes insufficient, what can work? Well, I like to use this graph to frame the discussion about value:
Usually, viable business models stabilize on a balance between two strategic objectives that are at inherent tension with each other. For example, Lower Costs vs Higher Quality. If you skimp too much on cost, quality goes down. If you pursue quality at all costs, your costs will rise.
Your business has not only picked a “sweet spot” on that curve in terms of profitability, but has also defined a curve that gives it the ability to adjust to changes in the environment:
If a new competitor enters the market, and they’re offering crazy quality, you can respond by “moving up the curve” and opening up a premium-service line that costs more. If a low-cost competitor enters the market and surprises you by gaining a lot of traction, you can “move down the curve” and cut costs at the expense of quality.
But what you can’t do is break that curve. Or can you?
That’s where technology is so important. Technology that’s truly innovative helps businesses break that curve:
There’s no great way to properly represent this in a traditional ROI (that doesn’t stop everyone from trying though!)
I think what we should do is actually produce a document that attempts to convince leadership that the model and tradeoffs your business has been operating under can be broken with the given technology. This forces technology leadership to explicitly answer some rather hard questions:
- In what ways do you think the model is broken?
- How, exactly, will the new technology change that?
- What does the new model look like, and how is it better?
These questions are the important ones to get everyone aligned on. These questions should be fought over and scrutinized from every angle. Because these are the questions that will actually predict the value of whatever ROI calculations you end up doing.
Wrap Up
I love this approach, because it forces technology leaders to explicitly define the innovation in opposition of the current business model. It jolts us out of our formulaic, boring by-the-book OI calculations that often aren’t accurate anyway.
Another benefit of this approach is that it often exposes the problem with the potential innovation investments right away: when we draw this graph and pick the axes, the other business leaders say: “well, I don’t actually care about the two axes you defined, because while they exist, they aren’t fundamental to the business vis a vis the investment you’re proposing.” Now you’ve arrived at truth, fairly quickly!
Let’s go back to the discussion about infrastructure migration. The ROI case we sketched out is very typical. There’s probably even templates floating around that CIOs and CTOs could use to make the case.
But I think if we’re being honest, the argument is missing something, isn’t it? If the ROI discussion was really so straightforward, when Netflix moved to AWS, everyone would have immediately piled on. But there were skeptics for years. What were they skeptical about?
Well, the reality is that cloud infrastructure isn’t a little better because it makes it a bit cheaper to run servers. It’s way better, because it largely removes infrastructure as the bottleneck for innovation. Your graph, then, looks like this:
How do you represent that as ROI? And yet, it’s the fundamental factor in cloud-enabled technology.