This article was co-authored with Duke Dyksterhouse, an Associate at Metis Strategy.
So long as time is limited and leaders must decide how best to use that time, capacity management will remain a cornerstone of organizational efficiency. And for digital and technology leaders in particular, it’s more important now than ever. Nearly every department in the organization is competing for IT’s resources, often to enable strategic moves they hope to make before the competition. Delaying the wrong project can break a company – so tech leaders must prioritize and they must prioritize well. Capacity management is integral to doing so.
But anyone who has tried to build a capacity model knows it’s deceptively challenging. Countless variables muddy the equation—PTO, unforeseen projects, workers that span multiple teams, even the occasional pandemic. The architect of the model will, in good faith, try to account for all of these nuances by layering in an ever-growing number of toggles, caveats, footnotes, and filters. But there’s always another variable. And the future submits to no spreadsheet.
The secret to capacity management, then, is not mathematical rigor but emotional discipline. It’s about keeping your eye on the ball—managing capacity—and accepting that broad strokes are enough to get you there. A lightweight model will usually do the trick and, aside from it being easier to build, has other advantages:
- Your employees will actually know how to use it, so wherever they must navigate it—entering forecasts, for example—they’ll be able to do so easily and consistently, meaning you’ll get better estimations from more people, faster.
- When it’s time to talk priorities with your department leads, a simpler model will keep your audience on track. Generally, people will accept the liberties your model takes if they can follow how it works, and that means the team can get on to weighing the merits of proposed initiatives rather than quibbling about how you calculated something. And this is the appropriate measure of your model’s success: the quality of the conversation it compels, not its precision.
So if a lightweight model is sufficient, how do you build a good one? One of our rapid-growth SaaS clients manages capacity with great success using classic principles of supply/demand. Their process can be summarized in the following four steps.
Step 1: Set the stage
Determine how often you will refresh your capacity model. The client refreshes their model quarterly to reflect the inherent uncertainty people encounter when they try to forecast anything more than a few months out. Next, identify what criterion should connect supply and demand. The default is skill type, though you might organize by product teams if you are working in a product-oriented operating model. Whatever criterion you pick, it must be one by which your people—not just you—can intuitively estimate both the supply and the demand side of the equation.
Step 2: Estimate supply
Getting supply is straightforward—you calculate the amount of productive time between two dates and multiply it by the number of resources you have. But beware of a few important considerations. First, rarely is all that time available for planned transformational work; you should reserve some for planned and unplanned keep-the-lights-on work. You can do this in the aggregate, by individual, or by product team. Then you need to cut what time you do have for planned work by general productivity (i.e., not every hour in a day will be hands-on-keyboard development time). Summarize your supply by skill set, product team, or whatever criterion you chose to link supply and demand.
Step 3: Estimate demand
List all the work and projects you might undertake in the quarter (or other time period) you’re forecasting—that’s work under way as well as new initiatives.
Then, you need to estimate the size of each piece of work for that quarter (or time period)—not the size of the entire project if you anticipate it will extend beyond the period you’re forecasting. And you must estimate the size by the same denominator by which you calculated supply—hours, FTEs, points, etc.
Finally, you need to estimate how that total size breaks down by the criterion you chose to connect supply and demand—that is, by skill set or by product team. As an example, you might plan to fulfill 100 hours of a 1,000-hour project in the upcoming quarter. And you might expect each of five unique skill sets to have an equal hand in fulfilling that work. Your demand estimate, then, would be 20 hours for each skill set.
Step 4: Prioritize
Once you have supply and demand, bring them together to see the surpluses and deficits. Have conversations. Determine whether you should onboard new workers to address deficits. If so, how many? If there is a surplus of hours for a particular product team, determine if it’s feasible to slate additional prioritized work for the upcoming quarter. Include and exclude different combinations of projects and assess whether you have the supply—across skill sets, or product teams—to cover them. Let leaders make their case as to why their projects should be included in this quarter’s slate of work. Look to someone with the mandate to make the final call. Or come to a decision as a group.
The ultimate goal: data-driven expectation management
At the end of the day, taking this simple approach to capacity management helps tech leaders align on expectations with their stakeholders. A directional, quantitative model that articulates the supply of hours available in a quarter can prevent tech leaders from overcommitting, or provide the data needed to build a case for additional hires when demand exceeds supply.
The quarterly cadence provides regular opportunities to pivot and make tradeoffs based on shifting priorities. But more importantly, it enables transparent, data-driven discussions around expectations with key stakeholders across the organization.