The Business Case & ROI

Smart-building value comes from five pools, and it helps to know which are easy to bank and which take faith. Energy is the classic: analytics and better control - fault detection, optimised HVAC scheduling, model-based control - typically shave a meaningful slice off a building's energy bill, and energy is metered, so the saving is measurable. Maintenance is the second bankable one: predictive maintenance and fault detection catch a failing chiller or a stuck damper before it becomes a callout, a comfort complaint or a catastrophic failure, cutting reactive repairs and extending equipment life.

The other three pools are larger but softer. Space and utilisation - occupancy sensing that reveals you are paying to heat, light and lease desks nobody uses - can dwarf energy savings, because real estate is expensive, but the saving only lands if someone actually gives up the space. Productivity and wellbeing - better air, light and comfort measurably improving how people work - is the biggest number of all and the hardest to defend, because salaries dwarf energy but the causal link is fuzzy. Asset value - a smart, certified, low-carbon building commanding higher rents, lower vacancy and a better valuation - is increasingly real as tenants and investors screen for it. A credible business case leans on the two provable pools and treats the softer three as upside, not the headline.

Money in a building splits into capex - the one-off capital to buy and install sensors, gateways, a platform, integration and commissioning - and opex - the ongoing cost of licences, connectivity, cybersecurity, and the people who actually run the system and act on what it says. The single most common business-case error is counting the capex and forgetting the opex: a twin that nobody is paid to watch quietly rots into a dashboard nobody opens.

Payback is the time for accumulated savings to repay the up-front spend, and it has a characteristic shape (see the figure). You dip negative at go-live as capex lands, climb as savings accrue, cross the break-even line at the payback point - often one to three years for a focused analytics-and-FDD project on a large building - and run net-positive after that. For a sharper decision, discount future savings to today with net present value (NPV) or an internal rate of return, because a rupee saved in year five is worth less than one saved now. Two honest cautions: savings are rarely permanent unless you keep tuning (buildings drift back), and a project scoped as the whole smart building at once almost always has a worse payback than a phased one that banks quick wins first - which is exactly the next lesson.

Take a 20,000 square-metre commercial office. A focused first phase - sub-metering, a cloud analytics platform with fault detection, and someone to act on the faults - might cost a capital sum plus an annual licence-and-labour opex. Suppose analytics and re-tuning trim energy by roughly a tenth and cut reactive maintenance noticeably. On a building with a large energy and maintenance spend, those two provable pools alone can repay the phase inside two to three years and keep paying after - a defensible case you can put in front of a finance director.

Notice what makes it defensible. Every number ties to a meter or a work order - something you can measure before and after. We did not headline the productivity saving, even though it might be larger, because we cannot prove it cleanly; we listed it as upside. We budgeted the opex explicitly, including the human who acts on the analytics, because software that nobody acts on saves nothing. And we scoped one phase, not the whole building, so the payback is fast and the risk is small. That is the difference between a business case and a wish: a business case survives being audited a year later.

It is worth stress-testing the same example the way a sceptical owner would. What if the energy saving comes in at half the estimate? With a conservative, two-pool case the payback stretches but the project still repays - which is exactly why you never headline the soft pools. What if the operator stops acting on the faults after six months? The saving flattens, which is why the opex line funds a named person, not just a licence. What if energy prices move? Model a low and a high case rather than a single hopeful number. A case that has already answered these questions before anyone asks them is far harder to cancel at the next budget review than one that quietly assumed the best.

Even a strong payback can stall on one obstacle: someone has to find the capex up front, and the person who pays is not always the person who saves. In a leased building the owner may fund the sensors while the tenant banks the lower energy bill - the classic split incentive that kills otherwise sound projects. Naming who pays and who benefits, and structuring the deal so they align, is often more decisive than the raw numbers. This is why phasing helps twice over: it shrinks the up-front cheque and it lets early savings fund later stages, so the owner is never asked to bet the whole sum at once.

There are also ways to make the capex someone else's problem. Energy performance contracts, where a provider funds and installs the upgrades and is paid out of the guaranteed savings, and as-a-service models, where analytics or even the whole smart layer is rented as opex rather than bought as capex, both move the risk off the owner. Green-building incentives, efficiency rebates and, increasingly, the plain regulatory necessity of decarbonisation can tilt a marginal case into a clear one. Whatever the structure, the discipline that makes any of it credible is measurement and verification (M&V): agreeing the baseline before you start and measuring against it after, so the saving is a proven fact rather than a hopeful estimate. A case backed by a real baseline and honest M&V survives scrutiny; one built on vendor projections does not. The point of the business case is not to win the pitch - it is to still be true a year later when someone audits it.

The graveyard of smart-building projects is full of technically impressive systems that were never tied to a decision. A twin was bought because a board wanted to look innovative; a platform was installed with no one funded to run it; a dashboard went live and nobody could say which cost it was meant to cut. When budgets tighten, these are the first casualties, because nobody can point to the money they save.

The cure is the discipline this whole course has pushed: start from the decision, not the technology. Name the one or two problems worth money - runaway energy, chronic breakdowns, empty leased space, tenant churn - size them, and buy only the sensing, data and analytics that serve them. A twin with a purpose and a measured baseline is one of the best investments a property owner can make. A twin without one is an expensive screen-saver, and no amount of 3D gloss will save it at the next budget review. The business case is not paperwork you do at the end; it is the reason the project exists.