By Dr. Mazlan Abbas | IoT World
What happens in the room right before an IoT project gets approved?
Someone stands up, presents a compelling vision, sensors connected, data flowing, dashboards lighting up in real time. The technology works. The use case is real. The vendor is credible. And then, somewhere around slide nine, a hand goes up from the CFO’s side of the table.
“What is the return on investment?”
The room goes quiet. The presenter flips to a slide with numbers that were calculated the night before. The figures look confident on screen. But the CFO has seen enough technology projects to know that IoT ROI calculations are often built on assumptions stacked on top of optimism stacked on top of vendor projections.
The project gets deferred. Or approved with a scaled-back pilot that never gets a second budget cycle. Or worse, it gets approved and then cancelled eighteen months later because nobody set the right expectations at the start.
This is not a technology problem. It is a language problem. Engineers speak in capabilities. Boardrooms speak in returns.
Here is a framework that bridges that gap.
Start With the Business Problem, Not the Technology
The first mistake most IoT ROI models make is starting with the solution. A team decides they want to deploy predictive maintenance sensors, then works backwards to justify the investment. The numbers end up reverse-engineered to support the decision that was already made.
A credible ROI model works in the opposite direction.
The starting point is a clearly defined business problem with a measurable current cost. Unplanned downtime is costing the plant RM 2.3 million per quarter. Water leakage in the distribution network is creating non-revenue water losses of 23%. Cold chain temperature excursions are resulting in a product rejection rate of 4.1%.
These numbers must come from finance, from operations, from audit reports. They cannot come from the vendor’s industry benchmark slides. If the business problem cannot be quantified before the technology is selected, the ROI model will not survive a serious boardroom review.
Theoretical Example: Consider a palm oil processing facility in Pahang. Before proposing any IoT investment, the operations team pulls twelve months of maintenance logs. The data shows that the steriliser vessels fail unexpectedly an average of eight times per year. Each failure takes the line offline for fourteen hours. At a throughput value of RM 180,000 per day, each incident costs roughly RM 105,000 in lost production, plus RM 30,000 in emergency repair labour and parts. The total annual cost of this single problem is RM 1.08 million. That number, drawn from internal records and signed off by finance, is now the anchor of the entire ROI model. The IoT solution comes after. The business problem came first.
The Four Components of a Credible IoT ROI
Once the baseline cost of the problem is established, the ROI model needs to capture four distinct categories of value. Most IoT proposals only capture one or two. All four are necessary for a complete picture.
Cost Reduction
This is the most straightforward category and the one most teams default to. If the IoT system reduces unplanned downtime by 40%, and unplanned downtime currently costs RM 500,000 per incident, and incidents occur approximately six times per year, then the addressable saving is RM 1.2 million annually. Apply a realistic capture rate. Not 40% of all incidents will be eliminated in year one. A more defensible assumption might be 25% in year one, scaling to 40% by year three.
Theoretical Example: A beverage manufacturer installs vibration sensors on its filling line motors. The predictive analytics platform flags abnormal bearing wear patterns fourteen days before failure would occur. In year one, the system catches three of the six incidents that would have happened without intervention, reducing unplanned downtime costs from RM 900,000 to RM 450,000. By year two, with refined alert thresholds and better operator response protocols, the catch rate improves to five out of six, bringing the annual saving to RM 750,000 against the baseline. The model does not claim perfection. It claims a defensible trajectory.
Revenue Protection or Enhancement
IoT often protects revenue rather than reducing cost, and these two categories are not interchangeable in a CFO’s model. An agriculture IoT deployment that prevents a crop disease outbreak is not reducing a cost. It is protecting a harvest that would otherwise be lost. A smart retail system that reduces stockouts increases sales that would otherwise not occur. Quantifying these numbers requires working closely with commercial teams, not just operations.
Theoretical Example: A frozen seafood exporter deploys temperature sensors across its cold chain, from the processing plant to the port warehouse. Historical records show that two to three shipments per year are rejected by overseas buyers due to temperature excursion documentation failures, not actual spoilage, just inadequate proof of an unbroken cold chain. Each rejected shipment represents RM 320,000 in deferred revenue and RM 45,000 in re-export logistics costs. The IoT system provides continuous, timestamped temperature logs that satisfy buyer audit requirements. The revenue protection value is RM 640,000 to RM 960,000 annually, a number that commercial leadership can stand behind because it maps directly to documented rejection history.
Asset and Capital Efficiency
When IoT data enables better utilisation of existing assets, it defers capital expenditure. A fleet management system that extends vehicle lifespan by two years does not produce a line on the income statement. But it reduces the capital budget required in year three. A preventive maintenance programme that extends machinery life from eight years to eleven years has a net present value that belongs in the ROI model, even if it never appears in an operating cost analysis.
Theoretical Example: A logistics company operating a fleet of 120 refrigerated trucks deploys GPS telematics combined with engine performance sensors. The data reveals that 30% of the fleet is being driven in ways that accelerate engine wear, specifically high RPM starts and extended idling with cooling units running at full load. A driver coaching programme based on the sensor data extends average engine overhaul intervals from 400,000 km to 520,000 km. With each engine overhaul costing RM 28,000, and the fleet averaging three overhauls per year previously, deferring two overhauls annually saves RM 56,000 per year in direct capital spend. Across a five-year model, that is RM 280,000 in deferred capital expenditure, a real number that belongs in the ROI calculation even though it never touches the P&L directly.
Risk Reduction
This is the most difficult category to quantify and the most commonly left out. Regulatory penalties, environmental incidents, safety failures, and reputational damage all carry financial exposure that IoT systems can reduce. Insurance actuaries do this kind of modelling routinely. IoT project teams should borrow the methodology. An HSE system that reduces reportable incidents has a calculable value based on the average cost of a reportable incident multiplied by the frequency reduction multiplied by the probability of regulatory action.
Theoretical Example: A chemical storage facility handles substances classified under Malaysia’s Environmental Quality Act. In the past three years, two minor containment incidents have resulted in DOSH investigations, each costing RM 85,000 in legal representation, remediation, and operational disruption while inspectors were on site. The facility installs environmental sensors monitoring pressure, temperature, and gas concentration at fourteen critical points, with automated alerts to the operations team and a secondary alert to the emergency response coordinator. The risk model assigns a 60% probability that the IoT system would have detected the conditions leading to both prior incidents early enough to prevent them. The annual risk reduction value is calculated at 2 incidents x RM 85,000 x 60% probability of prevention, or RM 102,000 per year. It is not a guaranteed saving. It is a probabilistic one, and that is an honest way to present it to a board.
Building the Cost Side of the Equation
ROI is a ratio. The numerator is benefits. The denominator is total cost. Most IoT proposals understate the denominator.
A credible IoT cost model includes five categories.
Hardware costs cover sensors, gateways, edge devices, and infrastructure. These are usually the easiest to quantify and the first number a vendor will provide.
Connectivity costs are frequently underestimated. Cellular SIM plans, private LoRaWAN infrastructure, or licensed spectrum all carry ongoing costs that compound over a five-year investment horizon.
Platform and software costs include IoT platforms, cloud hosting, storage, analytics tools, and any integration middleware. Enterprise software contracts have a habit of expanding. Build in a 20% buffer.
Implementation and integration costs are where most projects find their biggest surprises. Integrating IoT data into existing ERP, CMMS, or SCADA systems is almost never a plug-and-play exercise. Estimate implementation at no less than 30% of hardware and software cost combined, and validate that estimate with a systems integrator before presenting to the board.
Operational costs include maintenance contracts, firmware updates, device replacement cycles, and the internal staff time required to manage the system. An IoT deployment is not a set-and-forget investment. It is an operational commitment.
Theoretical Example: That same palm oil facility models its full cost stack for a predictive maintenance deployment across 40 sensor points. The vendor quotes RM 180,000 for hardware. Connectivity via cellular SIM across the remote site adds RM 36,000 over three years. The IoT platform licence and cloud hosting runs RM 48,000 per year, or RM 144,000 over three years. Integration with the existing CMMS takes four months and costs RM 95,000, more than the project champion expected but consistent with the 30% rule of thumb. Annual operational costs including one part-time technician for sensor maintenance and firmware updates are estimated at RM 24,000 per year. The true three-year total cost of ownership is RM 180,000 + RM 36,000 + RM 144,000 + RM 95,000 + RM 72,000, which comes to RM 527,000. That is 193% higher than the hardware quote alone. The ROI calculation looks very different depending on which number goes in the denominator.
The Time Horizon That Makes or Breaks the Model
A 12-month ROI model for an IoT deployment is almost always misleading. Most IoT systems take six to nine months to deploy properly, which means a 12-month model is measuring a system that has been running for three to six months at most.
A three-year model captures the full deployment cost, the ramp-up period where value is being realised at partial capacity, and the period of full operational maturity. A five-year model adds the asset replacement cycle and the second-generation technology refresh.
Present the model in all three time horizons. Twelve months shows the cash flow impact. Three years shows the business case. Five years shows the strategic value. Different members of the board will focus on different horizons. Give them all three.
Theoretical Example: The palm oil facility ROI model presented across three time horizons tells three distinct stories. At 12 months, the project is cash flow negative. Deployment takes seven months, and the system has been live for five months by year end. Benefits captured in that period amount to roughly RM 135,000 against a first-year cost of RM 420,000. The 12-month picture looks painful. At 36 months, cumulative benefits from cost reduction, one avoided capital overhaul, and partial risk reduction total RM 980,000 against total cost of RM 527,000. The project has returned 86% over its cost and is approaching payback. At 60 months, with sensor replacement in year four and a platform upgrade, total costs rise to RM 680,000. But cumulative benefits, now at full operational maturity, reach RM 1.9 million. The five-year ROI is 179%. Three different numbers, three different conversations, all from the same deployment. Giving the board all three prevents the CFO from anchoring on the 12-month loss and missing the full picture.
The Number That Matters Most
A sophisticated boardroom will not stop at total ROI percentage. The three numbers that carry the most weight are payback period, net present value, and internal rate of return.
Payback period answers the question that every CFO instinctively asks: how long until we get our money back? Anything beyond three years for an operational IoT deployment will face serious scrutiny. Two years or less is a strong business case. Under 18 months is compelling.
Net present value discounts future cash flows back to today’s dollars, accounting for the time value of money and the organisation’s cost of capital. A positive NPV means the project creates real value. A negative NPV means it destroys value regardless of how attractive the absolute numbers look.
Internal rate of return is the discount rate at which the project breaks even. If the organisation’s hurdle rate is 12% and the IoT project’s IRR is 18%, the project clears the bar. If the IRR is 9%, it does not matter how impressive the technology demonstration was.
Theoretical Example: Using the palm oil facility numbers, the financial model calculates a payback period of 28 months, just inside the three-year threshold that typically draws scrutiny. Applying a discount rate of 10%, which reflects the company’s cost of capital, the NPV of the project over five years is positive at RM 387,000. The IRR works out to 23%, comfortably above the company’s stated hurdle rate of 15%. Any one of these three numbers alone would be insufficient for a board presentation. Together, they tell a complete story: the money comes back in under two and a half years, the project creates real discounted value of RM 387,000, and the return exceeds the company’s own threshold for capital allocation by 8 percentage points. That is a business case, not a technology pitch.
What a Strong ROI Model Signals Beyond the Numbers
Here is something that rarely appears in project proposals but always appears in the minds of experienced board members.
The rigour of an ROI model is a proxy for the rigour of the team proposing it.
A team that has done the work to quantify baseline costs, build a multi-category benefits model, account for all five cost components, and present results across three time horizons is a team that understands the business problem deeply enough to be trusted with a significant investment.
A team that presents a single headline ROI number with a vendor case study attached is a team that has done marketing, not analysis.
Boards fund the former and defer the latter. Sometimes indefinitely.
The Question Worth Asking Before the Boardroom
The most powerful thing any IoT project champion can do before walking into a budget approval meeting is to ask one simple question internally: if the assumptions in this model are wrong by 30%, does the business case still hold?
Theoretical Example: The palm oil facility team stress-tests its model before the board meeting. They reduce all benefit assumptions by 30% and increase all cost assumptions by 30%. In this pessimistic scenario, the five-year cumulative benefit falls from RM 1.9 million to RM 1.33 million, and total costs rise from RM 680,000 to RM 884,000. The project still returns RM 446,000 over cost. The payback period extends from 28 months to 38 months, which now sits outside the preferred three-year window. The team decides to revisit the connectivity cost estimate, finds a LoRaWAN alternative that cuts five-year connectivity costs by RM 60,000, and resubmits. The revised stress-tested payback is 34 months. Not perfect, but defensible, and the team can show the board exactly what assumptions were tested and how the model responds. That level of transparency is itself a signal of competence.
If the answer is yes, the project has a defensible ROI. If the answer is no, the model needs another iteration before it is ready for the boardroom.
IoT technology has matured. The connectivity is more reliable. The platforms are more capable. The integration options are broader than they have ever been. The gap that remains is not in the technology. It is in the business case discipline that determines whether the technology gets a chance to deliver its value.
What does the next IoT proposal in your organisation look like? Is it being built around a technology demonstration, or around a business problem with a measurable cost and a credible path to return?
That is the difference between a project that gets approved and one that gets deferred until next year’s budget cycle. And the year after that.
Dr. Mazlan Abbas is the CEO of FAVORIOT and a pioneer of the Internet of Things ecosystem in Malaysia. IoT World covers the intersection of IoT strategy, technology, and business transformation across Southeast Asia and beyond.
Leave a Reply