How to Calculate Technology ROI (A Method for Nigerian Businesses)

ROI calculations fail in one of two ways. Either the cost side is understated, because only the invoice was counted, or the benefit side is invented, because somebody estimated a percentage improvement that nobody ever measured. Both produce numbers that look precise and mean nothing.

This article gives you a method you can defend to a sceptical finance colleague: how to scope the calculation, what belongs on each side, how to value time honestly, how to handle a currency that moves, and how to build the baseline that makes the whole exercise possible. It applies to any technology investment — a system, an automation project, a set of tools. Asset-specific versions for websites, apps and marketing follow the same skeleton with different metrics.

The formula, stated plainly

Technology ROI over a stated period is:

ROI (%) = (Total benefit − Total cost) ÷ Total cost × 100

Three companion figures make the result usable:

• Payback period = Total cost ÷ monthly net benefit. Expressed in months, this is often more persuasive than ROI itself.
• Net benefit = Total benefit − Total cost, in naira. This tells you the size of the prize, which a percentage hides.
• Benefit-cost ratio = Total benefit ÷ Total cost. A ratio above 1 means the investment returned more than it consumed.

An ROI of 150% over three years means the investment returned one and a half times its cost on top of paying for itself. The same investment might have a payback period of 14 months. Both statements describe the same project; quote both.

Step 1: Scope the calculation and set the horizon

Decide three things before you touch a number.

1. What exactly is being measured? One system, one project, or the whole technology budget? Mixing them makes the result meaningless. "The ROI of our new order system" is answerable. "The ROI of our technology" usually is not.
2. Over what period? Most business systems should be assessed over two to three years. One year understates the return because the build cost lands all at once. Five years overstates it, because business needs and technology both change.
3. Compared with what? ROI is always relative to an alternative: doing nothing, hiring more staff, or a different system. State the comparison explicitly, because it determines which benefits count.

Write these three lines at the top of the sheet. Every later argument about the number traces back to one of them.

Step 2: Build an honest cost side

The cost side is total cost of ownership over the horizon you chose, not the quotation you signed.

Indicative 2026 conventions; maintenance percentages and rates vary by vendor and scope.

Two disciplines make this credible. First, use written quotations rather than remembered figures. Second, include internal time at a real hourly cost. If a manager on roughly ₦600,000 a month spends 40 hours on implementation, that is close to ₦150,000 of cost, and leaving it out biases the result.

Step 3: Build the benefit side from four sources

Every genuine technology benefit falls into one of four categories. Anything that fits none of them is probably not a benefit you can defend.

Three rules keep this honest:

• Count only what changed because of the system. If revenue rose 20% while you also opened two new branches, the system did not cause all of it.
• Use measured values, not assumed percentages. "Staff save 30% of their time" is a guess. "Invoicing took 46 minutes a day and now takes 12" is a measurement.
• Count time saved only if it is redeployed or removed. If a freed-up hour is spent on higher-value work or lets you avoid a hire, it is a benefit. If it is simply idle time, it is not — say so rather than inflating the number.

Step 4: Put a defensible naira value on time saved

Time is where most ROI calculations lose credibility, so make the valuation explicit.

1. Calculate a loaded hourly cost per role. Take the monthly salary, add employer costs, divide by the number of working hours in the month. A staff member on ₦250,000 a month working about 176 hours has a base cost near ₦1,420 per hour; loaded, treat it as somewhat higher.
2. Measure the task, twice. Time the process before the change and again after it has been in use for a month. Use real observations, not recollection.
3. Multiply by frequency. Minutes saved per instance, times instances per month, times twelve.
4. Apply a realism discount. Not every saved minute converts to value. A discount of 25–50% on time-saving benefits is a defensible convention and makes the result far harder to dismiss.
5. State what happens to the freed time. Redeployment, avoided hiring or reduced overtime — name it.

A worked line looks like this: 34 minutes saved per day, across 3 staff, 22 working days a month, at a loaded ₦1,800 per hour, equals roughly ₦67,000 a month before discount and about ₦40,000 after a 40% realism discount.

Step 5: Calculate ROI, payback period and the monthly view

With both sides built, produce four figures.

Two adjustments improve accuracy without turning this into a finance exercise:

• Time value. Naira received in year three is worth less than naira spent today, particularly in an inflationary environment. If the horizon is three years or more, discount later benefits, or simply present a conservative case alongside the headline one.
• Sensitivity. Recalculate with benefits reduced by 30% and costs raised by 20%. If the investment still pays back inside the horizon, the case is robust. If it collapses, the case depends on optimism.

Set a baseline before you spend anything

You cannot calculate ROI after the fact if nobody recorded what "before" looked like. Capture the baseline in the two weeks before implementation starts.

• How long does the target task take today, measured not estimated?
• How many times does it happen per day, week or month?
• What does the current tooling cost per month, in naira?
• What is the current error, return or rework rate?
• What is current revenue for the affected channel or product line?
• How many enquiries are received and how many go unanswered?
• How many staff hours are involved, and at what roles?
• What is the seasonal pattern, so you do not compare December with February?

That last item matters more in Nigeria than people expect. Comparing a retail month before Christmas with one in the new year will flatter or damn any system unfairly.

Example (hypothetical): ROI on an inventory system

Example (hypothetical): a building-materials retailer in Ibadan with two branches invests in a custom inventory and order system. Figures are illustrative planning numbers, not quotations or client results.

Cost side over three years

Benefit side over three years

Result: net benefit of ₦3,740,000, ROI of about 39% over three years, and a payback period of roughly 27 months once the build cost and first-year running costs are set against monthly net benefit.

Interpretation: this is a defensible but not spectacular return, and it depends heavily on the stock-loss line. The owner's sensible next question is whether the stock-loss reduction is real and sustained, because a 30% reduction in that single line pushes payback past year three. That question is answerable only because a baseline was recorded before the project started.

What changes when you calculate ROI in Nigeria

• Currency movement affects both sides. Costs denominated in US dollars — hosting, SaaS, AI usage — rise in naira terms over the horizon, while benefits are usually in naira. Model the dollar lines at a conservative rate and show the result at two rates if the exposure is significant.
• Inflation flatters benefits. Revenue and cost-saving figures in later years are in weaker naira. Either discount them or present a like-for-like comparison so a nominal increase is not mistaken for a real one.
• Baseline data is often informal. Many Nigerian SMEs have order records in WhatsApp threads and notebooks. Spend a fortnight structuring that data before the project rather than reconstructing it afterwards.
• Power and connectivity affect realised benefit. A system that is unavailable during outages delivers less than one that works offline and syncs later. If availability is a risk, reflect it in the sensitivity case.
• Labour cost context. Where staff costs are comparatively low, time-saving benefits are worth less per hour than in higher-wage markets, which shifts the case towards revenue, loss avoidance and capacity benefits rather than headcount savings.

Benefits that are real but hard to quantify

Do not force everything into the naira column. List these separately as qualitative factors supporting the decision.

• Faster, more consistent customer response, which affects retention in ways that show up slowly.
• Better data for decisions, including data you may need for a loan application or an investor conversation.
• Reduced key-person risk, because knowledge sits in a system rather than in one person's head.
• Compliance readiness under the Nigeria Data Protection Act 2023 and cleaner record-keeping for tax purposes.
• Capacity to grow without proportional hiring, which is a strategic benefit more than a measurable one in year one.

A short qualitative list next to a conservative quantitative calculation is more convincing than a large number built from assumptions.

Mistakes that produce fake ROI numbers

• Counting only the invoice as cost. Implementation, training, internal time and maintenance are real money.
• Assuming a percentage improvement. Measure the task; do not estimate the gain.
• Claiming all revenue growth. Isolate what the system caused from what the market, the season or a new branch caused.
• Valuing every saved minute at full rate. Apply a realism discount and state it.
• Ignoring the transition dip. Productivity usually falls before it rises.
• Using a one-year horizon for a three-year asset. This kills good investments; using a five-year horizon rescues bad ones.
• Never revisiting the calculation. Recalculate at 90 days and at one year with actual figures, and correct the record when the forecast was wrong.
• Comparing seasons. December versus January tells you about Nigeria's retail calendar, not about your system.

Conclusion

A credible technology ROI calculation is mostly discipline, not mathematics. Scope one investment, set a horizon and a comparison, build the cost side as total cost of ownership including internal time, build the benefit side from revenue gained, cost removed, time saved and loss avoided, discount the time savings for realism, and state payback period alongside ROI. Record the baseline before you spend, run a sensitivity case, and recalculate at 90 days. Numbers built this way survive scrutiny; numbers built from assumed percentages do not.

If you are evaluating a system and want the cost side priced accurately before you model returns, Linestech can scope the work, separate one-off costs from recurring ones, and help you define the baseline measurements worth capturing first.

Frequently asked questions

What is a good technology ROI?

There is no universal threshold, and any figure quoted as a standard should be treated sceptically. A practical test for a Nigerian SME is whether the investment pays back within the horizon you would confidently forecast the business over — often 18 to 30 months — and whether it still pays back in the sensitivity case with benefits cut by 30%. Compare against what else the same money could do.

Can I calculate ROI before I spend the money?

Yes, and you should, but label it a forecast. Build the cost side from quotations and the benefit side from measured baselines and conservative assumptions, then state the assumptions in writing. The forecast's real value is that it forces you to record a baseline and define a success measure, which makes the after-the-fact calculation possible.

How do I calculate ROI when the benefit is better customer service?

Translate service into money through its consequences: retention, repeat purchase rate, and the number of enquiries that previously went unanswered multiplied by your conversion rate and average order value. Where a link cannot be measured, list it as a qualitative benefit rather than assigning a number you cannot support.

Should I include staff salaries in the cost of an internal project?

Include the incremental time staff spend on the project, valued at a loaded hourly rate, rather than their whole salary. Their salary is being paid regardless; what the project consumes is the portion of their time diverted from other work. Recording it keeps the calculation honest and usually reveals that internal cost is larger than expected.

How often should I recalculate?

At 90 days with early indicators, at one year with actual figures, and at the end of the horizon for a final view. The 90-day check is the important one, because it is early enough to act on. If the early numbers are far from forecast, investigate adoption first: most shortfalls are caused by people not using the system rather than by the system failing.

What if the ROI calculation says no but I still think we should do it?

That is a legitimate position, provided you say so explicitly. Some investments are made for compliance, risk reduction, competitive necessity or capability building, and their justification is not financial return. Record the real reason rather than manufacturing benefit numbers to support a decision that was made on other grounds.

How do I handle shared costs across several systems?

Allocate shared costs — hosting, an integration layer, a support retainer — across the systems that use them on a stated basis, such as usage share or an equal split, and note the method. Consistency matters more than precision. What you must avoid is charging shared costs to one system when it suits the argument and ignoring them when it does not.

Sources and further reading

Figures, platform rules and regulations change. These are the primary references behind this article and the places to check before you act on it.

• Nigeria Data Protection Act 2023 — NDPC
• Cloudflare Learning Center
• Nielsen Norman Group UX research