English
English (US)
English (AU)
English (NZ)
English (SA)
Français
Français (CA)
Italiano
Deutsch
Español
Nederlands
Português
Polski
العربية
עברית
Svenska
Dansk
Suomi
Norsk
Two things make a return-on-investment case in this market believable: it is built from your own operational data, and it counts every cost. This guide sets out how to build one that survives the twelve-month review — the six pools returns actually come from, the costs that get quietly omitted, three worked examples, and the measurement table to set up before anything is installed.
Why most ROI conversations in this market go wrong
Two failure modes dominate.
The remedy for both is the same: build the model from your own operational data, count all the costs, be deliberately conservative on the benefits, and show the arithmetic so that anyone can challenge a line.
The first question: what is the status quo already costing?
Investment decisions in this category are usually framed as “should we spend money on this?” That is the wrong frame, because it treats the current position as free. It is not.
An organisation without visibility of its vehicles and assets is already paying — in fuel burned while stationary, in claims settled because nothing could be proved, in equipment replaced because it could not be found, in hours spent reconstructing what happened, and in work not won because a safety record could not be evidenced.
The correct comparison is not cost versus zero. It is cost of the solution versus cost of the problem. In most operations that comparison is not close, which is why the honest business case in this market rarely needs exaggeration.
The exercise worth doing first: add up what the current situation costs annually. Put a single number on it. That number is the size of the prize, before any technology is discussed.
Where the returns actually come from
Returns cluster into six pools. Each behaves differently — different size, different speed, different certainty — and a credible model treats them separately rather than merging them into one optimistic total.
Fuel and energy
Mechanism. Less idling, fewer speeding miles, smoother driving, fewer unnecessary miles, less unauthorised use, better route sequencing.
How to calculate it. Take your annual fuel or energy spend and apply a conservative percentage. If you do not yet know your idling proportion, assume the improvement sits at the lower end. Do not model separate savings for idling, speeding and route efficiency and add them together — they overlap, and adding them will overstate the case substantially.
The honest caveat. The initial improvement fades if nothing reinforces it. Fleets that review and coach hold the saving; fleets that install and forget see consumption drift back within a year.
Collisions, claims and insurance
Mechanism. Four distinct effects — exoneration (disputed and 50/50 claims defended because the evidence exists), severity (early evidenced notification shortens the window in which third-party costs, credit hire and inflated claims accumulate), frequency (fewer collisions through coaching driven by real evidence) and premium and terms (realised at renewal, not before).
How to calculate it. Pull three years of claims. Separate at-fault from disputed from not-at-fault. Apply exoneration to the disputed pool only, frequency reduction to the at-fault pool only, and severity reduction to the remainder. Model premium improvement in year two, at a conservative rate, clearly flagged as an assumption.
The honest caveat. Exoneration is high-certainty if you have disputed claims history. Frequency reduction is medium and depends on execution. Premium improvement is the least certain and should never be the line the case rests on — treat it as upside, not as foundation.
Productivity and utilisation
Mechanism. More jobs per vehicle per day, less time locating vehicles and equipment, better scheduling, fewer wasted journeys, less waiting.
How to calculate it. The defensible version is narrow and specific: an additional job per vehicle per week, at your average job margin.
The honest caveat. Resist the temptation to model a 15% productivity uplift across the whole operation. Finance directors have seen that claim before and it will contaminate the credibility of the lines that are solid.
Administration
Mechanism. Automatic mileage capture, digital walkaround checks, automated timesheets, proof of attendance and delivery without paperwork, and the elimination of re-keying data between systems.
How to calculate it. Count the roles, the hours per week and the fully loaded hourly cost. Then state explicitly whether this is a cash saving or a capacity release.
The honest caveat. Time released is only cash if the headcount reduces or the released time is redeployed to revenue-generating work. Say which, honestly, and treat capacity release as a soft benefit sitting outside the payback calculation.
Maintenance and asset life
Mechanism. Servicing by actual use rather than by calendar, earlier fault detection, fewer roadside failures, less damage from harsh driving, better warranty and defect evidence.
How to calculate it. The most defensible line is avoided roadside failures. Count last year’s, apply a modest reduction, and use your actual cost per event including recovery, downtime and the knock-on to jobs.
The honest caveat. Extended component life and reduced wear are real and well-evidenced across the industry, but hard to attribute cleanly in a single year’s accounts — describe them rather than quantify them unless you have good historic data.
Loss prevention and recovery
Mechanism. Theft deterrence, rapid alerting, improved recovery rates, prevention of unauthorised use, and the end of off-hire leakage. See our guides to plant and machinery theft and tool theft and van security.
How to calculate it. Take your three-year loss history and gross it up for the costs that never reach the claims file — emergency hire, project delay, management time, excess. Then apply a conservative prevention-and-recovery rate. Model off-hire leakage separately.
The honest caveat. Honest modelling requires stating the probability rather than assuming every future theft is prevented. Off-hire leakage is often the larger number and it is close to certain once visibility exists.
The cost side: what to include
A credible model counts all of this. Omitting any of it is where business cases lose their authority at the twelve-month review.
| Cost | Notes |
|---|---|
| Hardware | Per device; higher for multi-camera and hardwired plant installations |
| Installation | Including vehicle downtime and the cost of getting vehicles to installers |
| Subscription | Software, connectivity and support, for the full contract term at the full rate — not the introductory rate |
| Data overage | What happens in a month with high footage demand |
| Integration | Building and maintaining connections to other systems |
| Training | Initial, plus re-training as people change roles |
| Internal resource | The largest routinely omitted cost. Someone must review events and run coaching |
| Refit and churn | Moving devices between vehicles as the fleet changes |
| Exit | Data extraction, de-installation, and what happens to bundled hardware |
On internal resource specifically. This is the line that most often turns an over-optimistic case into a realistic one, and it is the line that most determines whether the benefits appear at all. A programme with nobody assigned to run it delivers the passive benefits — exoneration, theft alerting, basic visibility — and none of the active ones. That may still be a positive return. It is simply a much smaller one, and the case should say so rather than assume resource that does not exist.
Three worked examples
Illustrative only. Every figure below is constructed to show the method — which lines to include, how to keep them from overlapping, and how conservative to be. Replace all of them with your own before use. Figures are in pounds sterling.
The shape of the return over time
Returns do not arrive evenly, and a business case that implies they do will underdeliver in month three and be judged on it.
What destroys the return
The five failure modes, in order of frequency.
Presenting the case
For finance
Lead with payback period, then total cost of ownership across the full term, then annual net benefit. Show every assumption as a separate line that can be challenged. Include a sensitivity analysis at 50% of modelled benefit. Separate hard cash savings from released capacity and exclude the latter from payback. Volunteer the risks before they are asked for.
For the board
Lead with risk, not savings. Work-related road risk is a governance and duty-of-care matter; equipment loss is a balance-sheet matter; demonstrable control affects insurability, contract eligibility and corporate exposure after a serious incident. The question that lands hardest is the simplest: if there were a fatality involving one of our vehicles next month, what could we produce to show we had managed the risk?
For operations
Lead with the problems it removes — the disputed jobs, the “where is it” calls, the paperwork, the arguments with customers — and be honest about the new work it creates.
For the workforce
Lead with protection, and mean it. Most footage that gets used clears the driver. Say so, and then prove it the first time it happens.
Measuring it afterwards
Set these up before installation, not after. Most are standard reports on a fleet software platform; the point is to record the starting value while you still can.
| Metric | Baseline | 3 months | 6 months | 12 months |
|---|---|---|---|---|
| Fuel or energy per mile | ||||
| Idling as % of engine hours | ||||
| Collisions per million miles | ||||
| Total claims cost, rolling 12 months | ||||
| Proportion of claims defended successfully | ||||
| Average days to notify insurer | ||||
| Unscheduled maintenance events | ||||
| Asset utilisation | ||||
| Equipment losses | ||||
| Admin hours on manual records | ||||
| Events reviewed and coaching sessions held |
That last row is the leading indicator. Every other line follows it. If coaching sessions are not happening, the rest of the table will not move — and you will know why before the renewal, rather than during it.
Returns that are real but do not appear in the model
Worth stating in the case even though they should not carry the payback.
- Contract eligibility. Where clients require evidence of managed road risk or a specific accreditation, the technology is not a cost saving — it is a condition of bidding. The relevant number is the value of the work that would otherwise be inaccessible.
- Corporate exposure after a serious incident. The difference between an organisation that monitored, identified risks and acted, and one that did not, is not a line in a spreadsheet. It is the difference between two very different outcomes.
- Recruitment and retention. A demonstrable commitment to protecting people is increasingly noticed by the people being recruited.
- Ending the arguments. Disputes settled by data rather than by whoever argues longest. Difficult to price, universally valued by the people who currently have those arguments.
- Knowing. The ability to answer a question about your operation with a fact instead of an estimate changes how an organisation makes decisions, and the effect extends well beyond the vehicles.
A closing note on figures
Every number in the worked examples above is illustrative. They are constructed to show the method, not to predict your result.
Replace all of them. A business case built on your own fuel spend, your own claims history, your own loss record and your own hourly costs will be smaller than the industry averages, harder to argue with, and correct. That combination is worth considerably more than an impressive slide.