Return on Investment is frequently required as the basis for decisions about technology directions and acquisitions — and equally frequently, it creates controversy and consternation. ROI is often misunderstood and misapplied; the computations can invoke uncertainty even in those who are not otherwise math-phobic.
To help you use ROI confidently and competently, we identify and suggest ways to avoid the following ten pitfalls that commonly affect the suitability of ROI calculations. We start with the simpler issues of calculation, then move to the harder — and often more important — issues of how ROI is interpreted and used.
The "return" in Return on Investment refers to financial benefits, which involve both costs and revenues — both must be counted. Cost savings stop money previously spent (often headcount). Cost avoidance sidesteps an expense that would otherwise become necessary — more palatable, but more prone to exaggeration. On the revenue side, revenue enhancement creates new revenue, while revenue protection preserves revenue that would otherwise be lost — to a competitor, or by failing a compliance requirement that's the price of continuing to operate.
ROI is often faulted for considering only the immediate term. Most solutions operate far longer — over their accounting useful life — and that horizon materially affects both costs and benefits. Include ongoing operation and maintenance: a higher upfront price can still yield a lower total cost of ownership. Revenue impacts vary over time too — initial high sales can fade; slow starters can compound.
A dollar returned a year from now is worth less than a dollar in hand — at 4% interest, you'd need $1,040 next year just to break even. ROI must reflect this time value of money: apply a discount factor to each year's net annual benefit to get its net present value, then sum the yearly NPVs for a single figure in today's dollars.
Many reasonable-looking ROI calculations grossly misrepresent benefits by assuming a revenue dollar offsets an expense dollar one-for-one. It doesn't. Each dollar of cost saved or avoided drops straight to the bottom line as profit — but only the profit-margin portion of a revenue dollar becomes profit, after the added cost of producing that revenue. Spending $1 to gain $2 of revenue can still be a poor investment.
ROI is often built on estimates with no rational link between the solution and the costs or revenues it's claimed to affect. The word "justification" is a tip-off — it means finding arguments for a predefined answer rather than objective analysis. Credible estimates come from someone who: is positioned to reliably know the impact; is aware of the factual factors affecting it; and has a reason to give figures that hold up (personal accountability). This is why buyers reflexively discount vendor ROI as understated cost and exaggerated benefit.
Reasonable estimates go unachieved because organizations don't follow through. Headcount "savings" evaporate when the staff stay on to "do something else" (and Parkinson's Law fills the time). The classic "five minutes a day" fallacy values freed minutes as payroll savings that never actually leave the payroll. And many solutions don't eliminate costs — they shift them: to the front office, to the customer (self-service that alienates), or to contractors who are often the same people at higher rates. Shifted costs aren't saved costs.
Quarantining intangibles into a separate qualitative section turns objective analysis into subjective justification — the decision then rides on the un-quantified part. Responsible ROI demands that all factors be quantified in money terms; the decision-maker is implicitly pricing them anyway. Risk belongs in the numbers too — risk exposure = impact × likelihood Opportunity cost is inherent (choosing one alternative forgoes the others); flexibility is investing now to enable future change. Prefer explicit scenarios over hand-waved likelihood percentages.
Both compare a solution against "No Change" — Business as Usual (BAU). A delta captures only the differences attributable to the solution; it's sufficient when the 80-20 rule holds and a few factors dominate. A scenario models the full set of costs and revenues for each alternative; it's necessary when benefits are small, factors are complex, or several alternatives must be compared. The hard part is knowing when a delta is — and isn't — appropriate.
Few organizations measure what actually happens — and fewer compare it to what they estimated. Without grounding in actuality, ROI estimates rapidly and continually diverge from reality. Matching actuals to estimates establishes confidence in the numbers and continuously improves the accuracy of every future estimate.
ROI benefits must map to what the customer or decision-maker actually values. Most selling touts product features and assumes they'll strike a chord — an unwise assumption. The deep trap is assuming what should be valued instead of ascertaining what is — a trap conventional ROI analyses fall into even when guided by supposed experts. When the value on the page isn't the buyer's value, the case simply isn't persuasive.
ProveIT.net's — now ValueLogics.ai's — ROI Value Modeling™ methodology uses: