# How small businesses should think about AI ROI Source: https://getstartedwithai.net/blog/how-small-businesses-should-think-about-ai-roi Site: Get Started With AI Tagline: Build AI automations that actually run your business. Audience: solo professionals, freelancers, consultants and AI implementers in training Contact: hello@getstartedwithai.co.uk This Markdown companion is generated from the prerendered page for AI crawlers. # How small businesses should think about AI ROI AI ROI for a small business is hours reclaimed per week, multiplied by the rate of the person doing the work, multiplied by 50 working weeks, minus the build cost, divided by the build cost. That single calculation produces a year-one return in percent that survives a finance partner reading it. AI ROI calculations from large consultancies do not transfer cleanly to small business decisions because they weigh against revenue uplift, customer lifetime value and category-level forecasts. None of those are useful when the question is whether a single owner-operator should spend £3,500 to save eight hours a week. This post is the small-business version, with the working, a worked spreadsheet example, the time-value adjustment for owner hours, the 12-month ROI timeline with milestones, and the cases where ROI is the wrong question to ask. ## The single calculation that matters For a UK SMB, AI ROI is hours reclaimed per week multiplied by the rate of the person who used to do the work, multiplied by 50 working weeks. Subtract the build cost. Divide by the build cost. That number is the year-one ROI in percent. The calculation survives a finance partner reading it because each input is verifiable: the hours come from the Time Tax conversation, the rate comes from the payroll figure or the consultants own day rate, the build cost comes from the audit recommendation. Most enterprise ROI calculations include uplift assumptions that cannot be verified in advance; the SMB version sticks to costs that already exist and savings that can be measured. ## A worked spreadsheet example A solo practitioner spends three hours a week drafting client follow-ups. Their effective rate is £80 an hour. Annual saving: 3 hours x £80 x 50 weeks = £12,000. Build cost: £3,500 for the audit and first build. Year-one ROI: (£12,000 minus £3,500) divided by £3,500 = 243 percent. The build is worth doing, with substantial margin for the build coming in over budget. A second example: a 15-person professional services firm with two fee earners each spending five hours a week on the same workflow. Annual saving: 10 hours x £150 x 50 = £75,000. Build cost: £8,500 for the audit and first build. Year-one ROI: (£75,000 minus £8,500) divided by £8,500 = 782 percent. The math runs at scale. ## The time-value adjustment most ROI posts skip The hours-saved calculation undercounts the real value of owner time. Owner hours are not interchangeable with employee hours; a founder hour spent on advisory work or business development typically generates two to four times the revenue of the same hour spent on administrative drafting. The right ROI calculation for owner-operator hours uses the next-best alternative use of the time, not the salary equivalent. A founder whose effective billable rate is £200 an hour but who could spend the saved hour on advisory work at £400 an hour should use the £400 figure in the ROI calculation. The honest test: what would a partner-level person genuinely do with the next saved hour. The answer drives the right rate. ## When ROI is the wrong question For builds where the value is non-time-based, the hours-saved calculation undersells the value. Defending a brand voice across client emails (the value is reputational, not a time saving). Catching compliance risks before they land (the value is avoided cost, not productive hours). Opening a new revenue line by serving prospects who could not previously be reached (the value is gross margin, not time). For each of these, the right move is to add a separate column to the same spreadsheet for the non-time benefits and assign a conservative pound value to each one. The conservative number tends to be plenty; the build still clears the threshold. ## A 12-month ROI timeline with milestones - Month 0: audit complete, candidate workflow chosen, Time Tax baseline measured. Build commissioned. - Months 1 to 2: build deployed, fortnight of output review complete, prompt and trigger updated based on real outputs. - Month 3: first measurement against the success metric. Time saved per week tracked. Most builds clear 80 percent of the projected saving by this point. - Months 4 to 6: monthly sample reviews, drift caught early, ROI tracking against projection. - Month 6: payback achieved on most builds with four-plus affected fee earners. ROI typically running 100 percent year-one even before second-half compounding. - Months 7 to 9: time freed up gets reinvested in higher-value work. The compounding benefit on owner-operator builds shows up here. - Months 10 to 12: full year-one ROI measurement. Most builds land between 200 percent and 800 percent depending on rate and volume. ## When not to measure ROI on AI A counter-intuitive point. Some early experiments in a business should not be ROI-measured because the experiment is itself the learning. A solo founder spending £30 a month on a Custom GPT to learn what AI assistants can do is not buying a business outcome; they are buying time on the technology before any client engagement starts. A finance team running a small Custom GPT on internal note drafting is not solving an ROI question; they are calibrating the firm voice against AI drafting before the higher-stakes use cases ship. The right framing for these is small bets at under £100 a month total, with no formal ROI tracking, where the goal is operational fluency rather than measured savings. The audit conversation usually surfaces both kinds of candidate; ROI math gates the bigger builds and not the small bets. ## The 200 percent threshold inside the WIRE Method Every WIRE Method engagement starts with this calculation in the audit conversation. If the math does not clear 200 percent year-one ROI on the chosen process, the build is not the right first project. We pick a different process from the candidate list or recommend the free Starter Kit instead. The 200 percent threshold gives the build margin to come in over budget by 30 to 50 percent and still clear payback inside year one, which is the realistic worst case for a first build. Most of the workflows that reach the audit conversation clear 400 to 800 percent year-one ROI; the ones that do not clear 200 percent are usually low-volume processes where a checklist or a calendar reminder is the right answer. ## What the calculation does not capture Three things the hours-saved calculation misses. The compounding effect of freed-up senior time, which often produces 2 to 4 times the value of the saved hours when reinvested in higher-value work. The risk reduction from removing single-person dependencies, which is hard to value but real. The cultural effect of the team seeing AI work in the operational layer, which makes the second and third builds easier to ship. None of these go into the spreadsheet because all three are hard to measure in advance. They are real and large, and they tend to make the actual ROI run higher than the projected ROI by month nine. The honest practitioner read: the calculation is conservative on purpose, and the conservatism usually under-reports the year-one outcome. ## Build your first AI system live with Richard A 60-minute live build with Richard Batt. Bring one repeatable task. Leave with a mapped automation, success metrics and a deploy plan you can test the next morning. Free seat, no upsell. ### Live build session The next live session covers the WIRE Method end to end on a real audience build. [Join the next session](/webinar) ## Questions people ask before they start ### undefined undefined ### undefined undefined ### undefined undefined ### undefined undefined ### undefined undefined ### undefined undefined