Before the instinct
Four steps to an honest automation ROI estimate.
This is not a complex financial model. It is a structured way of making explicit the numbers that are usually left implicit in automation decisions. Each step takes five minutes.
01
Calculate the current cost
Count how many times the manual task happens per week and how long it takes each time. Multiply by the blended hourly cost of the person doing it (salary + benefits + overhead, divided by working hours). This is the weekly cost of keeping it manual. The annual cost is this number times 50. Most teams discover this number is higher than they expected because they have been thinking about individual instances rather than aggregate volume.
02
Estimate the automation cost
What will the automation cost to build? This is either a development cost (hours times hourly rate, or a quoted project cost) or a tool subscription cost. Add any ongoing maintenance cost, typically 10-20 percent of the build cost per year for custom automation, or the subscription cost for no-code tools. This is your automation investment.
03
Calculate the payback period
Divide the automation cost by the weekly savings to get the payback period in weeks. An automation that costs GBP 2,000 to build and saves GBP 150 per week pays back in 13 weeks. After that, it is pure savings. An automation that costs GBP 8,000 and saves GBP 50 per week takes three years to pay back, which is long enough that the underlying process will probably have changed before the investment is recovered.
04
Apply the haircut
Automation rarely saves 100 percent of the time spent on the manual task. There is setup time. There is exception handling time. There is the time spent fixing things when the automation breaks. Apply a realistic haircut, typically 60-70 percent savings rather than 100 percent, and recalculate the payback period. If the investment is still justified after the haircut, the automation is worth building.
The hidden costs
What the basic calculation misses, and why it matters.
The four-step calculation is a good starting point. It misses some costs that are worth adding when the investment is large enough to warrant more careful analysis.
Add these when the stakes rise
Costs the basic model omits
Data cleanup cost: if the automation requires clean data to run correctly and the current data is not clean, add the data cleanup cost to the build cost. This is often the biggest hidden cost in automation projects.
Exception handling cost: some proportion of instances will not match the automation's happy path and will require manual intervention. Estimate this proportion and add the ongoing cost to the savings calculation.
Opportunity cost of the build: building the automation takes time that could be spent on other things. If the build takes three weeks of developer time, that is three weeks of other work that did not happen. Include this in the investment if it is real.
Risk cost: what is the cost if the automation misfires? If a misfired automation sends a message to the wrong person or updates the wrong record, that has a cost. Weight the investment by the risk profile of the automation.
The decision rule
If the honest payback is under six months, build it. Over two years, do not.
Six months is not magic. Most operational processes change enough within two years that an automation built for today needs reworking, a six-month payback recovers the cost before that first major update.
The realistic number
0%
Actual time savings after applying a conservative estimate
Teams that assume 100% savings and build to that number get surprised. Teams that build to 60% and hit it have a system that justified the investment.
Common questions
Questions about automation ROI
When the payback period after a conservative estimate is under six months. Use the four-step calculation: current cost per week minus realistic savings, divided into the build cost. If the investment is recovered in under six months, build it. Over two years, do not. Between six months and two years, the decision depends on how likely the underlying process is to change before the investment is recovered.
Overestimating savings and underestimating build cost. Teams assume 100 percent time savings when the realistic number is 60 percent after exception handling. They forget to include data cleanup costs. They do not account for the first few weeks of parallel running and adjustment. Running the numbers with a haircut applied before committing is the most useful thing the calculation can do.
Yes, if it is real. The cost of not automating includes the compounding time cost of the manual process, the opportunity cost of the attention it consumes, and the risk cost of tasks that get missed or delayed because they depended on someone remembering. For high-volume, time-sensitive follow-up processes, the cost of not acting is often larger than the build cost.
Before investing in the automation, run the four-step calculation.
Bring the task, the volume, and a sense of what the build would cost. The calculation usually produces a clear answer, and in the cases where the payback period is borderline, it surfaces the assumptions that need to be tested before committing.
Bring the page, report, or workflow as it is now.
We reply with the clearest next step, or an honest no.
