The automation ROI calculation
Before commissioning any automation project, you should be able to answer: how much is this process currently costing us, and how much will automating it save?
Calculating the cost of manual processes
Start with time. How many hours per week does your team spend on the process you want to automate? Multiply by the fully-loaded hourly cost of the people doing it (salary plus benefits, typically 1.3–1.5x the gross salary). Multiply by 48 working weeks per year.
Example: A 3-hour weekly manual reporting process done by a senior analyst costing £60k/year. - Fully-loaded hourly rate: £60,000 × 1.4 ÷ 1,920 hours = £43.75/hour - Annual cost: 3 hours × 48 weeks × £43.75 = £6,300/year
That is your baseline. Add error correction cost (how many hours per month do you spend fixing mistakes caused by manual data handling?), delay cost (what revenue or penalty is associated with slow processes?), and opportunity cost (what higher-value work could those people be doing instead?).
Calculating the automation cost
A well-scoped automation for a single process typically costs £3k–£15k to build and £100–£500/month to operate (hosting, third-party API costs, monitoring). Add an annual maintenance budget of 15–20% of the build cost for updates as your connected systems change.
Simple payback period
If the process costs £6,300/year and the automation costs £8,000 to build with £2,400/year in running costs, the net annual saving is £3,900. The payback period is 8,000 ÷ 3,900 = 2.1 years. Automation projects that pay back within 12–18 months are strong investments. 2–3 years is reasonable. Beyond 3 years, you should scrutinise the assumptions.
Which processes are good automation candidates
Not every manual process should be automated. The best candidates share several characteristics:
High repetition, low variation
Processes that follow the same steps every time, with predictable inputs and outputs, are ideal automation candidates. If a process requires human judgment, creativity, or handling of genuinely novel situations in every instance, it is not a good fit for automation.
Clear business rules
Automation encodes your business rules as code. If the rules are unclear, inconsistently applied, or frequently changing, the automation will be fragile and require constant maintenance. Before automating a process, document exactly what the rules are. This documentation exercise often reveals inconsistencies that the manual process was silently absorbing.
Measurable inputs and outputs
For an automation to be trustworthy, you need to be able to verify that it is working correctly. Processes where you can define success clearly — a report was generated, a record was created, an email was sent — are easier to automate reliably than processes where success is subjective.
The three automation traps
Over-engineering
The most common automation failure is building something unnecessarily complex. A process that currently takes 2 hours per week does not need a microservices architecture. Start with the simplest solution that works: a script, a spreadsheet macro, a no-code tool. Add complexity only when it is clearly justified.
Automating a broken process
Automation makes everything faster — including mistakes. If the underlying process has flaws (incorrect data, bad business logic, missing edge case handling), automating it scales those flaws. Fix the process first, then automate it.
Neglecting error handling
Manual processes have a built-in error recovery mechanism: a human notices when something goes wrong. Automated processes can fail silently for days or weeks if error handling and alerting are not built in from the start. Every automation needs monitoring, error notification, and a clear recovery procedure.
Building an automation roadmap
The most valuable automation programme is not a single large project — it is a portfolio of progressively more ambitious automations, each building on the infrastructure and learnings from the last.
Start by mapping all manual processes across your organisation and rating them on two axes: time cost (how much does it cost to do manually?) and automation complexity (how hard would it be to automate?). High-cost, low-complexity processes are your quick wins. Build those first, demonstrate ROI, and use those results to justify the more ambitious projects.
A mature automation programme typically delivers 150–300 hours per year in recovered capacity per automation, compounded across a portfolio of 10–20 automated processes. At that scale, the ROI is transformative.