Business process automation for SMEs: the practical guide

Business process automation means taking a task your business currently repeats by hand, such as copying figures between spreadsheets, chasing an email chain for an approval, or re-typing an order into a second system, and having software do it instead, the same way, every time. It isn't one big system replacing everything at once. In practice it's a series of specific changes: how a quote turns into a job, how stock levels update, how a customer's status moves from enquiry to invoice. Each change should be visible on its own, which is why the businesses that get this right tend to fix one painful process, prove it works, and use what they learn to plan the next one, rather than attempting a single company-wide overhaul. It's also worth being clear about what automation isn't: it isn't the same thing as artificial intelligence, and it doesn't require one. Automate the workflow first. Give it consistent rules and a reliable record of what happened. Any AI you add later needs that record. You cannot skip to AI and get a useful result.

Spreadsheets are a genuinely good tool for a genuinely small job, which is exactly why they're so hard to leave behind. Nobody builds a bad spreadsheet on purpose. The trouble is what happens as a business grows around one. Watch for a small set of recurring signs: only one person really understands how the sheet works, and the business quietly depends on them staying; the same figures get copied between two or three files, and they occasionally disagree with each other; getting a straight answer to a simple question means opening several tabs and reconciling them by eye; and formula errors or version mix-ups have caused a real problem at least once. Any one of these is survivable. Once two or three combine, the spreadsheet has stopped being a tool the business uses and started being a risk the business carries. A useful test: if the person who built the sheet went on holiday for a month, would the business notice immediately? If the honest answer is yes, that's less a spreadsheet problem than a single point of failure. That is usually the sign that the process should be a system rather than a file.

The instinct is to automate everything at once; the better approach is to pick the one process causing the most pain and fix that first. Look for the workflow where the cost of doing nothing is most visible. That is the one taking the most staff time, generating the most errors, or stopping someone from getting a straight answer when a customer asks for one. That's usually quoting, job tracking, or whatever step turns an enquiry into revenue. Fixing that one process first does two things: it delivers a result the business can actually feel within weeks rather than months, and it tells you far more about what the second process needs than trying to plan everything up front ever could. A phased approach beats a single big-bang launch far more often than it loses to one. If more than one process is causing pain, resist the urge to solve them together. Pick the single workflow with a result the rest of the business can see, so they have a reason to trust the next phase.

No-code and low-code platforms have made a genuine difference for small, well-defined jobs. A simple approval flow, a form that writes to a spreadsheet, or a single automated notification are examples. If that's the whole problem, a no-code tool is usually the faster and cheaper route, and we'll say so plainly. They start to strain once a workflow has real business logic behind it: conditional pricing, multi-step approvals that vary by customer type, or reporting that needs to pull consistent numbers from several sources. At that point you're often paying a monthly platform fee for a tool you're still working around, rather than a tool that fits. That is when bespoke software development is the better fit. The system follows how you work. You do not have to fit the work around a template built for thousands of other companies. The test is fit. If a no-code platform covers what you need today and can grow with you, we will say so.

For the systems we build, most bespoke automation projects land between £15,000 and £80,000, with the number driven by how many workflows are involved, how many existing tools need to talk to each other, and how much reporting the result has to support. Every project starts with a fixed-scope analysis phase, so you know the full cost before committing to a build. It is not an estimate that grows once development is underway. Delivery is then phased. The process that takes the most staff time goes live first, typically as a working first release within 8 to 12 weeks, and later phases are priced and agreed once you can see what the first one actually saved you. You're never committed to the whole spend up front, and you're not paying for a specification exercise that produces a document instead of working software.