How to automate business processes

Pick one task that happens every week, follows the same steps, and has a clear owner. Not the most complex one.

Business process automation is using software to run the repeatable steps of a process, so the work gets done the same way every time without someone doing it by hand. It works when the steps are the same every time and the decision is clear. That one rule decides most of what’s worth automating.

• Frequent.

  It happens every day or every week, so the time adds up.

• Rule-based.

  The steps are the same every time, with no judgment call in the middle.

• Slow or error-prone.

  It’s tedious by hand, or mistakes cause real problems.

• Clear owner.

  One role is responsible for it today and will own it after.

Where small businesses usually start

A quick test: if someone copies information from one screen and pastes it into another more than a few times a week, that’s a candidate.

Follow the task from trigger to finish, write the steps down, and cut what shouldn’t be there before you automate anything.

1. 01

   Map it.

   Follow one task from start to finish. Note every system it touches and every handoff.

2. 02

   Write it down.

   If you can’t write the steps, you can’t automate them. Our guide on how to write an SOP shows how.

3. 03

   Clean it up.

   Remove steps that shouldn’t exist. Automating a messy process just makes the mess faster.

4. 04

   Write down the before numbers.

   How long it takes by hand and how often it goes wrong. You’ll need them in step 5.

Use rules for anything structured, AI only where something needs reading or writing, and keep people on the final call.

• Rule-based

  What it doesMoves data and triggers actions on fixed rules, the same way every time.

  Use it whenThe inputs are structured and the steps never change.

• AI automation

  What it doesAdds reading, sorting or drafting to an automated process.

  Use it whenThe input is messy, like emails or documents, and a person can check the output.

• AI agent

  What it doesWorks through several steps toward a goal on its own, using your tools.

  Use it whenThe job is narrow, the access is limited, and a person approves what matters.

A simple example: a customer email comes in. A rule files it and starts the process. AI reads it, works out whether it’s an order, a question or a complaint, and drafts a reply. Rules route it to the right person, who checks the draft and sends it.

AI shouldn’t make the final call on pricing, hiring, promises to customers or payments. It can prepare all of them; a person decides. Build in your own accounts, on plans that don’t train on your data, and give each tool only the access it needs. More in our guide on where AI fits first.

Build the smallest version that removes the manual step, test it on real work, and add a person and alerts before it goes live.

1. 01

   Choose the trigger and tools.

   Use what you already have if it can do the job.

2. 02

   Build the small version.

   Add the edge cases later.

3. 03

   Test with real cases,

   including the odd ones. Fix what breaks.

4. 04

   Add a checkpoint and alerts.

   A person approves anything that reaches a customer or moves money, and someone hears about it when a run fails.

5. 05

   Hand it to an owner,

   with a short note on how it runs.

Off-the-shelf, no-code or custom?

• Built-in

  Good forAutomations inside tools you already use: your CRM, store or accounting system.

  TradeoffLimited to what that tool supports.

• No-code

  Good forConnecting tools with platforms like Zapier, Make or n8n.

  TradeoffCan sprawl if nobody owns it.

• Custom

  Good forCode for a process that’s core to the business or high volume.

  TradeoffTakes longer to build and needs maintaining.

For a fitness studio Lockhold worked with, booking, email and SMS sequences were built as systems from day one, connected to the front-end customer experience. The studio grew 11% month over month after those systems went live.

The task gets done without anyone touching it, the time is free, and the errors are gone. Measure all three against your before numbers.

• Time saved.

  How long it took by hand, and how much of that is gone.

• Errors.

  How often it went wrong by hand, and how often it goes wrong now.

• Response time.

  How long customers or your team wait for the next step.

Add them to your KPI dashboard and keep checking. Tools update and a form field gets renamed. A short monthly look at the failure alerts catches drift before customers notice.

The mistakes that break automations

• Automating a broken process.

  The automation inherits every problem.

• No owner.

  It drifts, then breaks quietly.

• No alerts.

  A failed run just stops the work.

• Too many tools.

  More to manage, not less.

• Automating judgment.

  If a person needs to decide, keep the person there.

Most owners know what they want automated. What they don’t have is the time to map it, clean it up and build it properly.

Our AI and automation work does exactly that: we map the process, cut what shouldn’t be there, build it in your accounts, and hand it over running.

Scope your project →