Plain answers
AI automation for businesses, explained
The questions businesses actually ask before they automate anything, answered without the sales pitch: what it can do, what it cannot, what it costs, what happens to your data, and how to tell which jobs are worth it.
Start with the work, not the technology #
Almost every useful answer in this section comes back to the same idea. A business does not have an automation problem, it has a list of jobs it repeats. Some of those jobs run on rules: the same trigger, the same steps, the same output, every time. Some need a person, because they need judgement, or a relationship, or someone to take responsibility for a decision.
Automation is worth doing on the first kind and nowhere else. That is why our free tools count hours and jobs rather than talking about models, and why the audit starts by walking through a normal week rather than by recommending anything.
What these pages will not do #
They will not tell you which product to buy, or hand you a recipe to wire two tools together. Partly because the honest answer depends entirely on what your week looks like, and partly because working that out properly is the job we get paid for. What they will do is let you judge, before you spend anything, whether there is enough repeated work in your business to be worth automating at all.
AI automation, explained
Plain answers to what businesses ask before they automate anything.
- What is AI automation? What it means in a business, and how it differs from the software you already run
- What is an AI agent? How an agent differs from a chatbot and from ordinary automation, and when you should want one
- Which jobs are worth automating The test that decides it: how repeatable the job is, and how much judgement it needs
- When not to automate something Rare work, unstable processes, judgement in disguise, and work nobody owns
- How much does AI automation cost? The four parts of the cost, what drives them up, and how to read a proposal
- UK GDPR and AI: what it asks of you The duties that apply to a business the moment an AI tool touches personal data
- Is ChatGPT GDPR compliant? Why the question has no yes or no answer, and how to check whether your data trains the model
- What the ICO says about AI Where the regulator has actually spoken, including the guidance that is under review
- What a data processing agreement is for Why UK GDPR requires one, and the eight things the ICO says it has to cover
- Do you need a DPIA? When an automation project triggers a data protection impact assessment, and when it does not
- Questions to ask an AI supplier The answers a supplier should be able to give in a sentence each, before you sign anything
- What happens when AI gets it wrong How these systems fail, why the failures look plausible, and who carries the consequence
- What should always wait for a person Why anything that sends, spends or deletes needs approval, and what a sensible check looks like
- What to tell your team How to announce an automation without the whole room hearing the word redundancies
- Will AI replace my staff? What happens to the person whose job was half admin, and what the good version looks like
- Your staff are already using AI What to do about the accounts nobody authorised, without turning it into a disciplinary matter
- When the tools change underneath you Why an automation that worked in March stops working in October, and what to expect of a supplier
- What being locked in looks like The four kinds of lock in, how to spot each one early, and the questions that prevent it
- Who owns what gets built Accounts, code and data, and what actually happens if your supplier disappears
- What a free audit involves What happens on the call, what to have ready, and what you keep whether or not you go ahead
Free audit
Find out what the repeated work costs you
The audit counts every job the business repeats, ranks them by the hours they eat, and maps the five worth automating first. It is free, you keep everything, and there is no obligation at the end of it.