A no-hype guide to using AI in your business
“AI” is doing a lot of heavy lifting in marketing right now. In a real business the useful version is narrow and practical: software that reads messy documents, drafts text, and classifies information — under human review. Here’s how we think about it.
What AI is genuinely good at here
- Reading documents. Pulling fields and figures out of invoices, forms, contracts, and emails — including scans and photos.
- Classifying and routing. Suggesting the right category or owner based on history, then learning from your corrections.
- Drafting communication. First drafts of customer emails, summaries, and reminders.
- Answering internal questions. “What’s our policy on X?” against your own documented procedures.
What to keep humans firmly on
- Judgement calls — anything requiring interpretation or sign-off.
- Final review — AI drafts, a human approves. Always.
- Customer relationships — the reassurance and advice people pay for.
The pattern is assist, not replace: AI does the first 80%, your team does the last 20% that matters most.
Staying safe with your data
This is non-negotiable. When we set up AI tools we insist on:
- Data residency and retention you control — know where data goes and how long it’s kept.
- No training on your data unless explicitly agreed.
- Access controls — the tool sees only what it needs to.
- An audit trail — who ran what, and when.
If a vendor can’t answer those four questions clearly, that’s your answer.
The honest summary
AI won’t run your business. Used well, it removes a layer of manual work and gives your people better first drafts to work from. That’s a smaller promise than the headlines — and a much more reliable one.