From the East Midlands Tech Week keynote
The Moat Playbook
Six steps to arm your unfair advantage with AI: one for this week, one for this month, one for this quarter, two for this year, and one for every plan you write from now on.
The frame in one paragraph
AI didn’t replace your moat. It made it scalable. Your customer history, trust, expertise and network are things no AI-native startup can copy. The winners of the next three years are relationship-native businesses with AI on top. This playbook is how you become one.
Unlock the full playbook
Checklists for all six steps, the Inventory worksheet, the Build / Buy / Borrow matrix and a 90-day timeline. Free, and it takes 20 seconds.
We’ll email you the playbook link and the worksheets. No spam; unsubscribe anytime.
✓ You’re in. The full playbook is unlocked below. We’ve also sent the link to your inbox.
Most businesses can name their tech stack faster than their moats. That’s the wrong way round. List what makes you defensible first, then layer AI onto what’s already worth multiplying.
Checklist
- Walk your team’s screens for one morning and note every repetitive task you see.
- List your moats: customer history, key relationships, regulatory know-how, supplier network.
- Build the two-column sheet: “what’s unique about us” vs “where AI can multiply it”.
- Do it yourself: don’t delegate it, don’t outsource it.
Done looks like: a one-page sheet, on your desk by Friday, that pairs each unfair advantage with one place AI could multiply it.
Pitfall: starting with tools. The tool list is the easy half; it’s worthless without the moat list.
Don’t pilot AI to replace the human moment; pilot it to prepare the human for the human moment. The agent does the prep; the human does the relationship.
Checklist
- Pick one customer-facing process (account reviews, renewals, support escalations).
- Have an agent assemble the prep: history, open items, contract terms, last issue.
- Keep the human delivery exactly as it was, just better armed.
- Ask customers afterwards: did that feel more personal, or less?
Done looks like: one process where the customer leaves feeling more seen, and your team spent half the prep time.
Pitfall: automating the conversation itself. That’s how you turn a moat into a chatbot.
80% of SMEs report no productivity gain from AI, yet almost none have an AI problem. They have a data problem. An agent is only as smart as the context you can give it.
Checklist
- Pick one team and one system of record, not a company-wide migration.
- Pull the scattered history in: spreadsheets, old CRMs, the inbox archives.
- Agree the one place new customer information lands from now on.
- Connect your first agent to it and test against real customer questions.
Done looks like: one team whose full customer history is in one queryable place: your template for every other team.
Pitfall: treating this as IT hygiene. It’s not; it’s literally the asset your AI-native competitors can’t have.
You’re not training people to use AI; you’re training them to bring AI to the relationship. Drudge work moves to the agent so they do more relationship work, not less.
Checklist
- Block 60 minutes a week, whole team, hands-on, not a webinar.
- Everyone automates one of their own tasks each month.
- Name an agent owner per team, never just the founder.
- Share wins in the open: what got delegated, what time it freed, where it went.
Done looks like: by Christmas, every team has someone who can confidently run an agent end-to-end.
Pitfall: being the only person who can delegate to an agent. That’s a single point of failure dressed up as a competitive advantage.
Agency content sounds like everyone’s agency; generic AI posts sound like everyone’s AI. Both shrink the moat. Train AI on your view, your stories, your client work, then publish at scale.
Checklist
- Collect 10 things you’ve actually said to clients this year: calls, emails, pitches.
- Write down your three contrarian positions: what you believe that competitors don’t.
- Build a voice doc and give it to the AI with every drafting task.
- Publish one post a week. You edit; nothing goes out you wouldn’t say out loud.
Done looks like: a weekly cadence in your voice that you spend under an hour on, and that sounds like nobody else.
Pitfall: going invisible exactly when the market needs to hear your view, or worse, sounding like everyone else’s AI.
For every workflow on your inventory: build, buy or borrow. No shame in any of the three, only in indecision. The lane decision is a revenue decision.
Checklist
- Build: only if it touches a moat. Your data, your relationships, your edge.
- Buy: if it’s commodity. Invoicing, transcription, scheduling: off the shelf.
- Borrow: partner when you need speed without the headcount.
- Mark every inventory item B / P / H before your next board meeting; revisit quarterly.
Done looks like: every AI initiative has a named lane and an owner, and no pilot quietly becomes a custom build.
Pitfall: drift. The most expensive AI projects are the ones nobody decided to start.
Worksheet · The Inventory
| What’s unique about us | Where AI can multiply it |
|---|---|
| e.g. 12 years of service history with our top 40 accounts | Agent preps every account review from full history |
Decision matrix · Build / Buy / Borrow
| Lane | When | Watch out for |
|---|---|---|
| Build | It touches a moat: your data, your customer relationships, your edge. | Scope creep. Build the smallest thing that multiplies the moat. |
| Buy | It’s commodity work any business has. | Per-seat pricing for agent-shaped work; renegotiate. |
| Borrow | You need speed without permanent headcount. | Make sure the knowledge lands in your systems, not just theirs. |
The first 90 days
Want a hand with any of the six?
NetMonkeys builds relationship-native AI for East Midlands businesses, from Business Central agents to full data consolidation. 30 minutes, no pitch: bring your inventory sheet and we’ll tell you what we’d do first.
Book a call