1. The Problem: Old Dogs, New Tech
Across the UK, boardrooms are buzzing with uneasy chatter: ‘Should we be using AI?’ ‘Is it safe?’, ‘What if we get it wrong?’ and occasionally, ‘Has anyone tried turning it off and on again?’
Many established businesses, especially those in traditional sectors like law, accountancy, manufacturing or professional services, are facing the same dilemma: they know AI is coming (or already here), but they’ve no idea how to integrate it meaningfully, safely, or profitably. And the tech is so very fast-moving, it’s virtually impossible to keep up to speed with new developments.
So how do you build a strategy for bringing generative AI into a decades-old business without causing chaos or missing the boat?
2. What We Used to Do: The Wait-and-See Approach
Historically, many firms dealt with technological change by ignoring it for as long as humanly possible – until clients, competitors or staff forced the issue. Fax machines replaced post. Email replaced fax. Then, a fancy intranet sat gathering digital dust.
There’s been a temptation to treat AI the same way. Some firms have:
- Bought a few AI licenses ‘just in case’;
- Experimented in isolated corners (‘Sandra in accounts is trialling it’);
- Or worse, banned it altogether out of fear of mistakes or data breaches.
But those half-measures are no longer enough. AI is reshaping workflows, client expectations and business models. Sitting it out is not a strategy. It’s a liability.
3. How to Formulate an AI Strategy: A Practical (Human-Led) Approach
Here’s a clear roadmap for business leaders ready to take AI seriously, without overreaching or undercooking it.
A. Understand the ‘Why’ First
Don’t start with tools. Start with problems. Ask:
- Where are our pain points? (e.g. report / document creation, customer support, compliance, marketing)
- What do clients expect from us that we’re struggling to deliver?
- What’s repetitive, predictable, or pattern-based in our work?
If you can answer those questions, you’ll find the right use cases for AI.
B. Appoint an AI Lead
Pick someone practical. Someone who understands your business processes and has curiosity, not fear, about technology. They’ll become your AI translator, the main experimenter and watchdog.
Bonus points if they already talk to both IT and operations.
C. Educate Internally Before Automating Externally
Bring your staff along. Offer demos. Set up play sessions. Show how AI (like ChatGPT or your own in-house tools) can help – not replace – them. Hold monthly meetings with everyone bringing forward a tool they’ve tried out (even if it doesn’t work as they’d like).
Human-in-the-loop isn’t just a best practice. It’s your safety net.
D. Test. Measure. Iterate.
Start small:
- Create a pilot project (e.g. using AI to draft marketing emails or produce first-draft file notes).
- Measure success against real KPIs (e.g. time saved, errors reduced).
- Get feedback from users.
- Tweak and improve.
Avoid the temptation to overhaul everything at once.
E. Build a Governance Layer
Yes, it sounds dry. But you need:
- Clear policies on AI use (especially around data protection and output review)
- Version control and record keeping for AI-generated work
- Defined responsibilities for review and sign-off
A bad AI strategy isn’t one that fails, it’s one that works too well and then causes a mess.
F. Make It Part of Business Planning, Not an IT Bolt-On
Your AI strategy should feed into:
- Hiring plans (what skills do we need?)
- Client engagement (what will impress or scare them?)
- Risk management
- Pricing and billing models (especially for professions moving away from time-based charging)
G. Train, Then Train Again
AI evolves fast. Your people need regular top-ups, not one 90-minute webinar from six months ago.
4. The Don’ts: What to Avoid
- Don’t just buy a tool and hope it works. AI isn’t a plug-and-play magic wand.
- Don’t leave it to IT alone. This is a cross-functional journey.
- Don’t copy what big companies like Google are doing. You’re not Google.
- Don’t forget compliance. AI tools can hallucinate, leak data or mislead if unchecked.
- Don’t ignore staff fears. This is a cultural shift as much as a technological one.
5. Timelines: Realistic, Not Revolutionary
Think in quarters, not weeks.
- Q1: Identify key use cases, select AI lead, run small pilot
- Q2: Evaluate pilot, refine approach, train staff
- Q3: Expand AI usage into new teams, develop policy frameworks
- Q4: Review ROI, embed into annual business planning
Repeat. Evolve. Stay nimble.
And reward staff for staying on the program – reduce their work hours, give them pay rises or bonuses to encourage them. Definitely do not give them pizza, no one falls for that old trick! And don’t give your staff Friday afternoons off and expect them to work extra hours Monday to Thursday as one local accounts firm has just done – it’s a sure fire way to lose the staff’s respect).
6. Alternatives (Hint: There Aren’t Many Good Ones)
- Do nothing – and risk falling behind in service, efficiency and client expectations.
- Outsource everything – which may cost more, lose control and offer little learning.
- Over-invest – and buy tech that staff don’t use or clients don’t value.
A balanced, iterative, people-first AI strategy is by far the safest – and most rewarding-course.
7. Consequences of Action… and Inaction
| Action Taken | Consequences |
|---|---|
| Thoughtful AI strategy | More efficiency, better client service, staff who feel empowered-not threatened |
| Ignore AI | Gradual decline in relevance, productivity, and eventually… market share |
| Panic-buy AI | Wasted budgets, ethical problems, staff confusion |
8. The Uncertainties (and Why They Shouldn’t Paralyse You)
- AI will change. New models, new risks, new laws.
- But the human skills of judgement, review and relationship-building won’t go away.
- The best strategy is to stay alert, flexible and in charge.
You don’t need all the answers. You just need to start asking better questions.
Conclusion: Make a Start – But Make It Smart
Integrating AI into an old-school business is less about coding and more about curiosity. It’s about taking practical steps that fit your business, not someone else’s.
If you want to understand how AI could work in your business (without a jargon overload), just give me a call. I don’t know all the answers because this technology is so very big and new, but I’d be glad to join you on the journey.
📞 Paul Brindley | 💻 paul@vaisolutions.co.uk
Let’s talk. There’s a strategy that’s right for you. Let’s build it together.
