AI Consultant vs DIY Tools — What UK SMEs Actually Need

The pitch for DIY AI is compelling. You've got ChatGPT, which can draft emails, summarise documents, and answer complex questions for free. You've got Microsoft Copilot baked into the tools your team already uses. You've got Zapier, Make, and a dozen other no-code platforms ready to automate workflows without a line of code.

So the obvious question is: why would any SME spend £5,000–£30,000 on an AI consultant?

The honest answer is that for many businesses, right now, you probably shouldn't. But that depends entirely on where you are, what problem you're solving, and what you're hoping to achieve. The decision isn't "consultant versus tools" — it's about which approach actually delivers a return given your specific situation.

I've seen both choices work well and both choices fail badly. Here's what I've learned about when each is the right call.

What DIY AI Tools Are Actually Good At

Off-the-shelf AI tools have become genuinely powerful, and businesses that use them well are getting real results. The key word is "well" — these tools reward users who know what they're asking for and have clear enough processes to point them at.

Here's where DIY AI consistently delivers:

• Repetitive text tasks — drafting routine emails, summarising meeting notes, generating first drafts of documents, rewriting copy in a consistent tone
• Research and synthesis — pulling together information from multiple sources, summarising long documents, producing structured briefings
• Single-step automation — if X happens, do Y. Moving data between well-documented systems, triggering notifications, formatting outputs
• Augmenting individual productivity — giving a skilled team member a tool that makes them 20–30% faster at their specific tasks

These are high-value activities that most SMEs can implement with a few days of internal effort and a modest subscription cost. If this is all you need, you don't need a consultant. A clear use case, a month of iteration, and the right tool subscription is more useful than an expensive engagement.

Where DIY AI Breaks Down

The failure mode I see most often isn't companies buying the wrong tools — it's companies buying the right tools and applying them to problems that aren't actually well-defined, or expecting them to fix things that aren't fixable with a subscription.

The patterns that reliably fail:

• Undefined processes. AI automates processes. If your process is "we handle it case by case," AI will faithfully automate chaos. The documentation problem has to be solved first.
• Data that isn't accessible. You can't get AI to analyse your customer churn if you can't produce a clean export of your customer records. Data access is a prerequisite, not a byproduct.
• Cross-system complexity. Most SMEs run 5–10 software tools that don't talk to each other cleanly. Connecting them requires either custom integration work or significant process compromise.
• "AI strategy." Tools don't give you a strategy. They execute tasks. If the question is "what should we be using AI for and in what order," that's a strategy question — and ChatGPT isn't the right tool to answer it.

The most common failure pattern isn't bad tools — it's good tools applied to the wrong problem, or to an undocumented problem, or to a problem that the business hasn't actually agreed they want to solve.

The Decision Matrix

Here's how I'd frame the choice. Neither option is categorically right — it depends on the specifics of your situation.

The Cost Comparison (Honestly)

The numbers look very different depending on how you count them.

Factor	DIY Tools	AI Consultant
Upfront cost	£50–£500/mo subscriptions	£5K–£30K engagement
Hidden cost	Staff time to configure, iterate, maintain	Internal effort to implement recommendations
Time to value	Weeks (if scope is clear)	4–12 weeks (includes discovery)
Best outcome	Faster individual tasks	Structural operational improvement
Risk of failure	Wasted subscription spend + staff time	Wasted engagement if scope was wrong
Scales with…	Individual clarity and discipline	Quality of brief and internal follow-through

The right framing is total cost of outcome, not cost of the tool or the engagement. A £200/month tool that absorbs 10 hours of a senior person's time every month — because it never quite does the right thing — is more expensive than a focused consultant engagement that fixes the underlying problem in six weeks.

The Sequencing Most Businesses Get Wrong

The mistake isn't usually choosing DIY over consulting or vice versa. It's getting the sequence wrong.

The pattern that works:

1. Diagnose first. Before spending anything, work out what problem you're actually solving and whether AI is the right tool for it. This doesn't need to be expensive — it can be a structured internal workshop or a focused discovery conversation with someone who can help you think it through.
2. Start narrow. Pick one process, not five. Get something working and measurable. Build internal confidence before expanding scope.
3. Use tools where tools work. There's genuine value in off-the-shelf AI for well-scoped tasks. Don't bring in external help for things you can do yourself in a few days.
4. Bring in expertise when you hit the ceiling. If you've done steps 2 and 3 and found that the real problem requires cross-system integration, process redesign, or a strategic decision about priorities — that's when outside expertise earns its cost.

What a Good Consultant Engagement Actually Delivers

Since the question is fundamentally about whether the cost justifies the outcome, it's worth being specific about what good consulting delivers that tools alone don't.

The things that are hard to buy with a subscription:

• Prioritisation. Which problems to solve first, based on ROI potential and implementation effort — not vendor marketing claims.
• Process clarity. Mapping what actually happens in your business versus what people think happens — which is usually different, and the gap is where AI projects die.
• Integration architecture. How to connect your existing systems in a way that's maintainable, not just technically possible.
• Change management. Getting your team to actually use new processes, not just having new processes on paper.
• Independent judgment. No tool vendor will tell you their product isn't the right solution for your problem. An independent consultant will.

If these aren't the things your business needs right now, you probably don't need a consultant. If they are — tools won't get you there.

The Honest Conclusion

For most UK SMEs at an early stage of AI adoption, the right move is to start with tools. Pick one or two well-scoped use cases, buy the relevant subscription, spend a month iterating, and measure what actually improves. This builds internal knowledge and gives you the evidence to make smarter decisions about what comes next.

Consulting makes sense when you've got a problem that tools alone can't solve — usually because the underlying issue is strategic, structural, or requires integration work that goes beyond what a subscription can handle.

The question isn't which one is better. The question is which one is right for where your business is today — and being honest about that is the most valuable thing you can do before spending any money.

Related reading: If you're trying to understand whether your business is ready to make AI work at all — regardless of which route you take — see 5 Signs Your SME Is Ready for AI Transformation. For the broader context on stages of readiness, AI Readiness for UK SMEs: Where to Start in 2026 walks through the Foundation → Growth → Scale framework.