5 Questions Every London Business Should Ask Before Buying AI Tools

After months of experimenting with various AI tools - from local models that produced endless repetitive text to development tools that created more problems than they solved - I’ve learned that the wrong AI tool can waste more time and money than doing things manually.

Most London SMBs I meet are asking “Which AI tool should we buy?” when they should be asking “Should we buy an AI tool at all?” Many are already spending thousands on failing AI experiments without understanding why.

Here are the five questions that could save you from expensive AI mistakes, based on real implementation experience.

Question 1: “Do I Actually Understand the Problem I’m Trying to Solve?”

Why This Matters: When I started experimenting with AI for product management work, I thought I wanted “better document generation.” What I actually needed was a way to overcome writer’s block and create structured first drafts.

The local LLM I tested failed because I hadn’t clearly defined what “success” looked like. Was it speed? Quality? Consistency? Without clear criteria, I couldn’t tell if any solution was genuinely useful. This experience taught me why building your own AI solution is rarely the answer for most SMBs.

Before buying any AI tool, write down:

• The specific task you want to improve
• How you currently handle this task
• What “better” would look like in measurable terms
• How you’ll know if the AI tool is actually helping

Red flag: If you can’t explain the problem in one sentence, you’re not ready for an AI solution.

Question 2: “Am I Trying to Replace Expertise I Don’t Have?”

The Hard Truth: During my app development experiment with Cursor, I discovered that AI coding tools don’t replace understanding - they amplify it. When I didn’t understand Swift programming, Cursor created compilation errors and duplicate files that made everything worse.

Many SMBs expect AI to work like a knowledgeable employee who can figure things out independently. In reality, AI tools work best when you understand the domain well enough to guide them effectively.

Ask yourself:

• Do I understand this area of work well enough to spot when AI gets it wrong?
• Can I provide meaningful feedback to improve AI outputs?
• Would I be able to fix problems if the AI tool breaks or produces poor results?

Reality check: If you’re hoping AI will teach you a new skill while solving your problem, you’re probably setting yourself up for frustration.

Question 3: “Have I Accounted for the Real Setup and Learning Costs?”

What Nobody Tells You: Setting up AI tools often takes much longer than advertised. Getting Xcode to work with Cursor took over an hour and killed my initial enthusiasm. Many teams give up during configuration rather than implementation.

Beyond setup time, effective AI use requires learning how to communicate with these tools. This isn’t intuitive - prompt engineering is a skill that takes practice.

Calculate the real costs:

• Initial setup and configuration time
• Learning curve for your team
• Integration with existing systems
• Hidden dependencies (like my £80 App Store surprise)
• Ongoing maintenance and updates

Warning sign: If the vendor says “just plug it in and it works,” they’re probably overselling.

Question 4: “Will This Tool Actually Work for My Specific Use Case?”

Lesson from Experience: Different AI models have different strengths and severe limitations. The local model I tested couldn’t handle general productivity tasks effectively, despite being designed for language work.

Many businesses choose AI tools based on impressive demos, then discover the tools don’t work for their actual use cases.

Test before you commit:

• Can you try the tool with your real data and workflows?
• Does it handle your specific industry terminology and requirements?
• How does it perform when you give it typical (not perfect) inputs?
• What happens when things go wrong?

Smart approach: Start with the free trial or cheapest option to test your actual use case before committing to expensive subscriptions.

Question 5: “Can I Start Smaller and Prove Value First?”

The Biggest Mistake: Most businesses try to automate entire workflows immediately, then get overwhelmed when nothing works as expected. This is particularly true when companies confuse RPA, AI, and simple process improvement.

The most successful AI implementation I completed was the childminder app - it focused on one specific task (converting photos and simple inputs into reports) rather than trying to solve every communication challenge.

Start smart:

• Pick the simplest possible task where AI could add value
• Perfect that before expanding to other areas
• Build confidence and understanding through small wins
• Scale up only after proving clear value

Success pattern: Every effective AI implementation I’ve seen started with solving one well-defined problem, not revolutionizing entire business operations.

The Reality Check

If your honest answers to these questions reveal gaps - unclear problems, missing expertise, underestimated complexity - you’re probably not ready for AI tools yet. That’s actually good news, because it means you can avoid costly mistakes.

The businesses succeeding with AI aren’t using the most sophisticated tools. They’re using appropriate tools effectively, with realistic expectations and proper preparation. And yes, your competitors might already be doing this - but that doesn’t mean you should rush into poor decisions.

Your Next Move

Before spending money on AI subscriptions, invest 30 minutes in honestly answering these five questions. Most AI failures can be traced back to skipping this basic preparation.

If you’re still unsure whether AI makes sense for your specific situation, book a free AI readiness assessment. We’ll work through these questions together and give you a realistic view of where AI could help your business - and where it probably won’t.

No sales pitch, no complex proposals. Just honest advice based on real implementation experience.

The goal isn’t to talk you into or out of AI tools - it’s to help you make decisions that actually improve your business rather than just adding to your monthly expenses.

QVXX helps London SMBs make smart technology decisions. We focus on what works for your specific situation, not what’s trending.