Before you add AI, fix your enquiry flow

Every week I see a small business owner who’s been pitched an AI tool. A chatbot. A “smart” CRM. An automated quoting system. Sometimes all three.

The pitch always works because the problem is real: there’s too much admin, too many things slipping through, not enough hours. The tool promises to make all of that easier. And it might. But only if the thing it’s sitting on top of works.

Most of the time it doesn’t. So you spend £79 a month on something that automates the wrong thing faster.

What “enquiry flow” actually means

The path between “a stranger heard about you” and “they’ve become a paying customer”. For most small businesses, that path looks roughly like this:

• They find you (Google, recommendation, social, AI assistant)
• They land somewhere (your website, your Instagram, your phone)
• They make first contact (call, form, DM, walk-in)
• You respond (or don’t)
• They get clarity on what happens next (price, timing, availability)
• They decide (book, ask for more info, ghost)
• You follow up (or don’t)
• They convert (or don’t)

Each arrow in that flow is a place where things break. AI tools tend to be sold for the middle bits (responding, following up). The bigger leaks are usually at the start (they can’t find you) or the end (you never followed up).

Five questions to ask before you spend a penny on AI

1. Can someone enquire on their phone in 30 seconds?

Open your website on your own phone. Time how long it takes you, knowing the site, to start an enquiry. If it’s more than 30 seconds, your customers aren’t doing it either.

2. Do all enquiries land in one place?

WhatsApp on one phone, Instagram on another, email on the laptop, voicemail from the landline, website form to an old address. If your enquiries are scattered, no AI tool will fix the fact that you can’t see them all in one view.

3. Is there a clear next step after the enquiry?

“Thanks for your enquiry, I’ll get back to you” is not a next step. “Thanks for your enquiry — I’ll send a quote by Friday lunchtime, please ring me on 07769 028475 if you don’t hear from me” is. Customers want to know what’s about to happen.

4. Is there a follow-up if they don’t reply?

For most small businesses the honest answer is no. The quote went out, the customer didn’t reply, that’s the end of it. A single follow-up message three days later, sent by anyone (you, a VA, a templated email), recovers a meaningful share of these.

5. Are the repeat questions already answered on your site?

How much do you cost? Do you cover my area? When are you available? Can you do weekends? If the website doesn’t answer these, customers ring to ask. You spend time answering. A clear FAQ page is the cheapest piece of marketing collateral you’ll ever make.

The fix-first checklist

None of these need AI. Most can be done in a weekend. Doing them is what makes AI worth adding later.

1. Add (or fix) a website enquiry form. Send yourself a test enquiry. Make sure it lands.
2. Consolidate where enquiries land. One inbox, one shared dashboard, one diary — whatever you can manage.
3. Set up a missed-call SMS reply if you take phone enquiries.
4. Write a one-page “what happens next” on your website.
5. Make a simple follow-up template (a single line: “Just checking the quote landed OK — let me know if you’ve got any questions.”).
6. Build the FAQ page using the questions you actually answer most weeks.

That’s a Saturday morning. It’s also worth more than most of the AI tools you’ll be pitched.

When AI actually helps

Once the foundation is there, AI is genuinely useful in a few places:

• Volume. You’re now drowning in qualified enquiries. AI can help triage, classify, and prioritise.
• Repetition. The same question, the same answer, a hundred times a month. Chatbot territory.
• Drafting. Quote replies, follow-ups, review responses. You still hit send.
• Analysis. Which channels actually convert. Where time is going. Which weeks are quiet.

Each of these only pays off once the underlying flow works. Otherwise the AI is just a faster version of nothing.

Where the audit looks at this

Part 1 of the AI Opportunity Audit (external visibility) and Part 2 (internal opportunities) are both deliberately designed in this order: foundation first, AI second. If the foundation isn’t there, the action plan leans heavily on the foundation. If it is, the action plan looks at where AI actually adds something.

This is the article version of why the audit isn’t a sales pitch for AI. It’s a sales pitch for fixing what’s leaking.