Where's the Money Hiding?

An SME’s AI Opportunity Audit

A Bisque whitepaper for SME owners and decision-makers — includes a printable scorecard

Executive summary

Most businesses approach AI backwards. They start with the technology — “should we get an AI thing?” — and go looking for a problem to point it at. That’s how you end up paying for a tool nobody uses.

The right order is the opposite. Find the most expensive, most repetitive, most error-prone work in your business first. Those are your AI opportunities, ranked by what they cost you. The technology is the easy part once you know what it’s for.

This paper is a practical self-audit. It walks through the four places money usually hides in an SME, shows you how to find them in your own business, and ends with a one-page scorecard you can print and fill in. Run it honestly and you’ll finish with a prioritised list of where AI would pay back fastest — whether you ever work with us or not.

Part 1 — The four cost centres AI attacks

Almost every AI opportunity in an SME falls into one of four buckets. Learn to see them and you’ll spot opportunities everywhere.

1. Wasted time. Repetitive tasks that fill your team’s day and produce nothing on their own: quoting, scheduling, data entry, chasing paperwork, compiling reports, answering the same questions. Every hour here is an hour not spent on work that grows the business. The cost is invisible because it’s spread across everyone — which is exactly why nobody adds it up.

2. Errors and rework. Mistakes that are expensive precisely because they’re easy to miss until it’s too late: the defect caught after production, the quote with the wrong figure, the invoice that doesn’t match the order. Rework costs twice — once to do it, once to fix it — plus the customer goodwill it burns.

3. Waste. Money leaking out through decisions made without good information: paying more than necessary because checking alternatives takes too long, over-ordering, under-using capacity, spend that isn’t tracked. Small leaks, constant flow.

4. Slow response. The lead that goes cold because nobody replied for hours. The quote that takes three days because someone has to look everything up by hand. In both cases the customer has moved on — often to a competitor who was simply faster. This is usually the most expensive bucket of all, because it costs you revenue you’d otherwise have won.

Part 2 — How to find them in your business

You don’t need consultants to do the first pass. You need a week and some honesty.

Shadow your own week. For five working days, have yourself and your key people note every task that fits any of these:

• It’s repetitive — you do it the same way, many times.
• It lives in one person’s head — when they’re off, it stops or slows.
• It involves chasing — paperwork, approvals, information, people.
• It’s a recurring error — the same kind of mistake keeps happening.
• It’s slow — customers wait, and waiting costs you.

Don’t filter as you go. Just capture. By Friday you’ll have a list that surprises you, because most of this work is so routine it’s become invisible.

Then attach numbers. For each item, estimate two things: how many hours a week it eats (across everyone involved), and what a single failure costs (a lost lead, a scrapped job, an overpayment). You don’t need precision. You need order-of-magnitude, because that’s enough to rank.

Part 3 — Worked examples

To show how this looks across different businesses:

A trades or service business. Enquiries arrive while everyone’s on a job. Half don’t get a callback the same day; some never do. Slow response — and it’s the biggest leak in the business, because every missed enquiry is lost revenue they already paid marketing to generate. First AI opportunity: instant, around-the-clock enquiry handling and follow-up.

A professional services firm. Quoting and proposals take a partner half a day each because everything’s looked up and assembled by hand. Wasted time, at the most expensive hourly rate in the building. First AI opportunity: quote and document automation, freeing senior people for billable work.

A small manufacturer. Defects are caught at final inspection — after the material’s used and the delivery’s already late. Errors and rework. First AI opportunity: catching the problem earlier in the process, before the cost is locked in.

A retailer or hospitality business. Reviews go unanswered, bookings need manual handling, and nobody has time to look at which products actually make money. A mix of slow response and wasted time. First AI opportunity: automating front-desk admin and surfacing the data already sitting in the till.

Different sectors, same four buckets. Yours will fit too.

Part 4 — The AI Opportunity Scorecard

Print this. Fill in a row for each task from your week of shadowing. Score each column 1–5.

How to read it:

• Total of 11–15: Top-priority AI opportunity. High cost, and repetitive/rules-based enough that AI handles it well. Start here.
• Total of 7–10: Worth doing, second wave. Solid return once the top items are sorted.
• Total of 3–6: Leave it. Either it doesn’t cost enough or it needs human judgement AI can’t replace. Automating it isn’t worth the effort.

The third column matters as much as the first two: a task can be costly and a bad fit for AI if it genuinely needs human judgement every time. The sweet spot is costly AND repetitive. That’s where AI pays back fastest.

Part 5 — What the scorecard usually reveals

When SMEs run this honestly, two things tend to surface at the top:

1. Slow response to leads — almost always scores high, because it’s frequent, costly, and highly automatable. It’s also the cheapest win, because you’ve already paid to generate those leads. (Our next paper is entirely about this one.)
2. A quoting or admin bottleneck — usually living in one person’s head, eating senior time, and perfectly suited to automation.

If those two come out on top for you, you’re in good company — they’re the most common high-return AI opportunities we see across sectors.

A note on doing this properly

You can run this audit yourself, and you should. But the gap between a list of opportunities and a working system is where most AI projects quietly die — usually because someone automated a broken process, or bought a tool before understanding the problem. That’s the part we do: turn the top of your scorecard into something that actually runs, with a number attached, that you own outright.

What to do with this

Run the audit. Even if you do nothing else, you’ll know — with numbers — where your money is hiding. Most owners have never had that list, and having it changes every conversation about spend.

If you’d like a second pair of eyes on your scorecard, that’s exactly what our free 30-minute audit is for: a straight read of where AI would pay back fastest in your business. No obligation. No sales pitch. Just the maths.

Book a free audit at wearebisque.co.uk — or reply to the email this came with, with your scorecard attached.

Bisque — AI automation and enquiry generation for UK businesses. We don’t sell activity. We build systems that produce results.