All plays › Legal, accounting and finance

Legal, accounting and financeHannah AI (LedgerLeads, CleanBooksAI)

Booked meetings from new leads up from 18 to 37 at a UK accountancy

Picked and checked by Olga Pearson, host of The Practical AI Show · Published

What was the result?

vendor-reported

Before AI, The Tax Guys handled about 102 leads a year and got 18 meetings. Since April 2025 an AI lead agent that replies in under 30 seconds, 24/7, handled 128 leads, got 74 replies and booked 37 meetings, and 16 became paying clients, a 12% conversion rate. The largest new engagement was worth £12,850.

The Tax Guys, a UK accountancy firm, had steady enquiries but slow, inconsistent follow-up and nobody to answer outside office hours. They put an AI lead agent called Hannah on first contact: it replies to every new enquiry in under 30 seconds, day or night, asks qualifying questions, gathers information and books meetings straight into the diary. The human team still runs the meetings and does the work.

Who did this?

How do I try it?

  1. Editorial adaptation: pull your last year of enquiries and count how many became meetings and clients.
  2. Set up an AI lead agent that answers every new web or email enquiry within a minute, including nights and weekends.
  3. Give it 3 to 5 qualifying questions and a link to your real diary so it can book a meeting.
  4. Keep people on the meeting and the work; the AI only does first contact and booking.
  5. Compare meetings and new clients after a few months against your old numbers.
Copy-paste prompt
Editorial adaptation, not the firm's setup. You reply to new enquiries for [firm], an accountancy practice. Reply within a minute, warm and short, and say you are an AI assistant. Ask: what kind of help they need, whether they run a business or need personal tax help, rough size, and their deadline. If they fit [criteria], offer a meeting using this booking link: [link]. Never quote fees or give tax advice; a person will cover that in the meeting.

Make it run every week

A play works once. Turn it into a routine in three small steps:

  1. Run it once and check the result yourself.
  2. Give your AI a schedule. Tell it when to run this play (every week, or every month) and to message you with what it found before you act.
  3. Keep a log. One line per run: what was checked, anything flagged, what you did.

See how owners turn plays into systems →

What should I watch out for?

Vendor-reported: the piece is written by the vendor (CleanBooksAI) about its own tool. The before figure is per year and the after figures are since April 2025, so the periods are not an exact match. No person at the firm is quoted. Safer version: have a person review AI conversations weekly. Build it with Page: an enquiry page that books the meeting in one step.

Did it work for you?

PageMotor keeps this site free.