How Coaches Use AI to Qualify Discovery Call Bookings

Put three to five qualifying questions in front of your calendar, then let AI read the answers. It can sort each booking into good fit, maybe or not yet, write you a short brief before every call, and draft a kind reply that points the not-yet group to something useful. You define what "fit" means; the AI applies it consistently.

Qualification is there to protect both diaries, not to keep people out. Ask too much and good prospects abandon the form; filter too hard and you turn away clients who couldn't describe their problem neatly. So keep the AI's role to preparing and sorting, and glance at every "not yet" before a reply goes out. You may not need AI at all for simple rules: a routing form that sends anyone without a budget to a free resource is just a rule.

Follow me on Instagram@sagnikteaches

Decide what "qualified" means before writing a single question

Most coaches can describe their ideal client in a sentence and their worst discovery calls in a paragraph. Write both down, then turn them into four or five fit factors. An illustrative filled-in version for a career coach working with mid-career professionals:

Connect on LinkedInSagnik Bhattacharya

A business coach working with small retail owners might swap "stage" for "trading at least two years" and "problem" for "wants to grow without working more hours". The point is that each factor must be something a prospect can answer about themselves in a sentence.

Subscribe on YouTube@codingliquids

Questions that qualify without putting people off

A qualifying question should be easy to answer honestly and useful to you either way. Some common versions, and better ones:

Four questions, one of them open, is usually enough. Include "not sure yet" wherever it's honest, because a forced choice produces bad data.

A rubric the AI applies the same way every time

Turn the fit factors into points so the AI isn't improvising:

6 to 8 points: good fit. 3 to 5: maybe, which still gets a call. 0 to 2: not yet. The prompt:

You are sorting discovery call bookings for a career coach.
Apply this rubric exactly: [paste rubric and thresholds]
For the answers below, return:
- score per factor with a one-line reason quoting the answer
- total and category (good fit / maybe / not yet)
- a 3-sentence pre-call brief: their goal in their words, what to
  explore, one question to open with
Rules: score only from the answers given. Do not judge spelling,
grammar or writing style. Do not guess age, gender, background or
anything not stated. If an answer is blank, score it 0 and say so.

ANSWERS:
[form responses]

For one booking, the result might be (illustrative):

Problem: 2 - "stuck as a senior analyst for 4 years, want to move
into product management"
Timing: 2 - "ASAP, my review is in 6 weeks"
Investment: 1 - "not sure yet"
Time: 1 - "under 1 hour"
Total 6: GOOD FIT
Brief: Wants to move from senior analyst into product management
after four years without progression. Explore what's blocked
previous moves and how the upcoming review fits in. Open with:
"What happens in your review in six weeks?"
Note: limited time availability suggests low commitment.

The scoring is right and the brief is useful. The last line isn't: "suggests low commitment" is an inference the rubric didn't ask for, and it could bias you before the call. Someone with a demanding job and a new baby may be very committed and short of time. Delete notes like that, and add "no notes beyond the brief" to the prompt if they keep appearing.

The harder bookings are the vague ones, and they're where you learn whether the brief is any good. Here the open answer was "I just feel stuck and don't really know what I want next", timing was "not sure", investment "not sure yet" and time "1-2 hours". An illustrative result:

Problem: 1 - "feel stuck and don't really know what I want next"
Timing: 1 - "not sure"
Investment: 1 - "not sure yet"
Time: 2 - "1-2 hours"
Total 5: MAYBE
Brief: Seeking career clarity and a renewed sense of direction.
Explore their values and long-term vision. Open with:
"Where do you see yourself in five years?"

The score is fair, but the brief has drifted into coaching-speak. "Career clarity" and "renewed sense of direction" are the AI's words, not the prospect's, and the five-year question is the one every stuck person dreads. A better brief keeps their phrasing: "Feels stuck, doesn't know what's next. Explore what has changed recently that made them book now. Open with: 'What made this the week you booked a call?'" To get that consistently, add a line to the prompt: "Quote the prospect's own words for their goal; if they haven't stated a goal, say 'goal not stated' and suggest a question about why now."

Three ways to wire it up

1. Routing inside your booking tool. Calendly's routing forms let you ask questions first and send people to different event types or pages based on their answers. They're available on the Teams and Enterprise plans, not the free or Standard ones; Calendly's help page on routing forms shows the setup. Teams is listed at roughly $16 to $20 a seat a month depending on billing, so check the pricing page. Routing is rules-based, not AI: "Under $1,500" goes to your free resources page, everything else to the booking page. For many coaches that alone does most of the job.

2. Form, automation and an AI step. A form tool you already use sends each response to Zapier, where an AI by Zapier step applies the rubric and writes the brief, a Paths step routes by category, and the result goes to your email or CRM. Multi-step Zaps need a paid plan: Zapier Professional starts at $19.99 a month if you pay annually ($29.99 paying monthly) for 750 tasks. The AI step uses 1, 3 or 5 tasks per run depending on the model tier you choose, while Filter and Paths steps use none. At 40 bookings a month, you're well inside the limit. Automations fail quietly when a form field is renamed, so add the checks in stopping automations from breaking silently.

Laid out step by step, an illustrative version for a life coach using Google Forms looks like this:

1. Trigger: new response in the "Discovery call questions" form.
2. AI by Zapier: the rubric prompt above, with the four answer fields mapped in, asked to return three separate fields: total, category and brief. Separate fields matter, because the next step needs to read the category on its own.
3. Paths: Path A if category is "good fit" or "maybe": email the booking link. Path B if "not yet": create a Gmail draft of the not-yet reply, not a sent email, so you read it first.
4. Google Sheets: add a row with date, name, score, category and brief, which becomes your log for checking drift later.

The task sum for 40 responses a month: the trigger and the Paths step are free, the AI step costs 3 tasks on a mid tier, the email or draft is 1 and the sheet row is 1. That's 5 tasks a booking, or 200 a month, comfortably inside Professional's 750.

3. Manual, once a day. Form responses land in a spreadsheet; each morning you paste the new rows into ChatGPT or Claude with the prompt above. Five minutes a day, no new subscriptions, and the best way to test your rubric for a month before automating anything.

Whichever route you pick, run a test batch before real prospects meet it. Write ten fake bookings that cover the awkward cases, decide the category you'd give each one yourself, then compare. A filled-in sample of the sheet:

Two mismatches out of ten is normal on a first run. Keep adjusting until your categories and the AI's agree on all ten, then go live.

The pre-call brief that changes the first five minutes

The brief is where qualification pays for itself even with good-fit bookings. Walking into a call knowing the person's goal in their own words, and with a question ready, changes the first five minutes. Keep it to what they told you. If you want background on a business-owner prospect, their company website and professional profile are fair game; their personal social media isn't. The research side is covered in using AI for client research before discovery calls.

For a business coach, a typical case: a toy shop owner books a call. Answers: trading six years, "want to stop working every Saturday without losing sales", ready to start next month, in the investment range. The brief tells the coach to explore who else could run Saturdays and what the shop's best-selling hours are, and to open with "What would you do with your Saturdays back?" That's a better first question than "So, tell me about your business."

A kind "not yet" reply

The not-yet group are future clients, referrers or reviewers, and a cold rejection wastes all three. Compare the default with a reply worth sending:

Automated default: "Unfortunately, based on your answers, you are not currently a good fit for our coaching programmes."

Better, drafted by the AI from your template and checked by you: "Thanks for telling me about your plans to move into your first graduate role. My one-to-one programmes are built for people further into their careers, so I don't think a call would be the best use of your time yet. What I'd suggest instead is my free guide to first-job interviews [link], and the monthly group Q&A, which is open to anyone. If things change, you're always welcome to book again."

Read every not-yet reply before it goes, at least for the first two months. It takes seconds and it's where a rubric's blind spots show up.

One coach's month, before and after

Here's an illustrative executive coach who gets about 30 discovery call bookings a month from a website booking link. Before qualifying, roughly a third of the calls were with people who wanted something she doesn't offer, weren't ready, or were shopping for free advice. At 30 minutes plus 15 minutes of preparation and notes each, those ten calls cost about seven and a half hours a month.

After adding four questions and the manual daily sort, bookings dropped slightly, to 26, because some mismatched prospects went to the free resources instead. Calls with poor-fit prospects fell to about three. Preparation time per call dropped from 15 minutes to 5, because the brief was waiting. Net result: around nine hours a month back, and calls that start in the right place.

Track four numbers monthly: bookings, show-up rate, calls that turned into proposals, and how many "maybe" calls became clients. If maybes convert as well as good fits, your rubric is too strict. Pair it with reminders from reducing no-shows with AI reminders if show-up rate is the weak point.

Rubric problems that surface early

• A rule that filters out good clients. A realistic example: one coach routed every "not sure yet" on investment to the free guide. Two of the next month's best-fit prospects had answered "not sure yet" simply because they hadn't seen prices. Moving "not sure" to the maybe group fixed it.
• Spam and bots. Public booking links attract junk. Add a required open question and let the AI flag nonsense answers for deletion rather than scoring them.
• Disclosures that need a person, not a score. Open questions sometimes get answers no rubric should touch. A wellbeing coach might read "I've been signed off work with anxiety and my doctor suggested coaching" or something more serious. Add a rule: "If an answer mentions a health condition, a crisis or harm to anyone, return category REVIEW, no score and no brief." Then reply to those yourself, and if the need sits outside coaching, say so kindly and point to appropriate professional help. The AI step should also be told not to copy health details into the brief or the sheet.
• Referrals scored like strangers. A prospect sent by a past client often answers vaguely because the client has already told them what you do, so they score 3 and land in "not yet". Add "How did you hear about me?" with a "Recommended by someone" option, and a rule that recommended bookings always get a call whatever the score.
• People gaming the form. Some prospects pick whatever answer gets a call. That's fine: the brief still tells you where to probe.
• Scoring drift. Every few weeks, score five bookings yourself and compare with the AI's. If they differ, tighten the rubric wording rather than overriding case by case.
• Losing warmth. If the process starts to feel like a funnel to the people going through it, keeping the personal touch while using AI has ideas for putting it back. And if you add a chat widget to your booking page, tell visitors they're talking to an AI; it's a legal requirement for people in the EU.

After the call itself, the notes and follow-up are a separate workflow: AI session notes and follow-ups for coaches picks up from there.

Further reads

• Are AI Tools Built for Coaches Worth Paying For Over ChatGPT? — Whether coach-specific AI tools beat a general assistant.
• Can AI Qualify Wedding Leads by Date, Budget and Guest Count? — The same qualifying logic applied in a very different business.
• How to Reply to Every Enquiry in Under Five Minutes With AI — Fast first replies for enquiries that don't come through the form.
• Should Coaches Build an AI Version of Their Method for Clients? — What to offer the not-yet group beyond a newsletter.
• What Is Lead Nurturing and How Can AI Automate It? — Turn the not-yet group into future clients over months.
• How to Build a Quiz Funnel With AI to Capture Leads — A quiz-style alternative to a plain qualifying form.
• AI Appointment Scheduling for Small Businesses: Tools and Setup — Booking engine first, rules second, AI at the front door last. Tools, prices, a six-stage setup and an architect practice's bookings worked through.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Calendly help pages (routing forms, plan availability) and pricing page; Zapier pricing and AI by Zapier task usage, checked September 2026.