Running Client Discovery Calls

Cornerstone Guide · 22 minutes · Introduces the Discovery Conversation Map

Summary

A discovery call is not a sales meeting, and it is not a questionnaire. It is the first serious argument an agency makes to a prospective client — and the client will assume the rest of the engagement feels like whatever this call felt like. This guide sets out the LaunchAnAEO methodology for running an AI Visibility discovery call and introduces The Discovery Conversation Map — a five-stage model that moves a first call from a stranger's business context to a considered commercial recommendation, without ever becoming a pitch.

Introduction

The discovery call does not exist to deliver a recommendation. It exists to earn the right to one.

Discovery calls fail in two directions.

Some collapse into a sales conversation before the agency has understood the business: within twenty minutes the founder is describing packages, pricing and timelines to a buyer who has not yet said, in their own words, what they are trying to change. Others collapse into a questionnaire: the agency arrives with a form, works through it politely, and hangs up with a full document and no shared understanding. Both calls end the same way — the buyer says they will "think about it", and the follow-up never quite finds a foothold.

The middle of a discovery call is where trust is actually built, and the middle is what agencies most often skip. The Discovery Conversation Map is a conversation model, not a script. Its job is to make the shape of the middle visible so the agency can hold it under pressure — including the pressure that comes from a buyer who wants a price on minute nine.

This guide is written for entrepreneurs, freelancers, consultants and early-stage agencies running their own AI Visibility discovery calls. By the end you will have the vocabulary of the Map, a stage-by-stage method for holding a first conversation, a way of handling the two questions that most often derail it, and a checklist you can run against your next call.

Why Discovery Is the Product Before the Product

Discovery is not the free part of the engagement. It is the first piece of the engagement. What the buyer experiences in that call is what they will assume the rest of the work feels like: the pace of thinking, the willingness to sit with a difficult question, the restraint about jumping to answers. A well-run discovery is the strongest signal a new agency can send about how it works, because it is the only signal the buyer has yet received directly.

New agencies prepare for a first call by rehearsing what they will say. Experienced ones prepare by rehearsing what they will ask, and in what order. The difference shows up in the first ten minutes. A call that opens with the agency describing its methodology puts the buyer in the role of an audience. A call that opens with a careful question puts the buyer in the role of a protagonist, which is the role they need to be in for the rest of the conversation to work.

The Map is the shape of that better conversation. It is designed for AI Visibility engagements specifically — where the buyer is often trying to close a risk they cannot yet name, in a category that did not exist eighteen months ago. A generic sales-discovery template can be adapted to fit that terrain, but the Map is built for it.

The Framework — The Discovery Conversation Map

The Map is the model LaunchAnAEO uses to structure a first AI Visibility call. It is deliberately small: five stages, in order, each of which earns the right to the next.

The Map is not a script. Two agencies running it will ask different questions, land on different diagnoses, and recommend different packages — and both can be right. What the Map will not tolerate is a call that skips a stage. Jumping from Context straight to Recommendation reads as a pitch. Skipping Ambition reads as a technical consultation. Skipping Diagnosis reads as an off-the-shelf sale. Each stage does work the next stage depends on; the sequence is the point.

Each of the five stages is described below using the same shape — purpose, questions, what the stage earns, what "done" sounds like, and its failure mode. That repeated shape is deliberate: readers should absorb the rhythm of the Map as they read it, because the rhythm is part of the methodology.

Stage 1 — Context

Purpose. Establish, in the buyer's own language, what the business does, who it sells to, and what has changed recently enough to make this conversation worth having now.

Questions. What does the business actually sell, described the way a good customer would describe it? Who tends to buy, and what were they trying to do when they came looking? What has changed in the last six months that made AI Visibility land on the agenda this quarter?

What this stage earns. The right to ask anything else. Without an accurate picture of the business in the buyer's own words, every later question sounds generic.

What "done" sounds like. The agency summarises the business back to the buyer in three sentences and the buyer nods without correction. That summary is the artefact of the stage; when it lands, the call moves on.

Failure mode. Thoroughness. New agencies over-invest in Context because it feels safe — it is the stage furthest from a commercial decision — and end up spending thirty minutes on a company overview the buyer could have written themselves.

Stage 2 — Constraints

Purpose. Surface what is true about the business — its resources, its history, its internal politics, its previous investments — that any recommendation will have to respect. This is not a negotiation about budget. It is a clarification of the terrain.

Questions. What has the business already tried in this space, and what did it learn? Who inside the organisation would need to be involved for a change to stick? What is the internal appetite for content, publishing or technical change over the next six months? Where has the buyer been burned before, and by whom?

What this stage earns. The right to make a recommendation the business can actually act on. A recommendation made without Constraints is a recommendation made against a business that does not exist.

What "done" sounds like. The agency describes, out loud, two or three specific realities the recommendation will have to fit inside, and the buyer confirms them.

Failure mode. Skipping straight from Context to Ambition. Recommendations made without Constraints tend to be technically correct and organisationally impossible; the buyer receives them politely and never acts on them.

Stage 3 — Ambition

Purpose. Draw out — in the buyer's own words — what they are actually trying to change, and what a good outcome would look like six or twelve months from now.

Questions. If this went unusually well over the next year, what would be different about the business? What conversation would the buyer be having with their board, their partners or their team that they cannot have today? What is the smallest version of that change that would still count as a real win?

What this stage earns. The right to diagnose. Without a stated ambition, any diagnosis is a solution in search of a problem.

What "done" sounds like. The buyer, in their own words, has described a future state, and the agency has reflected it back in a single sentence the buyer accepts without softening.

Failure mode. Generosity. New agencies, keen to sound ambitious, inflate the buyer's stated ambition into something grander — "so really what you are looking for is category leadership" — and lose the buyer's trust in a single sentence. The buyer said what they meant. Record it accurately. A modest ambition usually points at a smaller, cleaner first engagement, which is often the right first sale.

Prospect says: "We just want to make sure we are not being overlooked by whatever these AI answer engines are recommending." Think about: the ambition is defensive, not offensive. The buyer is trying to close a risk, not open a new channel. Notice: a defensive ambition asks for a diagnostic first, not a growth programme. Match the shape of the recommendation to the shape of the ambition.

Stage 4 — Diagnosis

Purpose. Name, out loud, the specific reason the buyer's ambition is not already true — using what has been learned in the previous three stages.

Questions. Fewer questions here; more listening back. Given what has been said, what is actually in the way? Which of the constraints is doing the most damage to the ambition? What would have to become true for the buyer's stated future state to be reachable?

What this stage earns. The right to recommend. A diagnosis the buyer confirms in their own words is the licence the Recommendation stage runs on.

What "done" sounds like. The agency states a specific reason grounded in this business's own context, and the buyer confirms — in words, not just a nod — that it matches how they see it.

Failure mode. Generality. Diagnoses that could apply to any business in the category are not diagnoses; they are marketing copy.

To make the difference visible, consider a single fictional prospect. Ledgerly is a mid-market accounting SaaS selling to finance operations leads inside 200- to 2,000-person companies. It has strong reviews on the usual comparison sites, a competent SEO agency, and a founder who has noticed that when procurement teams ask ChatGPT or Perplexity for shortlists, Ledgerly is rarely mentioned by name.

Weak diagnosis. "You need better AI Visibility." — True, but it could be pasted onto any of a thousand businesses. It names the category, not the cause. The buyer politely agrees and forgets it by the end of the day.

Better diagnosis. "You have strong reviews, but AI systems are still describing you using outdated third-party sources, which is why procurement shortlists miss you." — Specific to what the buyer said, and closer to a cause. Still slightly generic in its phrasing.

Excellent diagnosis. "The finance-ops buyers you are trying to reach are increasingly asking AI systems for shortlists before they ever visit G2 or your website — and the sources those systems are drawing on for your category are three years out of date, which is why Ledgerly, specifically, keeps missing shortlists it should be on." — Same business, same problem, sharper picture. It names the buyer, the channel, the mechanism and the consequence, in the buyer's own terms.

Stage 5 — Recommendation

Purpose. Translate the shared diagnosis into a specific piece of work the agency is proposing to do, described at a level the buyer can make a decision about.

Questions. Almost none. Recommendation is where the agency stops asking and starts naming. The one useful question comes at the end: does that shape sound right to you?

What this stage earns. A decision, or a clean reason it is not yet the time to decide. Either is a useful outcome.

What "done" sounds like. The buyer has heard a specific shape of work, has asked at least one clarifying question, and has agreed — or declined — a next step. If the answer is genuinely "we need to think about it", something further up the stack has been skipped; the honest thing is to name it and offer to revisit rather than to send a proposal into silence.

Failure mode. Over-delivery. Agencies keen to prove value describe three options, four workstreams and a phased plan — and hand the buyer the work of choosing. A single, considered recommendation trusts the buyer to say yes, no, or "tell me more about the second half". A menu is not a recommendation.

A recommendation, at this stage, is a verbal, single-sentence description of the engagement, tied directly back to the diagnosis and the ambition. "Given what you have described, what LaunchAnAEO recommends is a ninety-day representation engagement that rewrites how AI systems describe you, followed by a monthly review that keeps it accurate as the category evolves — I can send a written version tomorrow if that shape sounds right." Short. Specific. Testable in the room.

The Map at a Glance

Before moving from stage-by-stage into practice, the five stages are worth holding in view together. Read down the table and the discipline becomes obvious: each stage earns the right to the next, and skipping one weakens every stage that follows.

Stage	Purpose	Signature question	Signal you can move on	Common failure
1. Context	Understand the business in the buyer's own language.	What does the business actually sell, and who tends to buy it?	You can summarise it back in three sentences and the buyer nods.	Turning it into a company overview interview.
2. Constraints	Surface the realities any recommendation will have to fit inside.	What have you already tried in this space, and what did it teach you?	You can name two or three specific realities out loud and the buyer confirms them.	Skipping straight from Context to Ambition.
3. Ambition	Draw out the change the buyer is actually trying to make.	If this went unusually well over the next year, what would be different?	The buyer accepts a one-sentence reflection of their own ambition, unsoftened.	Inflating the buyer's stated ambition into your own.
4. Diagnosis	Name the specific gap between where the business is and where the buyer wants it.	Given what you have described, here is what I think is actually in the way — does that match how you see it?	The buyer confirms the diagnosis in words, not just a nod.	Collapsing diagnosis and recommendation into a single breath.
5. Recommendation	Translate the shared diagnosis into a specific piece of work.	Given all of that, what I would recommend is…	The buyer has heard a specific shape and agreed or declined a next step.	Presenting three options and handing over the choosing.

If any row is weaker than the one above it, the call is weaker than its strongest stage. The Map is only as credible as its least considered line.

When the SEO Question Arrives Mid-Call

Sooner or later — usually somewhere between Constraints and Ambition — a version of this question will land: "How is this different from what our SEO agency already does?" It is not an objection. It is a category-clarification question, and it is a good sign; buyers only ask it when they are taking the conversation seriously.

The instinct to answer inside a competitive frame — by explaining, at length, why AI Visibility is better than SEO — is the wrong instinct. It puts the buyer in the uncomfortable position of having to defend an existing supplier they mostly like. The stronger move is to answer as a category question rather than a competitive one: SEO optimises how a business is found by search engines; AI Visibility optimises how a business is described by answer engines. Both matter. Neither replaces the other. That is a sentence the buyer can carry back to their SEO agency without a fight.

Once the category has been separated cleanly, return to the stage of the Map you were in. The reframe is a two-minute detour, not a change of conversation.

From Recommendation to Package

The Recommendation stage is where the Map hands the call over to work that already exists. The Discovery Conversation Map and the AI Visibility Package Blueprint are complementary methodologies: one earns the right to recommend, the other defines what is being recommended. This guide teaches how to run the conversation. Cornerstone Guide 01 teaches how to design the offer that conversation lands into.

That order matters. Buyers do not choose the agency with the deepest capability; they choose the agency that gives them the highest confidence that a decision to hire has been well made. A recommendation that arrives into a pre-designed package signals that the agency has thought about businesses like this one before. A recommendation invented in the twenty minutes after the call signals the opposite, however impressive the final document.

The practical implication is that the quality of a first call is partly decided long before the call itself. An agency that has designed its offer with restraint can afford to run a discovery that is genuinely open. An agency that has not will feel the pull to close in the room — because the room is where the offer is being invented, and the offer is what needs the buyer's agreement to exist at all.

A Discovery Call Checklist

Three short field checklists — one for before the call, one for during, one for after. Use them as a practical reference alongside the Map, not as a substitute for it.

Before the call

• Read whatever the prospect has published — site, LinkedIn, recent talks — for no more than fifteen minutes, then stop.
• Write a single sentence describing what you think this business does and who it sells to; test it in Stage 1.
• Decide which shape of engagement is most likely to be the right recommendation, and be ready to abandon that guess by Stage 4.
• Choose one question you will not skip in each stage.

During the call

• Spend the first five minutes on Context only; resist the pull to move faster.
• Ask one question at a time, and let the silence after it do its work.
• Reflect the buyer's Ambition back in their own words, not yours.
• State the Diagnosis in one or two sentences and pause before the Recommendation.
• Make one specific recommendation, not three.
• Close the call with a specific next step and a specific date.

After the call

• Write a three-line summary within an hour: what the business does, what the buyer said they want, what you recommended.
• Send a short written recap the same day, in the buyer's language, not yours.
• Name the stage you suspect was skipped if the call ended in "we need to think about it", and revisit it explicitly next time.
• Log what worked and what did not; the Map improves through iteration, not memorisation.

If a call cannot be described, afterwards, in the language of these five stages, the call was probably a version of the two failures this guide opens with — a pitch or a questionnaire — and the next call is the place to reset it.

Closing Perspective

A discovery call is, in the end, an act of restraint. The agency that runs it well is not the one with the sharpest questions or the most polished methodology; it is the one willing to let the buyer be the protagonist of their own conversation for long enough that a recommendation, when it arrives, feels less like a sale and more like an obvious next step.

The Map's job is to make that restraint reproducible. First calls stop being lucky, and start being designed.

The right to recommend is earned in the discipline of understanding, not in the moment the recommendation is spoken.

[[key-takeaways]]

• Discovery is the first piece of the engagement, not the free part before it.
• Each stage of the Map earns the right to the next, and skipping one weakens every stage that follows.
• Questions build trust faster than answers, especially in the first ten minutes of a first call.
• A diagnosis is only useful when it is specific enough that it could only describe this business.
• A recommendation should land into a package the agency has already designed, not into a proposal it is about to invent.
• The right to recommend is earned in the discipline of understanding, not in the moment the recommendation is spoken. [[/key-takeaways]]

[[frameworks-used]]

[[mental-models-used]]

[[related-frameworks]]

Methodology & Sources

This guide sits inside the wider LaunchAnAEO methodology for building an AI Visibility agency. It introduces the Discovery Conversation Map as the model for running a first client call, and rests on two ideas developed elsewhere in the Knowledge Base: that questions build trust faster than answers, and that buyers pay for confidence before they pay for capability.

It is written to be read alongside Building Your First AI Visibility Package. Where this guide ends — with a recommendation landing into a pre-designed offer — the packaging guide begins. Together, the two describe a single motion: the package is what the agency has already designed; the discovery call is the conversation that tests whether it fits the buyer in the room.

The brief detour into the SEO reframe is a short preview of a distinction the LaunchAnAEO methodology treats in full elsewhere: that AI Visibility and SEO are different disciplines that serve the same business, and that the sharpest agencies teach their clients to see them as complementary rather than competing.