Run a Diagnostic Sales Call

A diagnostic call is not a performance in which you persuade the buyer to like you. It is a structured conversation that helps both people decide whether a small paid engagement is useful. You need to understand the current workflow, the consequence of the problem, the decision path, the data constraints, and the next action. The buyer needs to understand what you can honestly deliver.

The call can end with a paid pilot, a follow-up, a referral, or a no. A good no is useful because it prevents an expensive project built on a weak fit.

Agree on the Purpose

Open with a simple frame:

> I would like to understand what happens now, what a useful result would look like, and whether a small paid pilot makes sense. I will not try to determine your whole business strategy in twenty minutes.

This reduces the pressure to perform. It also gives the buyer permission to correct your assumptions.

Confirm:

• The time available.
• Whether the buyer can discuss the operational details.
• Whether someone else needs to approve a report or a purchase.
• Whether the conversation may be recorded. Do not record without permission and an appropriate lawful basis.

If the buyer says, "Just send me a proposal," ask one diagnostic question before making a document:

> I can do that. To avoid sending a generic proposal, what happens today when the workshop transcript is ready, and who owns the follow-up?

A proposal built before diagnosis is often a guess with a price attached.

Use a Five-Part Agenda

Keep the call focused:

1. Context and current process.
2. Trigger, frequency, and consequence.
3. Decision roles and constraints.
4. Desired result and evidence of done.
5. Fit, next step, or disqualification.

Take notes in short phrases. Do not hide behind a script. The script exists to remember the important questions, not to prevent the buyer from correcting the agenda.

Diagnose the Recent Past

Ask about what happened rather than what might happen:

• When did this last occur?
• What was the starting input?
• Who did which step?
• Where did the work wait?
• How was it checked?
• What went wrong or required rework?
• What did the buyer do instead when it was late?
• What happened next for the customer or team?

Past behavior is usually more concrete than a prediction. A buyer who has never spent time or money on the problem may still care, but the evidence for urgency is weaker.

Avoid leading questions such as "You must lose a lot of time on this, right?" Ask what happened and let the buyer supply the meaning.

Quantify Without Pretending to Be Certain

Use a small calculation when the buyer has numbers:

> The process happens twice a month, takes about three hours, and two people review it. Does the three hours include both reviews?

If the buyer says yes, record the assumption and calculate a scenario. If the buyer does not know, write "unknown." Do not build a sales case on a number that the buyer never confirmed.

A simple worksheet:

• Frequency per month: 2, stated.
• Time per cycle: about 3 hours, buyer estimate.
• People involved: 2, stated.
• Rework per cycle: about 1 hour in some months, uncertain.
• Deadline consequence: missed the next event once, stated.

This worksheet is more honest than a single "saves 96 hours a year" claim generated by multiplying assumptions.

Ask About the Current Workaround

The alternative tells you a lot about willingness and fit:

• Doing it manually.
• Asking an employee to work late.
• Using a generic template.
• Skipping the follow-up.
• Hiring a general assistant.
• Using an approved AI assistant with no review.
• Waiting for a larger vendor.

Ask:

> What have you already tried, and why did it stop working?

Do not criticize the workaround. It may be sensible for the buyer's current stage. Your job is to identify whether a different result is valuable and feasible.

Identify the Decision Path

Ask:

• Who experiences the problem day to day?
• Who supports a change?
• Who approves a small paid test?
• Who controls the data or access?
• Who can stop the project?
• What does the approver need to see?
• What budget or procurement step applies?

Do not ask a user to reveal confidential budget details they cannot share. Ask about process. If the decision owner is not on the call, ask what information would help that person decide and whether a short follow-up with them is appropriate.

Test the Buyer's Willingness to Act

Evidence of willingness can include:

• A named trigger and deadline.
• A person responsible for solving the problem.
• Availability of a real input.
• A stated path to approve a small purchase.
• A prior spend or action that addressed the same problem.

A buyer may be interested but not ready. You can offer a later follow-up date or a smaller diagnostic. Do not push a pilot that depends on a budget cycle outside your runway and call it a near-term sale.

Describe the Pilot Only After Diagnosis

When the buyer has described the current state, translate it back:

> So the recurring job is turning one approved transcript into four follow-up sequences before the next event. Today, the owner and one event coordinator draft manually, and the last sequence was late. The owner approves a small paid test, and the coordinator controls the source file. The result would be twelve source-checked drafts within five business days.

Then ask:

> What did I get wrong?

Use the correction to revise the scope. Only then explain a specific pilot, price, and next step. Keep the explanation to the same one-page offer structure:

• Result.
• Boundary.
• Evidence of done.
• Inputs and reviews.
• Price.
• Why now.

Do not dump every feature or use AI jargon to establish expertise. A small pilot should be easy to repeat back.

Handle Objections as Data

Common objections include:

• "We do not have budget."
• "We already use AI."
• "We need to wait."
• "Our data is too sensitive."
• "We need a large provider."
• "Can you do it for free first?"

Respond with a question before a counterargument:

> What budget path would a small pilot need to fit?

> Which part of the current AI workflow is still producing rework?

> What event would make this urgent?

> Which data would have to stay out of the workflow?

If the answer shows a hard exclusion, disqualify respectfully. If the objection is a missing unknown, record it and decide whether a smaller step can answer it. Do not solve every objection by lowering the price.

Decide by the End of the Call

End with one of these outcomes:

• Pilot fit: Confirm the result, input, owner, review dates, and buyer for a written offer.
• Diagnostic follow-up: Agree on the missing person, data, or decision and set a date.
• Referral: Ask for a named role and permission to mention the conversation.
• No fit: Explain why and close the loop.
• No decision: Record what remains unknown and when to check again.

A call without a recorded next action tends to decay. The action should have an owner and a date.

A Worked Diagnostic Call

The following is an illustrative composite, not a transcript of a real client.

Nina speaks with the owner of a small workshop company after a referral. The owner says the team is "looking into AI." Nina does not immediately explain her workflow.

She asks about the last workshop. The owner describes:

• The transcript was ready on a Thursday.
• An event coordinator removed attendee names.
• The owner spent about five hours drafting follow-up messages.
• A sequence was not sent before the next event.
• The owner approves contractor spending under a local threshold.
• The coordinator controls the shared drive and would need to approve access.

Nina repeats the situation and asks what she missed. The owner corrects one detail: the five hours included checking an AI-generated draft that contained two unverifiable claims.

Nina records the correction. The value hypothesis is not "AI saves hours." It is "a source-checked draft workflow may reduce the owner's review and prevent an unapproved claim from reaching the final sequence."

She offers a pilot with:

• One transcript and one notes file.
• Four customer moments.
• Twelve drafts with source notes.
• Two review rounds.
• A human accuracy and privacy check.
• A five-business-day delivery window.
• A fixed price and a written scope change rule.

The owner is interested but asks for a free sample. Nina responds:

> I understand why a sample feels safer. A free custom sample would still require your transcript and review time. The paid pilot is scoped so you can judge the workflow on one real input, and the first input check happens before we begin. If privacy approval is the concern, we can start with a redacted excerpt and confirm the controls before full access.

The owner agrees to send the one-page pilot and to include the coordinator in the approval. Nina records a follow-up date. She does not call this a closed sale.

Write a Useful Call Note

Use headings:

• Participants and consent.
• Current process.
• Trigger and frequency.
• Consequence and evidence.
• Workaround and prior spending.
• Decision roles.
• Data and access constraints.
• Desired outcome.
• Fit and unknowns.
• Agreed next step and owner.

Separate facts from interpretation. If the owner says "the sequence was late once," do not write "chronic missed deadlines." If the coordinator says "we usually redact names," do not write "privacy process is complete." Preserve the exact limitation.

Your Chapter Exercise

Run one 20-minute practice conversation with a peer or trusted colleague:

1. Open with the diagnostic purpose and time boundary.
2. Ask for a recent example of the workflow.
3. Trace the process from input to final result.
4. Ask about time, consequence, and current workaround.
5. Identify the user, champion, budget owner, and approval path.
6. Ask what data and access would be required.
7. Repeat the situation and invite corrections.
8. Explain one bounded pilot only after the buyer's situation is clear.
9. End with one named next action and date.
10. Write a note separating evidence from interpretation.

Chapter Deliverable

Create a call kit containing:

• Purpose statement.
• Five-part agenda.
• Recent-behavior questions.
• Role and approval questions.
• Input and privacy questions.
• A place to record facts and unknowns.
• A short pilot recap.
• Disqualification language.
• A call-note template.

Source and Decision Caution

Requirements for recording calls and handling personal information vary by jurisdiction. A diagnostic conversation is not a substitute for qualified legal, tax, financial, health, or benefits advice. If the buyer's problem involves a regulated decision or a high-consequence action, explain your limits and refer the relevant decision to an appropriate professional.

Do not record a call without the required permission, notice, and lawful basis. Keep notes in a controlled location, store only the minimum needed, and protect any personal or confidential information disclosed during the conversation. Do not paste transcripts or notes into an unapproved AI tool.

AI-generated summaries can omit a caveat, misstate a number, or attribute a statement to the wrong person. Review summaries against the actual notes, correct errors, and do not treat a generated action item as approved until the responsible person confirms it.

The next chapter turns the diagnosis into a proposal that gives the buyer one outcome, one price, one timeline, and one clear decision.