The buyer asks about price.
What happens next?
A cleaning business offers recurring weekday office cleaning. It needs a site walkthrough before preparing a quote. No price, discount or start date has been approved.
- 01Read the exchange
- 02Review the evidence
- 03Try a better opening
“We need office cleaning twice a week. How much is it?”
“Twice a week sounds sensible. We can do every office for $200 and start tomorrow. Shall I book it?”
Which commitments would you challenge before this reply reaches a customer?
Show the coaching review
The issue is the unsupported promise. The seller invents a price and start date, and skips the walkthrough needed for a quote.
- Discovery0 / 2
No questions about the rooms, access or schedule.
- Confirming the need1 / 2
Repeats the frequency; does not check the full requirement.
- Factual explanation0 / 2
The price and start date have no approved basis.
- Handling unknowns0 / 2
Fills gaps with promises instead of checking them.
- Agreed next step1 / 2
Asks permission, but skips the required walkthrough.
“Which rooms need cleaning, and when can someone access the office? We arrange a site walkthrough before preparing a quote. Would you like to discuss a suitable walkthrough time once we understand the schedule?”
Continue with the buyer’s reply. The seller still needs to confirm the requirements and agree on the next action.
Ask about the work and access before discussing the quote. Repeat that moment, then check the same five criteria.
Original fictional example. The score illustrates this coaching rubric; it is not a client result, validated assessment or sales forecast.
Use the full scenario, prompt and scoring anchorsThis exercise runs in your browser. No AI request, customer records or form submission.
Your working checklist
- Choose one difficult moment and write the approved offer facts.
- Keep the buyer scenario fictional and the scoring criteria observable.
- Separate buyer simulation from review; flag unsupported promises.
- Have a person check the feedback and agree on one action to practice.
- Track real business outcomes separately from exercise scores.
Practice the question your team keeps getting
If the team knows the offer but struggles when a buyer asks about price, timing or fit, use that moment as the exercise. Choose one objective: ask better discovery questions, explain a limitation or agree on the next step. Salesforce describes role-play and pitch practice as uses of AI sales training, with managers checking the coaching. Start with the task; choose the tool afterward.
Write a short buyer brief and approved facts
Give the simulated buyer a situation, need and practical constraint. Give the seller the offer facts they may use and mark what remains unknown. The cleaning example above uses a two-room office, twice-weekly cleaning and a preference for access after 6 PM. No price or start date is supplied. The free playbook keeps both sides together, with prompt drafts you can adapt to your own approved scenario.
Score five things you can actually observe
Review discovery, confirmation of the need, factual explanation, handling unknowns and an agreed next step. Use 0 for a missing or contradicted behavior, 1 for partial evidence and 2 for clear evidence. Record the seller’s actual words beside each rating. The playbook defines each scoring anchor. This original rubric is a coaching aid, not a validated assessment; do not use accent or personality as a proxy for the task.
Find the priority behind the score
In the worked exchange, the seller invents a price and start date before understanding the request. The useful coaching action is specific: remove those commitments and ask about the rooms and access. The total helps organize the review, but an unsupported promise still needs attention even if other criteria score well. Ask the reviewer to explain the evidence rather than simply return a number.
Repeat the difficult moment
After feedback, try the same part of the conversation again. Let the buyer answer, then check whether the seller confirms the requirements and agrees on an appropriate next action. A stronger opening is a starting point, not a complete successful conversation. Keep the scenario and facts consistent so the team can see which behavior changed. Once the method is clear, adapt the buyer situation and practice again.
Keep AI feedback open to correction
Separate the buyer simulation from the coaching review. Ask for a quote supporting each rating and “no evidence” when a behavior was not shown. NIST identifies confidently incorrect AI output as a risk, so check both the quoted exchange and the offer facts. Have a manager correct the feedback when necessary. Review the prompt drafts with your team and adjust them to the agreed learning task.
Agree on the review before involving the team
Have two reviewers score the same fictional exchange independently. Compare their evidence and clarify ambiguous criteria. Check for invented buyer facts, premature agreement and feedback that rewards a polished script over the agreed behavior. Keep the scenario, scoring anchors and review notes together. Use this for practice and coaching, with human judgment; it is not an automated hiring, pay or staff-ranking system.
Track the next action, then review real-world results separately
For each practice round, record the skill, the main feedback and one action to try next. A repeat-attempt score can reflect familiarity with the exercise. If you later assess business impact, define the period, inquiry sources and sales stages, and account for other changes. Completion, confidence, practice scores and real sales are different measures. Choose the next level of training support below around the work your team needs to do.
Before your first practice round.
Can we adapt this to a different industry?
Yes. Replace the fictional buyer and offer facts with information your business has approved. Keep the scoring tied to observable behavior. A retailer might practice a delivery question; a service business might practice discovery before a quote. Do not paste private customer records into the exercise.
Do we need a new training platform?
A colleague, a document and a clear rubric can be enough to try the method. Discuss a platform when you need structured access, lessons, practice records, progress views or integrations. Those requirements are a separate application scope.
What would we bring to a workshop discussion?
Describe one conversation the team finds difficult, who needs to practice and what they should be able to do afterward. Share your tool names and high-level needs. The inquiry does not require customer transcripts, employee records or confidential material.
Does a better practice score prove more sales?
No. It shows what was demonstrated in that exercise. Track real inquiry and sales stages separately, over an agreed period, and account for other changes. Have a person review AI feedback rather than treating a generated score as a business outcome.