What the job is
Building a sales playbook means writing down, for each moment of your sale, what the rep needs to find out from the buyer, how to find it out, and what good sounds like on the call. Enablement or a sales leader owns it, with two or three of your best reps and a manager. A first version covering four or five deal moments takes about three weeks.
Most playbooks are written as reference documents: company overview, personas, product sheets, collateral, a page on the comp plan. They get opened during onboarding and rarely again. A playbook gets used when it is organized the way a rep looks for help: by the moment of the deal they are in. The playbook is also the standard your reps are made ready against, which is why it sits at the center of the sales readiness guide.
The seven steps
1. Pick the deal moments
List the four to six conversations that decide your deals. A typical B2B set: first discovery, the business case, the technical evaluation, pricing and terms, the mutual plan to close. Use your stage definitions, and add a moment wherever deals stall.
- 01First discovery
- 02The business case
- 03The technical evaluation
- 04Pricing and terms
- 05The mutual plan to close
What good looks like: a manager agrees that a rep who handled these moments well would win most deals that should be won.
2. Write the exit evidence for each moment
For each moment, write what the buyer must have said or confirmed before the deal moves on. This is buyer evidence, written as the buyer would say it. For first discovery under MEDDPICC: the metric the buyer will judge success by, in their words; the cost of the current problem; who else is affected.
What good looks like: a manager can read a call transcript and say whether the exit evidence is there. If it cannot be checked in a transcript, rewrite it. This step is where your methodology, whether MEDDPICC, Gap Selling or the ValueSelling Framework®, meets your product and your buyers.
3. Mine your won deals for how the evidence was found
Pull transcripts from eight to ten recent won deals at each moment. Find the lines where the buyer gave the exit evidence and the rep question that came just before. Those questions are your plays.
What good looks like: every play in the book traces to a real call. Training built on people's own scenarios, backed by trained managers, held skills best across 117 studies (Taylor, Russ-Eft and Chan 2005).
4. Write what good sounds like, with reasons
For each play, write two or three levels: what a strong version sounds like, what an adequate one sounds like, and what a miss sounds like, with a one-line reason for each. The reasons matter more than the examples. In our call-scoring research with PeerLab, adding experienced assessors' short written reasons for each level lifted our AI scorer's average agreement with human scores from 0.62 to 0.72, the largest single gain we measured; textbook examples added nothing.
What good looks like: two managers reading the same call land on the same level, and can say why.
5. Add the objections and the next step for each moment
For each moment, the two or three objections that come up most often in your transcripts, with the question that worked, and the next step the rep should leave the call with.
What good looks like: the next step is a buyer action with a date, written so a rep can propose it word for word.
6. Test it on live deals before publishing
Give the draft to three reps with active deals at different moments. Each runs one conversation from it. Read the calls against the exit evidence and the levels.
What good looks like: the reps found what they needed in under a minute, and the calls show the evidence or show which play failed. Fix those plays before the rollout.
7. Turn it into the rubric and the coaching
The exit evidence and the levels are your certification rubric. The plays are the questions a coach asks before a call. Set an owner and a review date each quarter, and update a play when the transcripts show a better question.
What good looks like: the rubric used to certify reps is the same document, and nobody maintains a second version.
A three-week build schedule
The first version, four or five moments, fits in about three weeks if the people writing it have a few hours a week. A schedule that works:
| Week | Work | Who |
|---|---|---|
| 1, early | Agree the deal moments and draft exit evidence for each | Owner, a manager, two strong reps |
| 1, late | Pull eight to ten won-deal transcripts per moment | Owner or sales ops |
| 2 | Mark the lines where buyers gave the evidence and the questions before them; draft the plays and levels | Reps and manager, one moment each |
| 2, late | Add objections and next steps from the same transcripts | Same group |
| 3, early | Three reps run live conversations from the draft | Reps with active deals |
| 3, late | Read those calls against the levels, fix what failed, train managers | Owner and manager |
The reps who write the plays are the first to use them, which helps when the rest of the team asks whether any of it works on a real call.
Two things slow the build down. The first is trying to cover every situation at once; four or five deal moments are enough for a first release, and the situational plays can follow. The second is arguing over wording in a meeting. Let the transcripts settle it: the question that got the evidence on won deals is the play, whatever anyone thinks of its phrasing.
Worked example: one play, written out (illustrative)
The company and lines are illustrative. A logistics software company, running MEDDPICC, writes the first-discovery play.
MOMENT: First discovery
EXIT EVIDENCE (all three before the deal moves to business case)
1. Metric: the buyer names how they measure the problem today, with a number.
("On-time is 91% against a target of 96%.")
2. Cost: the buyer puts a cost on the gap, or agrees to get one.
("Each missed delivery is a re-route, about $180.")
3. Reach: the buyer names who else feels it.
("Customer service takes the calls.")
PLAY 1: Get the metric
Ask: "How do you measure [problem from their inbound note] today,
and where does it sit against target?"
Strong: buyer gives the measure and the number. Reason: metric in their words.
Adequate: buyer gives the measure, no number; rep agrees to get it.
Reason: metric named, number missing.
Miss: rep accepts "it's a big problem". Reason: no measure, deal cannot be sized.
OBJECTION: "We're just exploring."
Ask: "What made it worth an hour this month?"
NEXT STEP: buyer shares last quarter's on-time report before the business case call.
A rep opening this before a call gets one question to ask, the answer to listen for, and the next step. A manager reading the transcript afterwards can score it in two minutes.
What goes in the playbook, and in what order
The content most playbooks carry is still useful. It moves behind the deal moments.
| Section | What it holds | Where it sits |
|---|---|---|
| Deal moments | Exit evidence, plays, levels, objections and the next step for each | First, one page per moment |
| Situational plays | What to do when a deal goes quiet, a new stakeholder appears or a competitor enters | Second |
| Buyer roles | What each role cares about, in their words, taken from transcripts | Linked from the plays that need them |
| Competitive notes | Where each competitor tends to appear, and the question that worked against it | Linked from the objections |
| Product and pricing | The short version a rep needs on a call, and where the detail lives | Reference |
| Process and tools | Stage definitions, CRM fields, approval rules | Reference |
If a section cannot be linked from a deal moment, ask whether a rep would open it before a call. If not, it belongs in onboarding material.
Plays for moments outside the stages
Some of the most useful plays are for situations that can happen at any stage. Write them in the same format: the evidence to get, a question that has worked, the next step.
- The buyer goes quiet. Evidence: whether the problem still matters to them, and what changed. A question to use in place of "just checking in": "Last time you said the re-routes were costing you. Is that still where it sits, or has something else moved ahead of it?"
- A new stakeholder appears. Evidence: what this person measures and what would make them say no. Next step: a short call with them before the next group meeting.
- A competitor enters late. Evidence: what prompted the buyer to look, and which of their own criteria the competitor answers. Avoid arguing features before you know why the competitor came in.
- The champion leaves. Evidence: who now owns the problem, and whether they feel it. Next step: a meeting with that person to restate the case in their terms.
- Renewal or expansion. Evidence: the metric the buyer set at the start and where it sits now. The play is first discovery again, on the buyer's own results.
Mine these from transcripts the same way: find the calls where the deal recovered and the question that came just before.
Playbooks for launches and new segments
The core playbook covers your main sale. Two other situations usually need a short playbook of their own, built the same way.
A product launch changes what buyers ask at one or two moments, usually discovery and the business case. Write new exit evidence only for those moments and reuse the rest. Before launch there are no won deals to mine, so test the draft plays on a handful of live conversations and replace them with transcript-tested questions as calls come in.
A new segment or buyer role, such as selling to finance after years of selling to operations, changes the language. Keep the moments, rewrite the evidence in the new buyer's terms, and mine the first won deals in that segment as soon as they close.
Rolling it out
A playbook that is announced in an email gets read once. A rollout that ties it to the calls reps are already having lasts longer.
- Train managers first. Walk them through the moments, the levels and how to use a play when prepping a call. They are the ones who bring it into the deal conversation.
- Launch one moment at a time. Start with the moment where deals stall most. Two weeks later, add the next.
- Use it in the meetings you already run. Deal reviews ask about the exit evidence for the deal's current moment. One-on-ones open on one play.
- Certify against the levels. The same document is the rubric.
- Collect better questions. When a rep finds a question that gets the evidence more often, it goes into the play with the transcript it came from.
How to tell whether the playbook is working
Count evidence on live deals. For each moment, check a sample of calls each month: did the exit evidence appear, and which play was used? Track how many deals move to the next stage with all the exit evidence on file, and where deals stall. A playbook that is working shows the evidence appearing earlier in the deal and fewer deals advancing without it. Page views and completion rates only say the playbook was opened.
Where it goes wrong
- Organized by content. A rep with a call in an hour does not want the company overview. Put the deal moments at the top and the reference material behind them.
- Plays as scripts. Scripts break the first time a buyer goes off them. Write the evidence to collect and a question that has worked, and let the rep get there in their own words.
- Written from opinion. A play nobody can trace to a won deal is a guess. Mine the transcripts.
- No owner after launch. Products, pricing and competitors change. A playbook with no review date is out of date within a quarter.
- Published and left. What happens back at work predicts whether people use training better than what happens in the session (Saks and Belcourt 2006; Blume et al. 2010).
Sales playbook, sales process and GTM playbook
A sales process lists the stages a deal passes through and the rules for moving between them. A sales playbook tells the rep what to do inside each stage: the evidence to get, the questions that get it, and what good sounds like. The process is the map. The playbook is how to drive each stretch of it.
A go-to-market playbook sits above both. It covers how marketing, sales and customer success share one view of the buyer: the segments you sell to, the stages every team uses, the handoffs between them, and the numbers each team owns. When a company refreshes its go-to-market plan, usually because of a new segment, a new product or a new pricing model, the GTM playbook changes first and the sales playbook follows.
The practical test for a sales team: if a rep cannot open the document an hour before a call and find the question to ask, it is a process document or a GTM plan, however it is titled. Keep the three separate and link them. The sales playbook's deal moments should use the same stage names as the CRM and the GTM plan, so nobody has to translate.
Writing a playbook that AI can use
More teams now want AI agents to read the playbook too: to prep calls, score them, or update the CRM from what the buyer said. A playbook built with the steps above is most of the way there: exit evidence that can be checked in a transcript (step 2) and a reason for every level (step 4) are what a model needs as well. "Buyer names the metric with a number" can be checked; "rep builds strong rapport" cannot, and an AI will guess. Two more habits matter.
- Keep one source. If the playbook lives in a slide deck, a wiki page and an AI tool's prompt, the three drift within a quarter. Pick one document and point everything at it.
- Date every change. When a play changes, scores from before and after are no longer comparable. A dated change log tells managers which.
Getting the playbook onto every live deal
Reps open a playbook when it answers the question in front of them: what to ask this buyer on the next call. A manager who preps the call with the rep brings the right play to them. Most managers cannot do that for every rep on every deal.
Replicate Labs can build an AI sales expert around your own playbook to do it. The rep brings a live deal to the expert on a call, by voice or video, and it runs your methodology and your plays on it: which exit evidence is confirmed, which is missing, and the question for the next conversation. Typing is the fallback. The expert remembers the deals the rep has brought, and managers see through Bryce which parts of the playbook the team still skips.
For enablement teams, start at the enablers page or the build selling skills use case. To check the playbook is changing behavior, see how to measure sales enablement; the short definition is in what is a sales playbook?, and the readiness term in what is sales readiness?.
If your playbook has exit evidence written for every stage and the stuck deals still have none of it, talk to us about turning the plays into an expert your reps can call.
FAQ
How do you build a sales playbook? List the four to six conversations that decide your deals. For each, write the evidence the buyer must give before the deal moves on, then mine won-deal transcripts for the questions that got it. Add what good sounds like with a reason for each level, the common objections and the next step, test it on live deals, and turn it into your rubric.
What should a sales playbook include? For each deal moment: the exit evidence, the plays (questions that have worked), what strong, adequate and missed versions sound like with reasons, the common objections, and the next step. Personas, product detail and collateral belong behind those sections as reference. Put the deal moments first, because that is where a rep looks before a call.
What is the difference between a sales play and a sales playbook? A sales play is one repeatable approach to one situation, such as getting the buyer's success metric in first discovery. A sales playbook is the full set of plays for every moment of your sale, plus the evidence each moment needs and what good sounds like. The playbook is the standard; plays are the moves inside it.
Why don't reps use the sales playbook? Usually because it is organized like a reference manual and nobody brings it into the deal conversation. Reps look for help by deal moment, an hour before a call. Organize the playbook that way, write plays from your own won deals, and have a manager or coach use the play when prepping each important call.
Who should own the sales playbook? One named person, usually in enablement or sales leadership, with two or three strong reps and a manager writing it. The owner sets the quarterly review, decides which new questions go in, and keeps the playbook and the certification rubric as one document.
How long should a sales playbook be? One page per deal moment, plus the situational plays. A rep should find what they need for the next call in under a minute. Reference material can be as long as it needs to be, as long as it sits behind the moments.
How often should a sales playbook be updated? Review it every quarter with a named owner, and update a play as soon as transcripts show a better question or a new objection. Update the affected moments immediately when pricing, packaging, the product or the competitive set changes, and recertify reps on those moments only.