A sales coaching model is a repeatable structure a manager uses to coach a rep. The models people compare work at different layers: some shape a single conversation (GROW, CLEAR), some shape a coaching cycle around a buyer call (brief, observe, debrief), and some decide what the coaching is about (a skill, or a live deal).
Teams that pick "a model" from a list usually end up combining one from each layer. This page compares them by layer and says which fits which moment. The full framework we recommend is in our sales coaching framework.
Three layers of a coaching model
The conversation model is the order of questions inside one coaching conversation. GROW is the one many managers learn first.
- Conversation modelGROW, OSKAR, CLEAR, CIGAR
- Session cycleBrief, observe, debrief
- Coaching unitA skill or a deal
- Conversation modelGROW, OSKAR, CLEAR, CIGAR
- Session cycleBrief, observe, debrief
- Coaching unitA skill or a deal
The session cycle is the structure around a buyer conversation: prepare the rep before it, observe it, debrief it after.
The coaching unit is what the coaching is about. Skill coaching works on a capability across calls. Deal coaching works on one opportunity and its next buyer step.
A model at one layer says nothing about the others. GROW tells a manager what to ask. It does not say whether to ask it about a skill or a deal, or when.
The models compared
| Model | Layer | What it coaches | Where it works | Method support | Manager view |
|---|---|---|---|---|---|
| GROW (Goals, Reality, Options, Will) | Conversation | Whatever the goal is; often a skill or a target | Any one-on-one | None built in; the team's method can set the goal | One conversation at a time |
| OSKAR, CLEAR, CIGAR | Conversation | A goal or a gap the rep names | Any one-on-one | None built in | One conversation at a time |
| AOR (Activities, Objectives, Results) | Activity | Activity levels and their results | Dashboards and one-on-ones | None | Team activity, weak on which skill is missing |
| Brief, observe, debrief | Session cycle | The rep's performance on a specific buyer call | Live calls, before and after | The team's method is the standard for the brief and the debrief | Per call, rolled up per rep |
| Deal coaching | Coaching unit | One opportunity and its next buyer step | The live pipeline | The method's qualification criteria stage the deal | Per deal, across the pipeline |
| Peer and group coaching | Delivery | Shared gaps across a team | Call reviews, team meetings | Depends on the program | Team patterns |
AOR and peer coaching sit outside the three layers. AOR measures activity, and peer coaching is a way to deliver coaching at any layer. Both are in the table because teams often weigh them against the other models.
Conversation models: GROW, OSKAR, CLEAR, CIGAR
GROW runs Goals, Reality, Options, Will. It was developed by Sir John Whitmore and colleagues, and Performance Consultants holds it as a trademark. It keeps a manager asking and stops them telling. Its weakness is that it carries no standard: a rep can leave a GROW conversation with a goal and a promise and no idea what good looks like on the next call. In sales, the fix is to anchor Reality in what the buyer said on the last call, quoted. More in our entry on the GROW model.
OSKAR (Outcome, Scale, Know-how, Affirm and Action, Review) adds a 1 to 10 scale and a review in the next session. CLEAR (Contract, Listen, Explore, Action, Review) starts by agreeing how the session will run and ends with feedback on the coach. CIGAR (Current reality, Ideal, Gaps, Action, Review) puts the gap at the center. All four are question orders. Pick one, and spend the effort on what the conversation is about.
Activity models: AOR
AOR coaches on Activities, Objectives and Results: calls made, emails sent, deals closed. It is transparent and easy to track. It tells a manager who is busy and leaves open which skill is missing on which deal.
The session cycle: brief, observe, debrief
Brief, observe, debrief puts coaching around a real buyer conversation. Before the call, the manager and rep agree what the call has to produce and what good looks like in the team's method. During it, someone captures what the buyer said in their words. After it, the debrief finds one gap and one move, with a date.
The evidence favors coaching attached to real work. Across 117 studies, training that used the learner's own scenarios and had trained managers behind it held skills better (Taylor, Russ-Eft & Chan 2005, via our Game Winning Kickoffs field study). And manager skill matters: across 1,246 pharmaceutical reps, managers' coaching skill predicted goal attainment, and frequent coaching from less skilled managers was linked to worse results (Dahling et al. 2016, same source). A cycle with a clear standard is easier to do well.
The coaching unit: skill or deal
Skill coaching picks one capability, such as putting a cost on the buyer's problem, and works it across calls for a month. Deal coaching picks one opportunity and asks what it is missing: a named problem, a cost, a reason to act now, access to the person who signs, a committed next step. Our entry on deal coaching defines it.
The two work well together. The deal gives the skill somewhere to show up, and the skill explains why the deal is stuck. A rep who keeps accepting "budget is approved" from a champion has a qualifying gap, and it will appear on the next deal too.
Scaling models: peers, groups and AI
Most lists of "scalable models" are about delivery: peer call-review pods, mentorship pairs, call libraries tagged by skill, group coaching on a shared gap. Peer pods that review calls before a manager does are a sound idea. These programs extend whatever model a team runs. The session cycle and the coaching unit still have to be chosen.
AI coaching sits here as well. Our experts build roleplays from the rep's real deals and debrief them. The harder problem is coaching every rep on every live deal when there are never enough good managers. That is a deal-coaching problem, run in a brief, observe, debrief cycle.
Which model fits which moment
| Moment | Layer to lead with | Why |
|---|---|---|
| Before a decisive buyer call | Brief, with deal coaching | The call has a job; the deal says what is missing |
| After the call | Debrief, with GROW as the question order | Reality is now what the buyer said |
| Weekly one-on-one | Skill coaching, worked on one deal | One gap a month, seen on a live opportunity |
| Pipeline or forecast review | Deal coaching | Stage checked against buyer evidence |
| Team meeting | Group coaching on a shared gap | One example, many reps |
Common mistakes when adopting a model
Picking a conversation model and stopping there. A team that trains every manager on GROW has agreed the order of questions. It has not agreed when coaching happens or what it is about, so sessions drift back to status updates.
Running a different model per manager. If one manager runs GROW, another CLEAR and a third no model at all, reps who move teams start over, and nobody can compare coaching across the floor. Pick one per layer for the whole team.
Coaching the activity number. AOR-style dashboards show who is busy. A manager who coaches from them alone ends up asking for more calls when the gap is what happens on the calls.
Skipping the brief. The brief is the session teams drop first, though it is shorter than the debrief and changes the call itself, because the rep goes in knowing what the call has to produce.
No standard for the debrief. Without the team's method as the standard, "what went well" turns into opinion. Write down what good looks like at each stage of the deal, and debrief against it.
Changing models every quarter. A model needs a few months of steady use before anyone can tell whether it helps. Keep it long enough to judge.
Worked example: two models on one deal (illustrative)
Illustrative: the rep, buyer and deal are invented.
A rep has a first meeting with a buyer's CFO on Thursday. Brief (Tuesday): the manager and rep agree the call has to produce the CFO's own number for what the current process costs. Observe: the rep records the call. Debrief (that afternoon), run in GROW order: Goal, a cost figure from the CFO. Reality, she said, "It's significant, but I'd need my team to size it." Options: ask the team for the figure, send a sizing worksheet, offer a 20-minute working session. Will: the rep emails the CFO's analyst on Friday to book the session. The skill gap behind it, accepting a vague answer on cost, becomes the rep's skill for the month.
The brief and the debrief are the two sessions that drop when a manager runs out of hours. Our AI sales experts run both: the rep calls before the CFO meeting and again after it, and the expert works the deal in the team's own method. The leader's view is at for leaders.
FAQ
What is the best sales coaching model?
No single model is best on its own, because the common ones work at different layers. A strong default is brief, observe, debrief around live buyer calls, with deal coaching deciding what each session is about and GROW as the question order inside the debrief. Pick one conversation model and use it consistently.
What is the GROW model in sales coaching?
GROW is a four-step coaching conversation: Goals, Reality, Options and Will. The manager asks what the rep wants to achieve, where things stand, what the options are, and what the rep will do next. In sales it works well when Reality is anchored in what the buyer said on the last call.
What is the difference between GROW and CLEAR?
Both are question orders for one coaching conversation. GROW starts from the goal. CLEAR starts by agreeing how the session will run, spends more time listening and exploring, and ends with the rep's feedback on the coach. Neither says what the coaching should be about.
What is the difference between skill coaching and deal coaching?
Skill coaching works one capability across several calls, usually for a few weeks. Deal coaching works one opportunity and its next buyer step. Many teams need both: the deal shows where a skill is missing, and the skill is why the deal is stuck.
How do you choose a sales coaching model for a team?
Choose one model per layer and use it across every manager. Pick a conversation model the managers will remember (GROW is enough for most), make brief, observe, debrief the cycle around important buyer calls, and decide that most sessions work a live deal. Write the team's method down as the standard the brief and debrief use, then keep the choice for at least a few months before judging it.
Can AI run a sales coaching model?
AI can run parts of one: preparing a rep before a call, reading what the buyer said, and debriefing against the team's method. The manager still owns the one-on-one and the judgment about each rep. The useful test is whether the AI coaches the live deal, or only practice.