Coach AI adoption like any sales skill: pick one job for the AI, show it working on a real deal, attach it to a moment in the rep's week, and ask about it in every one-on-one.
What the job is, who does it, how long it takes
Coaching AI adoption means getting reps to use an AI tool in their real selling work, often enough that it changes how they prepare and follow up. Licenses and logins take a day. The habit takes weeks: a rep reaching for the tool before a discovery call without being told to.
The front-line sales manager owns it. Enablement can build the training, and RevOps can switch the tool on, but reps copy what their manager does and answer to what their manager asks about. A rollout that leaves managers out tends to show a spike of logins in week one and a slow fade after.
Plan for about six weeks per job. That is long enough to see whether a habit forms and short enough that the team still remembers why it started. Most teams need two or three of these cycles before AI is part of how they sell.
Why a launch event does not do it
Training days create awareness and little else. In our Game Winning Kickoffs study of public accounts, of about 4,800 first-hand accounts of sales kickoffs, 9 people described using something from the event afterwards. The research on training transfer says the same thing in a more careful way: what happens back at work is more strongly tied to whether a skill gets used than what happens in the training room (Saks and Belcourt 2006; Blume and colleagues 2010).
For AI adoption, "back at work" means the manager. A rep who is shown a tool once and then never asked about it has no reason to change a routine that already gets them through the week. A rep whose manager asks, every week, "what did the coach say about this deal?" has a reason.
Spacing helps too. Practice spread over weeks beats practice crammed into one session, with an effect size of 0.46 across 63 studies (Donovan and Radosevich 1999). Six weeks of small, repeated use is the spaced version of a launch day.
The six-week plan, step by step
The plan below assumes one manager, a team of six to ten reps, and one AI tool that is already switched on. It works for an AI sales coach, a call summarizer or a drafting assistant; the example later uses a coach.
- Week 1One job, shown live
- Pick the job
- Manager goes first on a real deal
- Weeks 2 to 3The habit
- Tie it to a calendar trigger
- Ask about it in every one-on-one
- Weeks 4 to 5Check the quality
- Review one output together
- Share the best example
- Week 6Add the second job
- Only once the first one sticks
1. Pick one job for the AI to do
Choose a single, frequent moment in the rep's week where the AI saves time or improves the result. Good first jobs are call preparation before every first meeting, a follow-up email draft after every discovery call, or a deal review before every proposal. Write it as one sentence: "Before every first meeting, the rep talks the deal through with the coach for five minutes."
Avoid starting with the job leadership cares about most, like forecast hygiene. Reps adopt tools that help them win the next call; they tolerate tools that help someone else.
2. Go first, on a real deal, in front of the team
In the next team meeting, the manager runs the job live on a deal from the team's own pipeline. If the tool is an AI coach, the manager starts a call with it by voice and talks a real deal through, out loud, with the team watching. Typing is the fallback for a room where talking is not possible.
This does two things. It shows the reps the tool on a deal they recognize, and it shows them the manager is willing to be coached in public. Pick a deal where the AI is likely to find a gap. A demonstration where the manager's deal comes out perfect teaches the team that the tool flatters.
3. Tie the job to a trigger in the calendar
A habit needs a cue. Attach the job to an event the rep already has: the calendar invite for a first meeting, the end of a discovery call, the moment a deal moves to proposal. Some teams add the coach link to their meeting template so it sits in the invite.
Keep the ask small. Five minutes before a call is a habit. Thirty minutes is a project, and reps skip projects when the week gets busy.
4. Ask about it in every one-on-one
This is the step most rollouts skip, and it is the one that works. In each one-on-one, the manager asks one question about the job: "What did the coach push you on before Tuesday's call?" Then listens to the answer.
Ask about the content of the session. Asking "did you use it?" turns adoption into compliance, and the rep learns to log in and close the tab. Asking what the coach said, and what the rep did with it, turns it into a coaching conversation about the deal.
5. Make the privacy rules explicit
Before week two, tell the team in plain words what the manager can and cannot see. Reps who suspect every rehearsal is scored and reported will stop rehearsing the hard parts.
Our own model is set out on how we use your data: each user-and-coach conversation is a window inside the account, and a manager's read from Bryce, our AI sales expert for managers, carries the pattern and the call moment while the transcript stays with the rep. Whatever tool you use, find out its rules and say them out loud. The question of whether AI sales coaching is just rep monitoring is the one every rep is asking quietly.
6. Review the quality of one output together
In weeks four and five, pick one AI output a rep produced and look at it together: a call plan, a draft email, a deal read. Ask whether it was right, whether it was specific to this buyer, and what the rep changed before using it.
This teaches judgment, which is the skill that matters most with AI. Generic AI produces average advice that sounds plausible. A rep needs to spot when an answer could apply to any deal and push back. Share the best example with the team, with the rep's permission.
7. Add the second job only when the first one sticks
At week six, check one thing: do most reps do the first job without being reminded? If yes, add the second job and run the cycle again. If no, find out why before adding anything. More features on top of a habit that has not formed spreads the team's attention thinner.
A worked example
This example is illustrative. Arrowmere, a 600-person logistics firm, gives its eight-rep mid-market team an AI sales coach. The team sells on MEDDPICC, so the coach is Caty, set up with the team's scorecard. The manager, Dana, picks the first job: "Before every first meeting, a five-minute call with Caty on the deal."
In the week-one team meeting, Dana starts a call with Caty in one click and talks through a stalled deal with a regional grocer. Caty asks who signs above the budget limit, and Dana does not know. The room laughs, and two reps admit they do not know either on their own deals. That moment does more for adoption than the slide deck enablement had prepared.
Dana adds the coach link to the team's first-meeting template. In weeks two and three, each one-on-one opens with the same question: "What did Caty push you on this week?" Five reps have answers by the second week. Three do not, and Dana asks each of them what got in the way. Two say the calls clash with travel; Dana suggests they type from the airport. One says he does not trust it. Dana asks him to bring the next answer he thinks is wrong.
He brings one in week four: the coach suggested asking about a competitor the buyer had already ruled out. Dana uses it in the team meeting as the example of checking an answer before acting on it. By week six, seven of the eight reps prepare first meetings with the coach without a reminder. Dana adds the second job: a deal review before every proposal.
Where it goes wrong
Five patterns account for most failed rollouts.
- Measuring logins. A login count tells a manager the tool is open. Reps learn to open it. Ask what the tool said and what the rep did with it.
- Too many jobs at once. A launch that shows ten features produces a team that uses none of them. One job, six weeks, then the next.
- The manager never uses it. Reps watch what the manager does. A manager who asks about the tool but has never been coached by it loses the argument with the first skeptic.
- Silence on privacy. If nobody explains who sees what, reps assume the worst and rehearse only the easy parts.
- Generic advice. A tool that gives every rep the same average advice loses the team within a month. AI that knows the team's method and buyers gives answers specific enough to test on the next call.
For the wider rollout, from pilot group to the whole team, see how to roll out AI sales coaching and how to get reps to use AI coaching. The AI sales coaching guide covers choosing a tool, and the page for sales leaders shows how our experts fit a team. To plan a rollout for your team, talk to us.
FAQ
How long does it take sales reps to adopt AI?
Plan for about six weeks per job you want the AI to do. Most teams need two or three of these cycles before AI is part of the way they sell.
Who should lead AI adoption on a sales team?
The front-line sales manager should lead it, with enablement building the training and RevOps switching the tools on. Reps copy what their manager does and answer to what the manager asks about.
What is the best first use of AI for sales reps?
Call preparation before first meetings is a strong first job, because it happens often, takes five minutes, and helps the rep win the next call. Follow-up email drafts are a close second.
How do you measure AI adoption on a sales team?
Count the reps who do the agreed job without a reminder, and ask in one-on-ones what the AI said and what the rep did with it. Login counts measure access and little else.
What if a rep refuses to use AI tools?
Ask the rep to bring the next answer they think is wrong and review it together. Skeptics often become the team's best checkers of AI output once their doubts are taken seriously.