Gong Coaching: Features, AI Trainer and Where Practice Fits
On this page
TL;DR
- Gong coaching today has two layers: review tools built on recorded calls (scorecards, AI Call Reviewer, comments, feedback requests, a coaching inbox, coaching metrics) and practice tools (AI Trainer roleplay and Dry Run rehearsal before a meeting).
- Gong's AI Trainer builds a practice buyer from a written scenario and, optionally, from linked Gong calls, then grades the session with the same scorecards used on real calls, according to Gong's help center.
- We found no gong.io page or help article that describes prompts shown to a rep during a live call. Gong's coaching works after the call (review) and before it (practice), so ask Gong directly if in-call cues matter to you.
- Gong publishes no prices. Its pricing page says licenses are priced per user, plus a base fee tied to user count (checked October 2026). The help center lists AI Trainer on Gong's full enablement plan only.
- Teams pair Gong with a separate roleplay tool when they need practice Gong does not describe, such as several buyers in one call or screen-share demo practice, or when their calls live in more than one recorder.
Gong coaching features at a glance
Feature | What it does | Works on | Where Gong documents it |
|---|---|---|---|
Scorecards | Managers answer scorecard questions about a recorded call | Real calls | |
AI Call Reviewer | Suggests scorecard answers, or scores every matching call automatically | Real calls and AI Trainer sessions | |
Comments and feedback requests | Timestamped notes on a call; reps can ask a manager for feedback | Real calls | |
Coaching inbox | One view of recent coaching activity and open requests per rep | Managers | |
Coaching metrics | Tracks whether coaches attended, listened to, commented on or scored each call | Managers of managers | |
AI Trainer | Timed, spoken roleplay with an AI customer persona, scored on a scorecard | Practice | |
Dry Run | Rehearse an upcoming customer meeting with an AI version of that buyer | Practice |
How we checked
We read Gong's coaching page at gong.io/sales-training, its revenue enablement solution page, its enablement product page, its pricing page, its newsroom (including the September 30, 2026 Celebrate '26 release), and the help center articles linked in the table above. All pages were checked in October 2026.
Quotain publishes this article and sells AI sales roleplay, so read the Quotain sections with that in mind. Gong features change often, and several of the help articles we read carry a stale flag in their page data. Confirm anything you plan to buy against a Gong demo.
What Gong's coaching features do today
Scorecards and AI Call Reviewer
Gong's coaching starts on the call page. A manager opens a recorded call, clicks the star icon and scores it against a scorecard the team defined, according to Gong's introduction to coaching. Gong suggests starting with one scored call per rep each week.
AI Call Reviewer speeds that up in two modes, per the AI Call Reviewer article. In suggested mode, Gong reads the transcript and proposes answers that the reviewer can accept, edit or replace. In automatic mode, Gong scores every call that matches the scorecard's filters as soon as it happens. Both depend on an admin building the scorecard first.
Coaching inbox, comments and feedback requests
Reps can request feedback from a manager on a call, and anyone can leave a timestamped comment and tag a teammate. The coaching workflows article describes the coaching inbox as the place where a manager reviews recent coaching activity, sees when they last gave feedback to each rep, and sees open requests. Coaching that happened outside Gong, such as in a 1:1, can be logged with "Mark as feedback given."
Coaching metrics
The coaching metrics report is about the coaches, not the reps. Per Gong's coaching metrics data article, it records five yes-or-no flags per call: whether the coach attended, listened, commented, scored the call, or marked feedback as given. Gong's coaching page also shows rep-level measures such as talk ratio, topic duration and question rate.
AI Trainer
AI Trainer is Gong's roleplay feature. An admin writes the scenario (customer company, meeting objective, relationship background, your positioning) and can link Gong calls and snippets, which Gong uses to build a customer persona with realistic behavior, according to the AI Trainer article. A difficulty setting makes the persona receptive, neutral or resistant.
Reps find assigned trainings under Coaching, then My trainings. Sessions are spoken, last 5 to 30 minutes, and can be repeated as many times as the rep likes before submitting, per the rep guide. After each session an AI coach gives private feedback, and the rep can ask follow-up questions. When a scorecard is attached, the session is scored by an automatic review scorecard, which can match the scorecard used on real calls.
Dry Run
Gong's enablement product page describes Dry Run as live AI roleplay for an upcoming customer meeting, with Gong copying the real customer persona and account context so the rep does no setup. The page does not say which plan includes it. Ask for a Dry Run demo on one of your own upcoming meetings before counting on it.
Live coaching during calls
Searchers type "gong live coaching" and "gong real time coaching," but the Gong pages we read describe review after the call and practice before it. The help center index returned no article on live or real-time coaching, and the coaching and enablement pages do not describe prompts shown to a rep while the buyer is talking. If in-call cues are a requirement, ask Gong to show them on a live call. Our guide to real-time AI sales coaching covers where live prompts help and where they get in the way.
Where call review stops and practice begins
Our post on AI roleplay vs conversation intelligence covers the general split. With Gong the line is less clean, because Gong sells both. A more useful way to draw it is by what each tool can produce.
Call review produces a finding: "reps quote price before they confirm scope," backed by scored calls and timestamps. It cannot give the rep a second attempt at that moment, because the real buyer has moved on.
Practice produces attempts: the rep runs the same moment five times and gets graded each time. It cannot tell you which moment matters, unless something feeds it real call evidence.
So the questions to ask of any Gong setup are practical ones:
- Does the finding from a scored call turn into a practice assignment without someone rebuilding the scenario by hand?
- Does the practice session use the same scorecard as the real call, so a pass in practice means the same thing as a pass with a buyer?
- Can the rep restart at the exact moment that went wrong, or only from the top of the call?
- Does a later real call get checked for the same behavior?
Gong's own answer to the second question is yes: AI Trainer sessions can use the scorecards applied to live calls. The first and third depend on how your team builds trainings, so test them in a pilot.
How teams pair Gong with roleplay
Pattern 1: Gong only
Teams on Gong's full enablement plan can keep coaching in one place. Managers score calls, spot a pattern in team stats, build an AI Trainer scenario from the calls that show it, and assign it. This works when one vendor, one login and one scorecard library matter more than practice depth.
Pattern 2: Gong for evidence, a roleplay tool for practice
Here Gong stays the system of record for calls, and a separate roleplay tool runs the practice. Teams choose this when they want practice formats that the Gong pages we read do not describe, such as several AI buyers in one session or sharing a screen during a demo, or when they want practice outside Gong's plan structure.
Quotain fits this pattern. Gong can be connected to Quotain for setup and research, so Quotain's assistants can use Gong conversations alongside other connected sources (Quotain changelog, May 12, 2026, quotain.com/changelog). The AI sales simulations page says Quotain reads org context, the web, the last Gong call and the knowledge library when it builds a buyer.
Pattern 3: Grade the real call, then practice the missed moment
The tightest loop grades the real call and the practice attempt on the same scorecard. Quotain's real-call grading grades recorded calls, including calls imported from Gong, against the team's own scorecard. Each requirement is marked Met, Not met or N/A with the transcript lines behind it, and a manager can correct a grade with a reason. From any transcript line, "Continue call from here" starts a practice session at that point with an AI buyer grounded in the recording, so the rep retries the exact moment. On Quotain's pricing page, real-call grading and buying-committee calls with several buyers are listed as on request for Team plans.
For the step between a recording and a scenario, see how to turn sales call recordings into roleplay scenarios. For a vendor-level comparison of a roleplay specialist and Gong, see Hyperbound vs Gong.
Where Gong fits better than adding a roleplay tool
Gong's own practice features are the better choice in several cases:
- You already pay for Gong's full enablement plan. AI Trainer and Dry Run are part of what you bought, and a second tool adds a contract, a login and a security review.
- Your scenarios come from Gong calls. AI Trainer can build personas from linked Gong calls and snippets without exporting anything.
- You want one scorecard library. Gong applies the same scorecards to real calls and practice sessions, and AI Call Reviewer grades both.
- Your managers live in the coaching inbox. Practice results and call feedback sit in the same place they already check each week.
A separate tool earns its place when the gap is practice depth: committee calls, demo practice with a shared screen, drills that restart at one moment, or practice tied to calls that sit outside Gong. Quotain also has limits buyers ask about. It does not yet have SOC 2 certification or SSO, per its trust page, and practice uses still buyer portraits, not video avatars.
How to choose: a two-week pilot scorecard
Run the same test on Gong's practice tools and on any roleplay tool you are considering. Use one team, one call type and one behavior you have already seen fail on recorded calls.
What to test | How to run it | Score 0, 1 or 2 |
|---|---|---|
Finding to assignment | Take one scored call with a clear miss. Time how long it takes to turn it into an assigned practice scenario. | 0 if rebuilt by hand over days, 2 if assigned the same day |
Buyer realism | Have two experienced reps run the scenario. Ask whether the buyer pushed back the way the real buyer did. | 0 if generic, 2 if reps say it matched |
Same scorecard | Grade a practice attempt and a real call on the same criteria. | 0 if separate rubrics, 2 if identical |
Retry at the moment | Ask the rep to restart at the missed moment only. | 0 if full call only, 2 if they can start at the moment |
Call format | Run the call type you sell: committee call, technical demo, renewal. | 0 if the format is not supported, 2 if it runs as sold |
Manager time | Log manager minutes per rep per week for review and assignment. | 0 if it rises, 2 if it falls |
Transfer | Score the same behavior on the next real call each rep runs. | 0 if no change, 2 if the miss is gone on most calls |
Keep the behavior definition fixed for the full two weeks. Record the score, the evidence (a call link or a session link) and the person who scored it. At the end, decide: keep Gong only, add a roleplay tool, or retest with a different scenario.
Frequently asked questions
What coaching features does Gong have?
Gong's coaching features, as described in its help center in October 2026, are call scorecards, AI Call Reviewer (suggested or automatic scoring), timestamped comments, feedback requests, a coaching inbox for managers, coaching metrics that track manager activity, and two practice features: AI Trainer roleplay and Dry Run rehearsal for upcoming meetings. Availability depends on the company's plan and each user's seat.
Does Gong have AI roleplay?
Yes. Gong's AI Trainer runs timed, spoken roleplay with an AI customer persona that admins build from a written scenario and, optionally, linked Gong calls. Sessions last 5 to 30 minutes, can be retried before submission, and can be scored with the same scorecards used on real calls. Gong's help center lists AI Trainer on its full enablement plan.
Does Gong offer live coaching during calls?
We found no Gong product page or help article in October 2026 that describes prompts shown to a rep during a live customer call. Gong's coaching pages describe review after the call and practice before it, including Dry Run rehearsal for upcoming meetings. If live in-call cues are a requirement, ask Gong to show the feature on a real call before you buy.
What is the Gong coaching inbox?
The coaching inbox is a manager view in Gong that shows recent coaching activity across the team, when the manager last gave feedback to each rep, and any open feedback requests. Gong's help center suggests reviewing it once a week. It sits under Coaching in the Gong app.
How much does Gong coaching cost?
Gong does not publish prices. Its pricing page, checked October 2026, says licenses are priced per user and that there is a base fee tied to the number of users, and it asks buyers to request a custom proposal by team size. The pricing page does not say which coaching features each plan includes, so ask for a written plan breakdown.
Can Gong scorecards grade practice calls?
Yes, for AI Trainer sessions. Gong's help center says AI Trainer sessions are graded by AI Call Reviewer with the same scorecards applied to live calls, and the rep sees an overall score and a breakdown when a scorecard is attached. Without a scorecard, the rep still gets AI coach feedback but no score.
Do you need a separate roleplay tool if you already use Gong?
Not always. If you are on Gong's full enablement plan and your practice needs are one buyer per scenario built from Gong calls, AI Trainer may cover them. Teams add a separate roleplay tool when they need committee calls, screen-share demo practice, drills that restart at one moment, or practice built from calls held outside Gong.
Decide where your coaching gap is
If the gap is visibility into real calls, Gong's review tools are built for that. If the gap is reps rehearsing the moment they keep missing, test practice on one real miss: grade the call, start practice at that line, and check the next call. Quotain's real-call grading and AI sales simulations run that loop with calls imported from Gong. Compare a roleplay specialist side by side in Quotain vs Hyperbound, or book a demo and bring one recorded call and the scorecard your managers already use.