A sales manager opens a library of recorded calls. Three reps keep missing the same cue: a buyer says "we're already looking at someone else" and the rep moves straight on to pricing. The conversation intelligence tool tags the moments as "competitor mention" on two of the calls and "objection" on the third. The manager wonders whether these are the same gap, or whether the labels are simply inconsistent.
This is an illustrative scenario, not a documented customer case. It captures the problem this guide addresses. Conversation intelligence tools, such as Gong, can surface a lot of call data. Turning it into the right coaching is a separate job. This guide is for heads of sales enablement and sales managers. It explains how to use call evidence to identify skill gaps across Bloom's levels, how to check the automated labels before acting on them, and how to turn what you find into targeted practice. It does not describe any integration with a specific product: no integration with Gong or any other conversation intelligence product is claimed, and everything here works with whatever call data you already have. Before using any call data, confirm consent, data handling and the rules in your region with your legal team.
What is sales conversation skill analysis?
It is the practice of reviewing sales conversations, whether recorded calls or transcripts, against a rubric to find which specific skills a rep needs to build, and then matching the coaching to the gap. Conversation intelligence platforms can transcribe calls, tag topics and surface patterns at scale. The tags tell you where to look. They do not tell you why a rep missed a cue or what kind of help would fix it. A person with a rubric still has to judge that.
빠른 사실
| 문제 | 짧은 답변 |
|---|---|
| 그 임무는 무엇인가요? | Translate call evidence into targeted coaching needs. |
| What does the software give you? | Transcripts, tags and patterns that show where to look. |
| What does it not give you? | A verified reason for a missed cue, or the right coaching. |
| How do you map gaps to Bloom levels? | Ask whether the rep misremembered, misunderstood, misapplied, misread or misjudged. |
| What must you check first? | Recording consent, data handling and the accuracy of the labels. |
| 무엇을 측정하시나요? | Accuracy of gap identification and coaching relevance on a new case. |
| What can analysis not fix? | A weak product-market fit, poor leads or a pipeline process that rewards the wrong behaviour. |
1. 성능 문제를 설명하세요
Start with the task the manager must do: translate call evidence into a coaching need a rep can act on. "Needs to improve discovery" is not a coaching need. "Does not ask a follow-up question when the buyer mentions a competitor" is. Write that as a one-line standard the team can check, and agree it with the sales enablement lead.
2. Check the data before you diagnose
Automated tags are a starting point. Before you act on them:
- Sample and compare. Take 15 to 20 calls and have two people tag the same moments against your rubric without looking at the tool's labels. Compare. Where the tool and the people disagree, find out whether the tool's category is too broad or your rubric is ambiguous.
- Check consent and handling. Rules on recording differ by place. In the United States, the Justia 50-state survey on recording phone calls and conversations is a starting point for how states differ, and calls across regions can involve several sets of rules. Your legal team, not this article, decides what is allowed.
- 맥락을 찾아보세요. A missed cue can follow from a rushed call slot, a poor lead or a deal stage that is not the rep's call. Check staffing, tools and process before you assume a skill gap.
3. Map call evidence to Bloom levels
Different gaps need different help, and Bloom's revised taxonomy is a handy way to tell them apart. Our guide to 학습 활동을 블룸의 분류 체계에 맞추기 covers it in more detail. For sales calls, the objectives that matter most are apply, analyse and evaluate.
| Pattern in the calls | Likely level | 그것이 암시하는 것 |
|---|---|---|
| Rep cannot name the competitor's main claim | Remember or understand | Short self-paced input |
| Rep knows the framework but skips it live | 신청 | Practice in realistic role-plays |
| Rep hears the cue but does not link it to the buyer's goal | 분석 | Case review with comparison of calls |
| Rep makes a poor call on when to push or pause | 평가 | Coaching on judgement with a manager |
This is a proposed mapping to test, not a validated model. A single call rarely proves a level, so look for the same pattern across several.
4. 설계 준비 및 실습
Match the help to the gap.
필수 조건 확인. Use a short diagnostic to confirm the rep knows the framework and the competitor facts. A recall check confirms prerequisites only.
Branching conversational scenarios. For apply and analyse gaps, run scenarios where the rep chooses a response to a buyer line and sees what the buyer says next. The scenario can be built from a pattern in your own calls, with names and details removed.
피드백. Tie feedback to the decision the rep made. "You moved to pricing before asking what the buyer meant by 'looking at someone else'" is more useful than "improve discovery". Our guide to 유용한 피드백을 제공하는 상호작용적 연습 설계 이것을 다룹니다.
실시간 입력 방식인가요, 아니면 비동기식 검토 방식인가요? Live coaching is worth the time for evaluate-level gaps, where a manager can explain how they would read the situation. Asynchronous review works for comparing a rep's call with an annotated example. For ideas on call-based practice, see our guide to 고객 발굴 통화 감사 연구소 sales objection handling training.
5. 현실적인 예를 통해 살펴보세요
예시입니다. 영업 지원 담당자는 사용 전에 사례와 답변 기준을 검증해야 합니다.
경우. Recorded calls show a recurring missed cue and inconsistent automated labels. In each, the buyer says they are "already looking at someone else". The tool labels one call "competitor mention", one "objection" and one nothing at all.
첫 시도입니다. The manager reads the labels and decides the team has an objection-handling gap, so books a session on objection scripts.
피드백 후 다시 시도하세요. An enablement lead asks, "What did the rep do after the cue in each call?" Reviewing the three transcripts against the rubric, the manager finds that in all three the rep acknowledged the comment and moved straight to features. Nobody asked what the buyer was comparing.
좀 더 나은 시도였습니다. "The gap is not objection scripts. It is a missing follow-up question when a buyer mentions an alternative, which is an apply gap, because the reps know to ask discovery questions. I will run a scenario where the rep chooses a response to that line, and I will review five more calls to confirm the pattern before I coach anyone. I will also ask the tool owner why the same line gets three different labels."
각 답변이 타당한지 또는 불완전한지 판단하는 이유. The first attempt relies on labels that were not checked, so it may be solving the wrong problem. The stronger attempt reads the evidence, names the level, plans a check and flags the data issue. Judge it against agreed criteria.
6. 역량을 평가하고 이전 계획을 수립합니다.
Measure the manager's analysis as well as the rep's skill. Use two measures on a new set of calls: the accuracy of gap identification (does the manager's diagnosis match an expert's review of the same calls?) and coaching relevance (does the coaching address the gap that was found?). A simple 0 to 2 rubric:
| 표준 | 관찰 가능한 증거 |
|---|---|
| 작업 실행 | Names a specific gap from call evidence and records the examples. |
| 추리 | Explains the alternative readings, the level of the gap and what is uncertain. |
| 경계 | Uses approved data and handling rules, and seeks help when information is missing. |
Anchors: 0 is absent or unsupported, 1 is partial with relevant omissions, and 2 is complete and justified against the agreed criteria. Define critical errors separately, such as using call data a rep did not consent to share. This rubric is a proposal, not a validated standard. Then follow up: after coaching, review a new sample of the rep's calls for the specific behaviour, not for the overall win rate.
상호작용적 전달이 도움이 되는 경우
The scenario practice and the manager calibration both suit interactive formats. In a live session, show a call excerpt, ask every manager to type the gap they see and its level, then reveal the spread before discussing. Differences show where your rubric is unclear. The same excerpts can run self-paced for new managers, with a live session for the ambiguous ones. AhaSlides supports interactive learning and moving between live and self-paced delivery. It is a presentation and learning tool, not a call recording or analysis platform, and no integration with Gong or any other conversation intelligence product is claimed. Check any specific authoring, feedback or reporting function you plan to rely on against your own workflow. For help choosing formats, see our guide to 적응형 학습, 상호작용형 학습, 실시간 학습은 언제 사용해야 할까요?.
변화된 것을 측정하세요
Decide in advance what would show the analysis is working. Compare diagnosis accuracy and coaching relevance on an unseen set of calls with a baseline, with a defined group, denominator and timing. If rep behaviour on calls improves, check what else changed, such as lead quality, pricing or a new product, before crediting the coaching. Avoid using win rate alone, since many factors affect it.
남아 있을 수 있는 제약 조건
- 라벨 정확도. Automated tags can be wrong or too broad. Sample and verify.
- Consent and privacy. Rules vary and may limit what you can use.
- 표본 크기. A few calls do not prove a pattern. Look across reps and weeks.
- 문맥. Deal stage, lead quality and pricing affect conversations as much as skill.
- Manager time. Careful review takes hours. Focus on one gap at a time.
자주하는 질문
Can conversation intelligence data identify skill gaps automatically?
It can point to calls and patterns worth reviewing, but a person with a rubric needs to confirm the gap and its cause before coaching.
How do you map sales skill gaps to Bloom levels?
Ask what the rep failed at: recalling a fact, applying a framework live, analysing a buyer cue or judging when to push. Each suggests a different kind of practice.
가요 AhaSlides work with Gong?
No integration is claimed. You can use any call evidence you already have to design practice and run it in AhaSlides as a live or self-paced session.
How many calls should you review?
Enough to see the same pattern across several reps and weeks. A small sample is a prompt to look further, not a conclusion.
어디서 시작하나요
Pick one recurring moment in your calls, such as how reps respond when a buyer mentions a competitor. Check consent first. Have two people tag 15 calls against a one-line standard, compare with the tool's labels, name the level of the gap and run one practice scenario before you coach at scale.
To see how a coaching scenario could run live with your own team, and be reused self-paced afterwards, explore AhaSlides.
