Insurance Agent Call Coaching: A Manager’s Playbook
Insurance agent call coaching is a weekly routine: you pull a few real recorded calls per agent, score them against a short checklist, and give each agent one behavior to fix before the next review. Done that way, it changes what agents say on the phone. Generic training days rarely do. This guide covers how to pick calls, what to score, how to run the session, where AI scoring fits, which recording rules apply and how to tell whether any of it worked.
TL;DR
- Coach from real calls, not from memory. Review a few recorded calls per agent each week, mixing wins, losses and one the agent picks.
- Score every call on the same short scorecard: opening, discovery, objection handling, next step, compliance language.
- Give one behavior to change per session. Agents can fix one thing before the next call. They cannot fix seven.
- Know your recording rules before you review anything. Consent, notices and retention differ by state and by line of business.
- Measure one observable behavior against a baseline, not gut feel.
Table of Contents
- What call coaching covers on an insurance floor
- Choose which calls to review
- Build a scorecard agents can remember
- Run the feedback session
- Coach new phone agents on openings, discovery and objections
- Use recordings, scorecards and AI scoring together
- Recording, consent and compliance rules to coach against
- Perspective: what we see on insurance call floors
- Measure whether coaching is working
- Set up the workflow in your tools
- Run coaching on RevRing.com
- FAQ
- Sources
What call coaching covers on an insurance floor
Call coaching is a manager listening to an agent’s actual conversations and giving specific, repeatable feedback. It sits between onboarding (teaching the product and the script) and performance management (hitting quota). Coaching asks a narrower question: what did this agent do on this call, and what should they do differently on the next one?
On an outbound or inbound insurance floor, coaching should improve five things:
- The opening. Identity, reason for the call, and any required notice delivered clearly in the first seconds.
- Discovery. Enough questions to understand the household, current coverage, budget and timing before quoting.
- Objection handling. Acknowledging the concern before answering it.
- The next step. A booked appointment, a submitted application or a dated callback. “I’ll think about it” is not a next step.
- Compliant language. Required disclosures, consent and honest representation of the product.
Notice what is missing: motivation speeches, pipeline reviews and generic sales theory. Those matter, but they are not call coaching. If the session does not start with a recording, it is a different meeting.
Choose which calls to review
Most managers coach the calls they happen to hear, which means the loudest agent and the most recent blowup. Pick calls on purpose.
A workable weekly sample per agent:
- One call that closed or booked. Find out what the agent did that worked, and name it so they repeat it.
- One call that died in qualification or discovery. This is where most lost revenue hides.
- One call the agent chooses. It shows what they think is hard, and it makes the session feel like a conversation rather than an audit.
Three calls a week per agent is a manageable minimum for a manager with a team of eight to ten. On a larger floor, team leads take a pod each, and the sales manager reviews only the team leads’ coaching quality and a sample of calls.
Skip calls under about 45 seconds unless the pattern is a problem in itself. Many short calls are hangups at the opening, which is a coachable issue. A few are wrong numbers.
For new hires, shift the mix. Review more calls and shorter calls, early and often. In the first weeks, listen for openings and the transition into discovery, because those are the moments new phone agents stumble most.
Build a scorecard agents can remember
A scorecard with 30 line items gets ignored. Build one with five to eight items and clear pass or fail definitions.
| Area | What “good” sounds like | Typical miss |
|---|---|---|
| Opening | Name, agency, reason for call and any required notice in the first 15 to 20 seconds | Rushing the notice or burying it |
| Discovery | Asks about current coverage, household, budget, timeline before pitching | Quoting after two questions |
| Listening | Reflects the prospect’s words back before moving on | Talking over the prospect |
| Objections | Acknowledges, asks a clarifying question, then responds | Arguing or dropping price immediately |
| Next step | Specific date, time, and what happens then | Vague “I’ll follow up” |
| Compliance | Required statements and consent handled as your agency defines them | Skipping disclosures when the call gets friendly |
Write each row with one example of a pass and one example of a fail, taken from your own recordings. Agents learn faster from a call they recognize than from a definition.
Keep the scorecard separate from the compensation plan. If agents believe a low score costs them money, they start gaming the calls they know will be reviewed.
Run the feedback session
A 20 to 30 minute one-on-one works for most agents. Structure it the same way every week so agents know what to expect.
- Start with the agent’s pick. Ask what they heard in their own call before you say anything. Agents often name the problem themselves.
- Play the exact moment. Cue up the 20 to 30 seconds in question. Do not replay the whole call.
- Name the behavior, not the person. “You moved to the quote before asking about current coverage” is feedback. “You rush” is a label.
- Practice it right there. Have the agent say the better version out loud, or role-play the objection with you. Practice is the part most coaching skips, and it is the part that sticks.
- Set one target. Write down one behavior and how you will check it next week.
- Close with what worked. Specific praise tied to a call is worth more than general encouragement.
Coaching skills that matter most
The manager skills that make these sessions useful are simple and learnable:
- Listening before talking. Let the agent diagnose first.
- Asking open questions. “What was the prospect worried about there?” beats “Why did you do that?”
- Being specific. Timestamp and quote.
- Staying consistent. Same day, same format, same scorecard.
- Following up. Next week’s first call reviewed is the one that tests last week’s target.
Coach new phone agents on openings, discovery and objections
Newly licensed agents usually know the product better than they know how to run a conversation. Coach in the order the call unfolds.
Openings. Give the agent one opening that fits your lead type, whether that is an inbound web lead, a purchased live transfer or an outbound dial to a past client. Have them say it until it sounds like speech. Then review live openings, not just roleplay, because real prospects interrupt.
Discovery. Give a short list of must-ask questions in a sensible order and let the agent adapt the wording. Coach the pause after each answer. Agents who fill silence tend to answer their own questions.
Objections. Pick the three or four objections your floor hears most, such as “I’m already covered,” “send me something in the mail,” “I need to talk to my spouse” and “what does it cost?” Pull real recordings of each, and build a short library of good responses from your own best agents. A response that works in your market, in your agents’ voices, beats a script from a book.
Daily practice. Ten minutes of role-play at the start of a shift, rotating who plays the prospect, does more than an hour-long monthly training. Pair a new agent with a strong one for call listening for the first couple of weeks.
Use recordings, scorecards and AI scoring together
Recording gives you the raw material. Scorecards give you the standard. AI scoring gives you coverage.
A manager can honestly review a handful of calls per agent per week. On a floor making hundreds or thousands of dials, that is a small slice. AI call scoring helps by scoring every connected call against your criteria, flagging the ones worth a human listen: missed disclosures, long monologues, no next step, strong closes worth sharing. Agencies weighing this kind of tooling can read about AI productivity results for agencies to see what to expect from automation more broadly.
Use AI scores as a triage tool, not a verdict:
- Sort by lowest and highest scores, then listen to a sample of each.
- Check that the AI’s read matches your own on a dozen calls before you trust it for coaching decisions.
- Never discipline an agent on a score alone. A person confirms what happened.
Good coaching software also lets you attach a comment to a timestamp, so the agent can jump straight to the moment you mean. That saves the first five minutes of every session.
Recording, consent and compliance rules to coach against
Coaching depends on recording, and recording is regulated. General workflow advice here is not legal advice. Have your compliance officer or counsel confirm what applies to your agency, your states and your carriers.
Consent to record. Some states require every party to consent. California, for example, bars recording a confidential communication without the consent of all parties under Penal Code section 632. If your agents call across state lines, plan for the strictest state you reach. Our own guides on two-party consent states and call recording laws in the U.S. walk through the details for agencies.
Notices up front. Deliver the recording notice at the start of the call, every call, in the same words. Coach it like any other opening skill, because a rushed notice is the one that gets challenged later.
Medicare marketing and sales calls. If your agents sell Medicare Advantage or Part D, federal rules require these calls to be recorded in their entirety and retained. The current text of 42 CFR 422.2274 sets a minimum retention period of 6 years. Retention periods have changed in recent years, so confirm the current rule and each carrier’s instructions before you set your storage policy. Our overview of Medicare sales call rules covers what agencies need to follow. For Medicare calls, scorecards should also check permission to contact, Scope of Appointment, prohibited terminology and licensure.
Calling hours and consent. Telemarketing calls to consumers are generally limited to the window between 8 a.m. and 9 p.m. at the called party’s location under the FCC’s TCPA rules, and some states are stricter. Add a scorecard line for consent and do-not-call handling so coaching reinforces what your compliance process requires.
Using AI monitoring. Automated review can catch missing disclosures across every call instead of a sample. See our write-up on AI compliance monitoring for the U.S. rules regulated call teams work under.
Treat compliance misses as coaching items first, escalating to your compliance lead when a call may have crossed a legal line. Agents who fear being reported stop flagging their own mistakes.
Perspective: what we see on insurance call floors
We talk to agency owners who record every call and listen to almost none of them. The recordings pile up. Coaching never starts, because pulling the right call takes longer than the meeting it is for.
Our view is that coaching should come out of the same system as the dialing. When the dialer, the CRM, texting and the scorecard live in one place, a manager can go from a disposition to the recording to the timestamp in a couple of clicks. That is when weekly reviews stay weekly.
AI call scoring is the other half. It does not replace the manager’s ear. It tells the manager which calls deserve it. Score everything, listen to the extremes, coach one behavior at a time.
Compliance belongs in the same loop. Our platform runs DNC and litigator checks and calling-hours rules before dials, and keeps consent records and recording notices in the workflow. Those are tools to support your process, not a guarantee of legal compliance, and your counsel still owns that call.
Marc
Measure whether coaching is working
Feeling better about the floor is not a result. Pick one observable behavior and track it.
- Choose one behavior. Discovery questions per call, percentage of calls ending with a booked next step, or time to acknowledge an objection. Avoid starting with revenue, which moves slowly and has many causes.
- Set a baseline. Score two to four weeks of calls before coaching starts.
- Coach, then re-measure. Check the same behavior for four to six weeks afterward.
- Compare against a group you did not coach, if you can. On a large floor, coach one pod first and use another as a comparison. Published guidance on measuring sales coaching suggests tracking a single behavior over four to six weeks against a baseline and a matched control group, with enough reps in each group for the numbers to mean something. Most single agencies will not have those sample sizes, so treat small-team results as directional.
- Then look at outcomes. Once the behavior shifts, check quote rate, appointment show rate, application rate and cancellations or early lapses on the business written.
Write the result down. If a behavior did not move, the problem is usually that the target was too vague or the practice step got skipped, not that the agent cannot learn.
Set up the workflow in your tools
Whether you run a spreadsheet or a platform, the weekly routine needs these pieces:
- Call recordings tied to the lead record, so context is one click away.
- A call sample for each agent, pulled by outcome: booked, closed, lost in qualification.
- A scorecard template managers fill in the same way every time.
- Notes and targets stored per agent, so last week’s target is the first thing you see.
- Consent and notice records, so the recordings you coach from are ones you are allowed to hold.
- A dial approach that gives agents enough live conversations to learn from. Preview, progressive or power dialing all work. A power dialer cuts idle time between calls, which means more conversations to review per shift.
If your agency buys leads or live transfers, add lead source to your review. An agent who struggles on one source may be fine on another, and a quality problem in a lead source is not a coaching problem. Routing the right lead types to the right agents also makes coaching easier, which is covered in our lead management and routing page.
Run coaching on RevRing.com
RevRing.com puts CRM, dialing, two-way texting, AI call scoring, lead routing and compliance tooling in one platform for insurance agencies. Agencies can run on RevRing CRM or connect the CRM they already use, such as HubSpot, Pipedrive or GoHighLevel, at no extra cost.
On the published pricing page, AI call scoring is $0.20 per scored call, and outbound calls via the Power Dialer are $0.05 per minute. Seat plans run from $49.99 per seat per month for Basic Dialer to $189.99 for Pro. Enterprise has no published price. See the current RevRing.com pricing and how the platform works for insurance agencies.
If you want to see call scoring and the dialer working on your own call flow, schedule a live demo from the RevRing.com dialer page.
FAQ
How many calls should a manager review per agent each week?
Three is a practical minimum: one that closed or booked, one that failed in qualification, and one the agent picks. New hires need more, and shorter, reviews. On high-volume floors, use AI scoring to flag which calls deserve a human listen.
Can I record coaching calls in every state?
No. Some states require every party to consent, and the strictest state on a call generally sets your standard. Deliver a recording notice at the start of each call and have your counsel confirm the rules for the states you call. Medicare calls carry their own recording and retention requirements.
Should agents see their call scores?
Yes. Share the scorecard and the criteria before you start, and review scores together in the one-on-one. Agents improve faster when the standard is visible, and they trust AI scoring more when a manager explains and confirms it.
How long before coaching shows up in results?
Behavior changes first, usually within a few weeks of consistent practice. Revenue metrics lag and are noisier. Track one observable behavior for four to six weeks against your baseline, then look at quote, appointment and application rates.