November 1, 2026 · By Jeremy Nason
How to Tell if Sales Training Is Working for Canadian Agents
You finished the course. You got the certificate. So why does your pipeline look exactly the same? For Canadian agents, sales training is not a scoreboard of modules completed. It is a behaviour-change experiment. This post will show you how to tell whether training is working by tracking four observable indicators each month: appointment-to-close rate by stage, cases designed before delivery, discovery questions per appointment, and referral asks per week. Before you judge anything, you’ll learn how to set a 30-day baseline. You’ll also see why course completion is not a scoreboard, how to read the results with a simple decision rule, three traps that can fool you, and why the numbers, not more training, should decide your next move. That is especially true in Canada, where licensing expectations and insurer priorities keep shifting, but your clients still experience only the conversation you have with them. If none of those indicators move, the answer probably is not another course. It is applying what you already learned.
Why Course Completion Is Not a Scoreboard
CE credits, attendance records and completion certificates measure exposure to instruction. They do not measure behaviour. A certificate confirms you sat through a module; it says nothing about whether your next client appointment went differently.
Training only becomes real when a new behaviour survives contact with a client who says no, goes quiet, or changes the subject. That moment is the whole point, and no completion email captures it. The same gap applies to good ideas filed rather than used, including CRM follow-up sequences for Canadian life insurance agents: saving a sequence is not running one.
The four indicators in this piece are internal tracking measures, not industry norms. No published Canadian benchmark exists for appointment-to-close rates, discovery-question counts, referral rates or training-impact measures. Your own prior month is the honest comparison.
For scale, not a target: CLHIA represents 99% of Canada’s life and health insurers, and its October 2026 news release puts life, health and retirement benefits supplied by Canadian insurers at $157 billion. That describes the market you sell into, not a conversion rate you can hold yourself to.
For external context, LIMRA describes its research as unbiased sales and sales-practice benchmarks. Use it for orientation and check the figures directly. Don’t convert someone else’s number into your target.
Prerequisites: Set a 30-Day Baseline Before You Judge Anything
An honest comparison point requires three things set up before you measure anything.
Pull your baseline from the 30 days before the training started, not the month after. That month is the only real control you have, and it is the difference between a measurement and a memory.
Define stage in writing before you count anything. Pick four or five stages that match how you actually work: booked appointment, fact find complete, case designed, proposal presented, closed. If your stages overlap, every indicator built on them is noise.
Build one tracking sheet, a row per appointment and a column per indicator, and fill it in the same day. Two minutes per appointment. Retro-fitted numbers are guesses.
Book a fixed 30-minute monthly review on the same date each month. A review that moves whenever you are busy will not happen, and cadence is what turns single data points into a trend. If your appointments come from people you know, a consistent prospecting routine keeps those rows filled.
Accept that month one will be messy. Consistency beats precision early: a rough number you actually collect beats an exact number you never do.
Indicator 1: Appointment-to-Close Rate by Stage
Appointment-to-close rate by stage is the indicator that tells you the most. For each stage, divide the cases that moved to the next stage by the appointments that reached that stage during the calendar month. Do that stage by stage, not as one overall ratio.
An overall close rate hides where training actually landed. A sharper fact find shows up in fact-find-to-case-designed conversion first, and it can take another month to reach a close. Your headline number can look flat while behaviour inside the appointment has already changed. If your cases are stalling later, in the presentation itself, it is worth examining why cases tend to stall at the presentation stage.
Take 12 fact finds. Last month 4 became designed cases; this month 7, with appointment volume unchanged. Something behavioural changed, and the number says so. Now reverse it: volume flat and close rate flat or lower means the training never reached the client conversation. If appointment volume itself dropped, the training may have crowded out prospecting time instead.
Cadence rule: compare this month against your baseline and against last month. One up month is noise. Two consecutive months in the same direction is a signal you can act on.
Indicator 2: Cases Designed Before Delivery

This indicator tells you whether cases were built before you walked in.
A case counts as designed before delivery when the design work happened before the client appointment, not in the car afterwards. Record it as a two-value field, designed before or designed after, and report it monthly as a proportion.
Preparation separates an agent following a process from one reacting to whatever the client asks. It is the direct output case design work exists to produce, and the discovery conversation that keeps Canadian cases small has to happen first.
Moving: 3 of 11 cases designed before the meeting last month; 8 of 10 this month, client mix unchanged. The appointment is now prepared for rather than survived.
Not moving: 9 of your last 10 cases still assembled afterwards. The training produced knowledge, not a process, and more content will not fix a missing routine.
The CLHIA module Serving Clients Through Needs-Based Sales Practices was developed with Advocis, the Independent Financial Brokers of Canada and the Canadian Association of Independent Life Brokerage Agencies. Designing before you deliver reflects the kind of disciplined disclosure practice that supports responsible client-facing conduct. That standard looks like preparation in practice.
Indicator 3: Discovery Questions Per Appointment
The third indicator is discovery questions per appointment.
Count only questions you asked and the client answered at length. A question the client deflects, or one you answer yourself, does not count. Log the tally immediately after each appointment, one line per appointment, then average across the month. Recall at month end tends to inflate the number, log immediately to avoid this.
For example, if your monthly average moves from 6 to 11 questions per appointment, that is a change in behaviour, treat these numbers as illustrative; your own baseline sets the real target. That kind of shift tends to reach Indicator 1 a month or two later. An average that stays flat while you complete more training says you are consuming content rather than applying it. The knowledge is not the missing piece; the repetition is. A consultative selling approach is a skill you rehearse, not a module you finish.
Quality guard: 20 rapid-fire recall questions are not 20 discovery questions. Flag appointments where you asked a lot and still left without the client’s real concern on the table.
Indicator 4: Referral Asks Per Week
The fourth indicator is referral asks per week.
What counts as an ask. An ask is an explicit, audible request for an introduction. “Let me know if you know anyone” does not count, because it is a statement of availability, not a request.
Count asks, not referrals. Tally asks per week. You control the ask; you do not control the answer. The ask is the behaviour training is meant to change, so measuring referrals received turns a behaviour measure into a mood measure.
Set the target from your own baseline. Choose a number at baseline, say five asks a week, and record the actual beside it. A target pulled from your own baseline is defensible; one pulled from a stranger’s claim is not.
Read the two numbers together. Asks flat means the behaviour did not change. Asks rising while referrals stay flat points at how you ask, not whether you ask, which is a different coaching problem.
Asking is one way in; inbound lead generation that makes prospects come to you is another.
Give it three months. Do not let one low-referral month stop the asking. Referral timing lags the ask by weeks or months, so three months of asking data is the minimum useful read. The decision rule comes next.
Reading the Results: The Decision Rule
Read all four indicators against one decision rule.
One indicator moved, same direction, two months running. The training is doing something. Keep going and keep tracking; watch whether the other three follow, usually a month or two later.
None of the four moved while appointment volume held steady. More training is unlikely to be the answer. Treat this as a working rule rather than a published finding. It follows from how behaviour change works, and it is a recommendation, not a sourced result.
None moved and appointments fell. Look at capacity before content. Time in training displaced prospecting, and the fix is scheduling, not another module.
Now separate instruction problems from application problems. Instruction problems sound like “I do not know how to do this.” Application problems sound like “I know how, and I did not do it.” Only the first responds to new content.
Application problems respond to repeated, supervised practice: group coaching, rehearsal, case design work and feedback on real appointments. Format matters as much as content, so understand the structural difference between group and one-on-one training before you book anything.
One timing note: agents working through MGAs in Ontario may see licensing expectations shift as regulatory oversight of the MGA channel develops. Under changing licensing expectations, evidence of how you actually practise is more useful, not less, so keep the tracking running.
Three Traps That Will Fool You
The decision rule only works if the data underneath it is clean. Four traps can still fool you.
Attribution trap. If you changed your lead source, market, hours or assistant in the same month, don’t credit training for the result. Change one variable at a time, or note both changes so the month is flagged as mixed.
Small-sample trap. At 10 to 15 appointments a month, one large case can swing your close rate. Read every indicator as a proportion and a trend rather than an absolute, and resist making decisions on a single month.
Seasonal trap. Canadian appointment volume bends around the winter holidays and summer, and many agents see a shift during RRSP filing season, note the period beside each review date rather than assuming the pattern is universal. Seasonal patterns in Canadian financial services broadly follow this shape, though your own calendar may differ.
Consistency trap. Missing two weeks of logging corrupts the whole month. If you fall behind, log what you remember, mark it incomplete, and resume. Don’t reconstruct plausible numbers.
Finally, write one line a month about what you changed and what you observed. Twelve lines of context make a year of indicators interpretable. Without them, you’re guessing at your own history.

Conclusion: Let the Numbers Decide
Strip out the traps and what’s left is a simple discipline. Track four numbers monthly: appointment-to-close by stage, cases designed before delivery, discovery questions per appointment, and referral asks per week. Define each in writing, then compare them against your own 30-day baseline.
Give it two consecutive months before you call it. If indicators are moving, the training is reaching the client conversation. Keep going and watch the laggards.
If nothing moved, the gap is application, practice with feedback, not more content.
These four measures are internal tracking tools, your baseline is the only fair comparison you have.