Your agents almost never ask for the next step
Across a single window of 7,715 insurance calls, read end to end rather than sampled, the clearest revenue pattern was not a bad pitch or a hard objection. On 6,955 of the 7,020 calls the check could score, the agent never asked for the next step.
We analysed a single window of 7,715 insurance phone calls end to end, every call, not a sample, for one anonymised Australian life-insurance business that runs Metricsense across its call floor.
Read every call in a window rather than a sample of them, and patterns appear that no sample can show. Here is the clearest one, and it is not the one most sales leaders would guess.
The reframe: in phone-based sales, the deal is rarely lost in a hard conversation. It is lost in a quiet one, the moment the agent informs, answers, and lets the customer go without ever asking for the next step.
Call it the unasked close: not a no from the customer, not a fumbled objection, just an ask that never happens. It is the most common revenue leak we see, and the easiest to miss.
What we actually found
You cannot see this in a sample. A handful of calls per agent tells you how a few went; it cannot tell you what is missing across all of them. Read the whole window and the shape appears. Across the 7,020 calls where the Recommendation Response check returned a result, 6,955 recorded no next-step ask at all: the customer was never invited to move the application forward, never prompted for a recommendation. On the 26 calls where an agent did ask directly, customers mostly did not say no. Nobody was asking the question.
calls where the Recommendation Response check recorded no next-step ask, which is 99.1% of the calls it could score. In a business where the phone call is the sale, that is a revenue gap, not an objection problem.
Why the ask goes missing
It is not laziness. On a regulated call, agents are trained to be careful: do not overstep, do not pressure, do not say the wrong thing. That instinct is right, and it quietly turns into asking for nothing at all. The safest-feeling call is the one where you inform, resolve, and let the customer go. It is also the one that does not convert. At scale, the same careful restraint ends thousands of calls just before the ask.
And the fix is not pressure, which is the opposite of a pushy-sales message. Asking a customer once, clearly, whether they would like to take the next step is neither a breach nor a hard sell. It is the difference between a call that helped and a call that also converted.
An absence is invisible one call at a time
A small QA sample could never surface this. Ask a reviewer to listen to a few calls per agent and they will grade tone, accuracy and compliance on the calls they hear. What they cannot grade is an absence: something that never happened, repeated quietly across the calls they did not hear. The unasked close is invisible on any single call. It only shows up as a pattern once you have read them all.
You cannot spot a missing question in a sample. An absence only becomes a pattern once you have read everything.
Why an absence is the cheapest problem you have
An absence is also the cheapest gap to close. You do not need better objection handling or a new script; you need the question asked. And because every finding is pinned to the exact call, the fix is specific instead of a floor-wide lecture: here are the calls that ended without an ask, here is where it should have come, here is the coaching conversation to have. It is why one pass over the calls pays twice: the audit a compliance team runs for risk hands the revenue team its clearest conversion gap, for free.
The honest limits
Three caveats, because an honest number should not become a sales pitch. First, a not-asked call is not always a missed sale. Plenty were pure service, or customers who were not ready, and forcing an ask on them would be worse than the silence. The signal is the scale of the pattern, not any single call. Second, we cannot tell you those unasked calls became lost sales; proving that needs your own outcome data joined to the call content. We can show you the gap. What it is worth is yours to measure. Third, 6,955 is the model's count, not a validated estimate: the check is a language model applied to a definition we wrote, and its precision and recall are not yet published.
One lesson from the calls
If reading that many calls taught us one thing, it is that the lost sale is rarely a dramatic no. It is a good, careful, compliant conversation that ended just before the ask. You will not catch it by listening harder to a handful of calls. You catch it by reading all of them, and counting the question that never came.
Head of Product, Metricsense
Gaurav Soni leads product for Metricsense. He has spent 13 years in product at the intersection of data, analytics and AI, starting in engineering, building APIs, data pipelines and analytics products, and moving steadily closer to the decisions leaders actually make. The thread through all of it is the same: turning messy, unstructured conversation into evidence a team can act on. At Metricsense that means reading 100% of a company's calls, reviews and tickets, and linking every finding to the exact quote.
Find the ask your calls are missing
Metricsense reads a week of your own calls, in your own cloud, and shows you where the next-step ask never came, the exact calls and the exact moment, right alongside the compliance picture, usually within a day.
Keep reading
Is it an adviser problem, or is the script teaching everyone the same mistake?
When the best adviser on the floor still gets it partly wrong on a quarter of explanations, you do not have a coaching problem. You have a wording problem.
The call that becomes a complaint is the call nobody chose to review
QA samples are not random. They are convenient. And convenience selects against exactly the calls that create conduct risk.
Your QA score only measures the calls you chose to hear
A 93% QA score is a fact about the calls your team selected. It is not a fact about your call floor, and ASIC has now said so in writing.