Thursday, and the phones are only half as busy as they should be
The index is down nineteen percent over five weeks. The firm has sent a market commentary, a fund manager letter and a webinar invitation. Eleven clients have rung in.
The book has two hundred and forty.
The advisor spends Thursday on the eleven. They're anxious, they're engaged, and three of them are quite demanding — and every hour spent on them is an hour not spent on the two hundred and twenty-nine who said nothing, opened the app four times a day, and are currently forming a view about whether this relationship is worth what it costs.
Nine months later, when the outflows are reviewed, the departures will not mostly be from the group who called.
The failure is silence, not panic
Almost all financial advisor client communication training for market stress is built around the inbound call: the client who rings up wanting to sell, and how to talk them out of it. That conversation matters and it's well covered.
It's also not where the book is lost.
Four things make the proactive call its own scenario, and a harder one.
You're initiating. Nobody has asked for this call. There's no complaint to respond to, no question to answer, and the advisor has to open a conversation about a subject the client may be actively avoiding.
You have no new information. The market fell. You don't know when it stops. Everything you could send in writing has been sent. The call cannot be justified by content, which is precisely why advisors don't make it.
The client's response is uninformative. "Yeah, no, I'm fine, thanks for ringing" comes from the person who is fine and from the person with a meeting booked elsewhere on Tuesday. Reading past it is the actual skill.
And it's a volume problem. Two hundred conversations in a week means six minutes each, which changes the structure entirely. This is not a review meeting; treating it as one is how advisors get through thirty and stop.
Triage, because you cannot call everyone first
The order is the decision, and most firms make it by account size, which is the wrong axis on its own.
Proximity to drawing. Someone two years from retirement, or already drawing, is in a different situation from someone twenty-five years out — not because the arithmetic is worse but because the fear is legitimate.
Recency of the relationship. A client of eight months has no experience of you in a bad market, and nothing to weigh this against.
Whether this has happened to them before, and how it went. Anyone who sold at the bottom once is at risk of doing it again, and they usually told you about it at onboarding.
And the ones who went quiet. Reduced contact, unanswered emails, a cancelled review. Disengagement before a drawdown is the strongest single signal in the book, and it's usually sitting in a CRM nobody has queried.
Portfolio value tells you what an exit costs. It doesn't tell you who's leaving.
The six-minute call
1. Open by naming it. "I'm ringing because markets have had a rough few weeks and I didn't want you hearing from me only when things are calm." No preamble about the weather. The reason for the call is the call.
2. Say the number before they do. "You're down about fourteen percent since the start of August." They've already looked. An advisor who doesn't say it out loud looks like someone hoping it won't come up.
3. Do not forecast. "It'll bounce back" is a prediction you cannot make, and when the next leg down arrives the client remembers who said it. "I don't know when it turns" costs nothing and is the only sentence in this call you can be certain is true.
4. Connect it to their thing, in their words. Not to a risk profile — to the boat, the school fees, the date in 2031. "This doesn't change the 2031 plan, and here's why" is a different sentence from "you're a balanced investor with a long horizon."
5. Ask one question, then stop talking. "What's the bit that's bothering you most?" Then silence. The commonest failure on this call is an advisor who fills six minutes with reassurance and learns nothing.
6. Listen for the thing underneath. It is often not the portfolio. It's a bonus that isn't coming, a partner who's asking questions, a parent needing care. The market is the occasion for the anxiety rather than the cause, and the advisor who finds that out is the one who keeps the client.
7. Close with a specific next contact. "I'll ring you again in three weeks whatever happens." Then do it. The promise kept in a calm week is what makes the next bad one survivable.
The tell to listen for
Brevity from someone who is normally chatty. Politeness where there is usually banter. "No, all good" delivered slightly too fast.
That's the call to extend rather than end — one more question, gently: "Can I ask — has it crossed your mind to change anything?" Direct, a little uncomfortable, and it surfaces in thirty seconds what would otherwise surface as a transfer request in March.
Four ways it goes wrong
The non-caller, who handles the inbound queue impeccably and never rings the quiet two hundred. The most common failure by a wide margin.
The reassurer, whose entire call is "it always comes back" — unprovable, faintly dismissive, and it teaches the client not to raise things.
The data-briefer, who delivers five minutes of market analysis to someone who wanted to be asked how they were.
The forecaster, who offers a view on timing to sound useful and is held to it for years.
Why this isn't trained
Training is built around advice, not contact. Planning, suitability, product, tax. The six-minute call with no new information isn't a technical competency, so it has no owner and no curriculum.
Firms substitute written commentary. Market updates go out and are counted as communication. They're broadcast, they're read by a minority, and they are not the thing that retains a relationship.
Nobody rehearses calling someone who hasn't complained. All the practice material assumes an inbound problem. The proactive call is harder and gets no attention.
And peer role play gives you an engaged counterpart. A colleague playing a worried client engages, explains, and gives the advisor something to work with. The real difficulty is the client who says "fine, thanks" and means something else entirely — a register a helpful colleague will not produce.
What downturn communication training can rehearse
A simulation can hold the flat, polite, minimally informative client — the one who reveals nothing on the first pass and something real on the third — and run the same advisor through twenty of them in an afternoon, which is the volume condition that makes this hard and which no role play can reproduce. Foretell AI provides the counterparty configuration, transcripts and rubric-based scoring; the investment views, suitability requirements and client data stay with the firm.
Four to build:
- The quiet one, who says everything’s fine and is already in conversation with another firm.
- The near-retiree, whose fear is proportionate and who should not be talked out of it, only talked through it.
- The one who’s done this before, who sold in a previous fall and remembers it as the sensible decision.
- The one whose real problem isn’t the market, where a redundancy or a family situation is doing the work.
Design caution. Scenarios must not contain investment recommendations, performance projections or product-specific advice. The exercise rehearses communication only; suitability, advice standards and all regulatory obligations remain the firm's, and vary by jurisdiction.
Designing the module
Pass one — the opening. Score whether the advisor named the reason for the call, stated the position before the client did, and avoided a forecast.
Pass two — the question. Score whether a question was asked, how long the advisor was silent afterwards, and whether the answer was pursued.
Pass three — the flat response. Score whether the advisor extended the call when the client's register suggested something unsaid.
Rubric on observable behavior: Was the loss stated by the advisor first? Was any prediction about timing made? Was the plan referenced in the client's own terms? Was an open question asked? Advisor talk-time as a share of the call. Was a specific next contact date given?
Talk-time share is the measure to build the module around. It's extractable from any recording, it correlates with everything else that matters on this call, and advisors are consistently surprised by their own number.
The operator case
The audit is sitting in your CRM. Match client departures in the twelve months after a drawdown against contact records from the drawdown weeks. Firms that run this usually find that outbound contact, not portfolio outcome, separates the leavers from the stayers — and it's a query, not a project.
The call list should exist before it's needed. Proximity to drawing, tenure, prior behaviour in a fall, recent disengagement. Building that list in the third week of a correction is too late; it's a standing report.
Written commentary is not contact and shouldn't be counted as it. Most firms measure distribution. Almost none measure conversations, which is the thing that predicts retention.
And the capacity constraint is solvable. Two hundred six-minute calls is thirty hours across a team — achievable with an agreed structure, impossible if everyone improvises a twenty-minute review.
For advisory programmes, this is a useful correction to how client management is taught: the relationship is not principally maintained in the annual review, it's maintained in the unprompted call in a bad month, and that call has a structure nobody writes down.
Frequently asked questions
Should advisers call clients during a market downturn? Yes, proactively and in a deliberate order — the clients who leave after a fall are frequently the ones who never rang in and heard nothing beyond a market commentary.
What should you say to a client in a market crash? Name the reason for the call, state the loss before they do, say plainly that you don't know when it turns, connect it to their actual plan, ask one open question and let the silence run.
How do you know which clients are at risk of leaving? Proximity to drawing, short tenure, a history of selling in a previous fall, and recent disengagement — usually visible in the CRM before the drawdown starts.
What shouldn't an adviser say during a downturn? Anything predictive about timing, and anything that reads as reassurance in place of engagement. "It always comes back" is unprovable and it teaches clients not to raise concerns with you.
The short version
Eleven people rang. Two hundred and twenty-nine didn't, and some of them are the ones who go.
The call takes six minutes: here's why I'm ringing, here's your number, I don't know when it turns, here's what it means for 2031, what's bothering you most — and then say nothing for as long as it takes.
The ones who sound completely fine in a nineteen percent fall are the calls to extend, not the ones to get off quickly.
Foretell AI lets wealth firms build conversational simulations — including proactive downturn calls, disengaged-client conversations and retention scenarios like the one above — with configurable counterparties, transcripts, recordings, and rubric-based evaluation. If your retention analysis has never been matched against who actually got called, we're happy to walk through how other firms have structured it.