Delivering a Decision You Disagree With

"He's said no to the boiler"

The tenant has reported the heating three times in five weeks. The engineer's report recommends replacement. The landlord has authorised a repair instead, for the second time, because it's cheaper — and the agent thinks he's wrong.

She now has to ring a tenant who has been cold since October and deliver a decision she'd have made differently.

Two responses come naturally, and both are wrong.

"I've pushed and pushed, honestly, he just won't spend the money." Disloyal to the client, and worse, useless — it tells the tenant that going around her is the only route left.

"The engineer thinks a repair will hold it." Dishonest, and she'll be quoted on it in three weeks when it doesn't.

Nobody has ever told her that a third option exists.

The intermediary's problem, named

This is the closing scenario in the series and it generalises the thread running through several of them: delivering a landlord's decision — or a credit committee's, or an insurer's, or a family client's — to someone who will hold you personally responsible for it.

Four features define it.

The professional has no authority and complete ownership. The decision belongs elsewhere. The relationship, the delivery and the consequences belong to her.

She disagrees, and the counterparty can tell. People are good at detecting a professional delivering something they don't believe. The question isn't whether it shows; it's what she does with it.

The two obvious moves are both damaging. Disowning the principal destroys the agency's position and invites the tenant to bypass her. Feigning agreement destroys her own credibility the moment the decision turns out badly.

And nobody has defined what she may say about her own view. Firms tell staff to represent the client. They almost never say what an honest professional may express about a decision they think is poor, so everyone improvises at the two extremes.

First: is this a preference or a problem?

Before anything else, a categorisation that changes the whole response.

A preference is a commercial decision the client is entitled to make. Repair rather than replace, a cheaper contractor, a slower timescale. The agent may disagree; it's the client's money and the client's call, and delivering it is the job.

A problem is a decision that may cross a legal or safety line. Those are not conversations to manage — they escalate internally, immediately, through the firm's process and to its legal advice, and they are not things an individual should be negotiating with either party.

The single most important skill here is telling the two apart quickly, and it's the one agents are least equipped for because nobody has written down where the line sits.

Then: the third option

For a genuine preference, there is a position between disowning and pretending, and it's available to anyone once they've been told it exists.

Deliver the decision, not the debate. "The landlord has decided to repair rather than replace it, and that's his call to make." Clear, attributed, no editorial.

Don't claim agreement you don't have. You are not required to say you think it's right. Omission is honest; endorsement is not.

Be neutral rather than warm or cold about it. Neutrality is a real option and most people have never considered it one. "That's the decision" delivered evenly is neither betrayal nor performance.

Say what you can do. Chase the timescale, get it in the diary, escalate if it fails again, review it at the next report. Concrete, within her authority, and it's the part that actually helps.

Commit to the review point. "If this doesn't hold, I'll go back to him with that." That's the honest version of "I'll push for you," and unlike the other version it's deliverable.

And if asked directly what you think — "do you think a repair is going to fix it?" — the answer is bounded and honest: "I'm not the engineer and it's not my decision. What I can tell you is what I'll do if it fails again." Not a dodge, and not a disclosure that undermines the client.

Then: report the consequence back

This is the half that almost never happens, and it's where the agent's disagreement is actually useful.

Not "the tenant's upset" — that's moral and it will be discounted. Commercial: this is the third call-out, the cumulative cost is approaching the replacement, a tenant of four years is asking about ending the tenancy, and a void plus re-letting would cost more than the boiler.

Landlords make short-term decisions because the long-term arithmetic has never been put in front of them. Putting it there in writing is the agent's actual influence, and it's a great deal more effective than anything she could say to the tenant.

And keep a record of what you advised. It protects the agency and it changes the client's decision-making over time.

Four ways it goes wrong

The disowner, who tells the tenant the landlord won't spend the money. Feels like solidarity, destroys the agency's position and invites escalation around her.

The fake-agreer, who defends a decision she thinks is wrong and is held to it when it fails.

The silent absorber, who delivers the decision competently and never tells the landlord what it cost — so the same decision is made again next quarter.

The boundary-misser, who treats a safety or legal issue as a client preference and manages it as a conversation instead of escalating it.

Why this isn't trained

Firms train loyalty, not honesty within loyalty. Staff are told to represent the client and never told what they may say about their own view, so they choose between disloyalty and dishonesty.

The escalation line is undefined. Where a client's decision stops being a preference is the most consequential judgement in the role and it is almost never written down.

The feedback loop is nobody's job. Reporting the consequence of a client's decision back to them, commercially, is the mechanism by which decisions improve — and it isn't in anyone's objectives.

And peer role play produces acceptance. A colleague hears the decision and moves on. The real counterpart is cold, has been patient twice already, and asks the direct question — which is the moment the whole scenario turns on.

What intermediary-role training can rehearse

A simulation can hold a counterparty who asks the direct question, presses for the professional's own view, and reacts to evasion — and score whether the professional found the position between disowning and pretending. Foretell AI supplies the counterparty configuration, transcripts and rubric-based scoring; the agency's client obligations, escalation thresholds, legal position and all regulatory requirements stay with the firm.

Four to build:

Design caution. Where a decision may engage safety, habitability or legal obligations, the correct behaviour is escalation through the firm's process and legal advice, not conversational management — and modules must be built so that this is the scored outcome rather than a well-handled delivery. Obligations vary by jurisdiction and nothing here states any. Modules must use the firm's own escalation thresholds and client-care standards.

Designing the module

Pass one — the categorisation. Score whether the professional correctly identified a preference from a matter requiring escalation.

Pass two — the delivery. Score whether the decision was attributed cleanly, whether the principal was disowned, and whether agreement was falsely claimed.

Pass three — the direct question. Score the answer to "what do you think?" — bounded honesty, evasion, disloyalty or false endorsement.

Rubric on observable behavior: Was the matter correctly categorised? Was the decision attributed without editorial? Was the client disowned? Was agreement falsely claimed? Was a concrete action within authority offered? Was a review point committed to? Was the consequence reported back to the client in commercial terms?

The response to the direct question is the measure. It's a single exchange, it's where every version of this scenario resolves, and the four possible answers are easy to distinguish in a transcript.

The operator case

Define where a preference stops. The threshold between a client's commercial decision and a matter requiring escalation is the most important line in the role, and in most firms it exists only in individual judgement. Writing it down is a risk control and a training shortcut simultaneously.

Tell staff what they may say about their own view. In the absence of guidance they pick one of two bad options. Two approved sentences resolve it.

Build the consequence report into the client cycle. Cumulative repair costs, call-out frequency, tenancy risk — presented commercially. It's the only mechanism that improves client decisions, and it currently depends on individual initiative.

And absorbed frustration is an attrition problem. Staff who spend their week delivering decisions they disagree with, with no route to influence them, leave. The feedback loop is a retention measure as much as a commercial one.

For property and professional services programmes alike, this is the general case behind several scenarios in this series: the person with no decision rights determines how the decision lands, and is the only one positioned to improve the next one.

Frequently asked questions

What should you say when you disagree with a client's decision you have to deliver? Attribute it cleanly without editorial, don't claim agreement you don't have, offer what you can actually do, and commit to a review point. Neutrality is an available and honest position.

Should an agent tell a tenant the landlord refused something? Attributing the decision accurately is appropriate; characterising the landlord's motives isn't. Disowning your client invites the tenant to bypass you and damages the agency's position without helping anyone.

What if a tenant asks the agent's personal opinion? Answer within the boundary — you're not the engineer, it isn't your decision — and pivot to what you will do if the problem recurs. That's honest without undermining the client.

When should a client's decision be escalated rather than delivered? Whenever it may engage safety, habitability or legal obligations. That threshold should be written down by the firm rather than left to individual judgement in the moment.

The short version

The landlord's entitled to make a commercial decision she thinks is wrong, and she has to be the one to say it to a tenant who's been cold since October.

Work out first whether it's a preference or something that has to go up — because those aren't the same conversation at all. Then attribute the decision cleanly, don't claim to agree, offer what's actually in your gift, and commit to a review point you'll honour.

And write to the landlord with the arithmetic: three call-outs, a four-year tenant asking about leaving, and a void that costs more than the boiler. That's where her disagreement is worth something — not in the phone call, where saying it would only make things worse for everyone including the tenant.

Foretell AI lets property and professional services firms build conversational simulations — including delivering decisions made elsewhere, boundary-holding and escalation judgement like the one above — with configurable counterparties, transcripts, recordings, and rubric-based evaluation. If your people are choosing between disowning a client and defending something they don't believe, we're happy to walk through how other operators have structured it.