Handling Questions, Objections and Dissent in a Finance Meeting

A question wants information, an objection wants a different outcome, and dissent is a recorded disagreement with a decision that has been taken. Answer a question from the record, answer an objection by naming the interest underneath it, and record dissent rather than defeating it. The same words carry all three, so telling them apart is most of the skill.

Almost everything that comes back across a table arrives in the same grammar. Is that certain. Where did that come from. But surely. Sitting underneath that identical surface are three different objects with three different mechanics, and the single most expensive habit in this whole area is treating all three as the first one. Sorting what has been handed over comes before answering it. An answer aimed at the wrong kind of challenge does not simply miss; it tells the room that the speaker could not tell what was being asked.

The Kavery research desk sits inside Kavery Capital Services Private Limited, and the work concerned Meenakshi Tubes Private Limited, a borrower on the books of Kavery Capital Services Private Limited. Sharada Iyer wrote it, seven months into the role. Prakash Nadar reviewed it. Latha Menon chaired the twenty minute item on 12 March at which it was taken through in front of six people, four of whom had read the three sheet pre-read. Twelve of those twenty minutes went on one slide. Five questions were asked inside those twelve minutes, and each one calls for a different move.

The finding on the slide was one sentence long. The operating margin had fallen from 12.0 per cent to 10.8 per cent, and the desk had an internal floor of 11 per cent, so the fall crossed a line rather than moving along one. All five questions were aimed at that single sentence, and not one of the five was answered by saying it again.

What is the difference between a question, an objection and dissent?

The other two are defined against the easiest of the three, so a question comes first. A question is a request for information. A question has a shape that can be felt from the inside: something is missing, the information is either held or it is not, and once it has been supplied the person who asked stops asking. The same thing happens outside the office. Somebody at home asks how much the plumber charged. The number is said out loud. Nobody asks again. The missing number is now present, and the exchange had a floor underneath it all along.

An objectionA push for a different outcome, said in the grammar of a question. Information was not what was short, so supplying information does not settle it. wears the same clothes and does something entirely different. An objection pushes for a different outcome. Same household: somebody asks whether it was really necessary to call the plumber at all. The sentence is not a request for the number. Information was never what they wanted. The number, the two other quotes taken at the time and a photograph of the wet floor can all be supplied, and the person will be standing precisely where they were. The objector wanted the outcome to have been different, or wants the next one to be. An objection is not a badly worded question and it is not bad faith; it is a different object entirely, and the reason good answers bounce off it is that it was never short of an answer in the first place.

DissentA disagreement with a decision that has already been taken, written into the record of that decision. Dissent asks for a line in a document rather than for a reply. is the third, and it is separated from the other two by timing rather than by content. Dissent attaches to a decision. Somebody argued for a different course, the decision went the other way, and they want their disagreement to exist inside the record of what was decided. Dissent asks for something narrow. Dissent is not asking to reopen anything and not asking for an answer; it is asking for a line in a document. The household version gets said at the front door: I still think the second quote was worth waiting for, and I would like it remembered that I said so. Nothing in that sentence needs a reply, and a reply is very often what wrecks it.

The three failures of sorting are not symmetrical, and one of them is much worse than the other two. Treating a question as an objection produces a defence of something nobody was attacking: embarrassing, and recoverable. Treating dissent as a question produces an explanation of a decision to somebody who already understood it perfectly and simply disagreed: patronising, and also recoverable. Treating an objection as a question is the expensive one. The answer will be a good one, the concern will come straight back in different words, the second answer will be good too, and the room will watch this happen two or three times. The room does not conclude that the questioner is being difficult. The room concludes that the speaker cannot hear what is being said, and inviting that conclusion is far more damaging than being wrong about a number.

How can the three be told apart?

Telling them apart requires no mind reading and no talent for reading a room. A listening skill helps enormously here, and it can be learned; it is set out under active listening. The sorting test that follows runs on the words actually said out loud, in the second and a half available before a reply is expected.

The tellThe one test that separates a question from an objection: whether supplying a straight answer would end the exchange. is this. Would a straight answer end this exchange? If a fact, said plainly, with its source attached, would cause the person to nod and the item to move on, the thing being handled is a question, and it should be answered. If that fact can be pictured as supplied perfectly with the concern still sitting there in the room afterwards, it is not a question, whatever the punctuation suggested. The test is not about tone, motive or seniority. The test asks whether information is what the person is short of, and it runs on the sentence itself without requiring anything to be known about the person who said it.

Then there is a second test, and it takes one clause. Has the decision already been taken? If it has not, what has arrived is an objection, and it is aimed at changing the outcome while the outcome can still move. If it has, what has arrived is dissent, and it is aimed at the record rather than at the outcome. The two need completely different responses, and mistaking dissent for a live objection is how people end up arguing about something that was settled a week ago.

Sorting is sometimes impossible to do in advance, and being unable to do it is normal rather than a failure of technique. When it cannot be told, the thing is answered once as though it were a question. Thirty seconds is cheap and the answer is diagnostic. If the exchange closes, it was a question and nothing has been lost. If the same concern comes back in different words, that second arrival settles it: information was not what was short. The second arrival of a concern in new clothing is the most reliable signal available, and it costs one short answer to obtain.

How to Respond to a Challenge in a Finance Meeting: What Are the Four Parts of an Answer?

Three of the five questions asked on 12 March were straightforward requests for information, and each of the three was answered in under thirty seconds. Thirty seconds sounds like a boast. Thirty seconds is a consequence. An answer takes thirty seconds when nothing in it is being invented at the moment of speaking, and all four parts had been written down days earlier, so nothing was.

The four parts, in this order. First the answer itself, said plainly and without preamble. Second the source, meaning the item number in the source log. The room can then go and look rather than take the speaker on trust. Third the confidenceHow firmly a statement is held and on what basis. Stated as a fact about the evidence rather than as a feeling about the person stating it., meaning how firmly it is held and on what basis. Fourth the falsifierThe single observation that would show a statement to be wrong. Naming it is what turns a claim into something a reader can test., meaning the one observation that would show the answer to be wrong. Answering with all four parts converts a claim into something the room can test, and a room that can test a claim does not have to decide whether to believe the person making it.

The ordering is not decoration. Putting the answer first is Barbara Minto's rule from The Pyramid Principle, 1978, usually taught for documents, and it works out loud for exactly the same reason. A listener who has the answer spends the next twenty seconds testing it. A listener who does not spends them wondering where the speaker is going. The answer comes first and the support follows it, rather than building up to it.

The third part is the one people skip, and skipping it is expensive in a way that is invisible at the time. Confidence stated plainly is information. There is a real difference between a figure that is arithmetic off two logged numbers and a figure that rests on somebody's estimate of what a distributor said, and when that difference goes unmarked, the room has to assume the two are the same. The room usually assumes the weaker one. Saying that part is arithmetic and it moves only if one of the two inputs is restated tells the room more in eleven words than another two minutes of explanation would.

The fourth part sounds like a gift to the other side of the table and it is the opposite. Naming what would change the answer does two things at once. Naming it shows the room that the speaker knows where the work is load bearing, and that is the single thing the room is most interested in finding out. The load bearing part they were about to go hunting for has already been handed over with a label on it, so the hunt loses its fun.

What is said when the answer is not known?

The fourth of the five questions had no answer available in the room. Nobody on the desk could say at that moment which document a particular series had come from. People dread that moment. The correct response is fully scripted and takes about eleven words, so it is the cheapest moment in the whole meeting.

The first move is to say that the answer is not known. The second is to name the day it will be confirmed. The third is to deliver on that day. Three moves are the entire formula, and the first two are free. Everybody manages to say they do not know and everybody manages to name a day; the third step is the only one that costs anything and it is the only one anybody remembers.

Sharada Iyer said: I do not know, I will confirm by Thursday. Two days later, on Thursday, the confirmation went out by email with the source log item number attached. Anybody who wanted to check it could do so without writing back. Four sentences. Nobody had chased it.

Now the three versions that sound identical and do nothing. I will come back on that has no day in it, so there is nothing to keep and nothing to fail. Having nothing to fail is exactly why it is the comfortable one to say. I think it is around, followed by a number, converts a known unknown into an unmarked guess sitting inside the record, and it is the worst of the three by a distance. Let me check and revert is the same as the first with more syllables. All three feel more competent in the moment than admitting the answer is not known. None of them creates anything the room can watch being kept, so all three are weaker.

How to Handle a Challenge to Financial Analysis: What Is the Room Actually Testing?

Some challenges are not aimed at the answer at all. Such a challenge is aimed at the machinery that produced the answer, and no item number settles it. The useful move here is to notice that an analysis has three layers, and that a challenge always lands on exactly one of them.

The top layer is the answer: the operating margin is 10.8 per cent, below the internal floor of 11 per cent. The middle layer is the method: operating profit taken over the revenue of the year just reported, rather than over any other year. The bottom layer is the inputs and the assumptions resting on them, and on this assignment that meant nine assumptions written into the register, of which three were carrying the conclusion. Almost every real challenge lands on the bottom layer, and almost every weak response is delivered to the top one.

Here is the concrete case. One order placed in December, worth Rs 2,96,80,000, accounted for 14 per cent of the year's revenue of Rs 21,20,00,000. Strip that order out and the margin on what is left computes at 5.0 per cent rather than 10.8 per cent. The concentration is not a small technicality hiding in a footnote; it is the largest single thing anybody could say against the work, and the risk colleague in the room was always going to say it. When the biggest challenge to an analysis is something the analyst already knows about, the only question left is whether the room hears it from the analyst or from somebody else, and those two versions of the same fact do completely different things to how the rest of the work is read.

So it was said first, on the slide, in the same breath as the finding. One order carries 14 per cent of revenue, item 5 of the source log, and without it the margin computes at 5.0 per cent. Notice what that sentence does. The sentence does not defend the 10.8 per cent, does not soften it, and does not argue that the concentration is unimportant. The sentence simply hands over the load bearing part of the work with a label on it. The risk colleague raised the concentration anyway. Both people were now looking at the same fact, so the disagreement had nowhere left to go and the exchange lasted about twenty seconds.

There is a general form here and it is worth carrying around. Under challenge, the room is not testing whether the analyst is right. The room is testing whether the analyst can tell the difference between the parts of the work that are solid and the parts that rest on an assumption. Somebody who can say that is a finding and this is an assumption, and here is which of the two the conclusion is standing on, has demonstrated the thing the room needed to know. Somebody who defends everything with equal energy has demonstrated the opposite, whether or not every word they said was true.

How is the interest underneath an objection found?

The fifth question was not a question. The relationship manager asked whether a fall of 1.2 points really justified touching a four year relationship. The tell applies. Suppose the arithmetic is supplied perfectly: 12.0 per cent falling to 10.8 per cent, below an internal floor of 11 per cent, computed off two filed numbers. Does the concern go away? The concern does not go away, and everybody in the room can feel that it does not. The fifth question took a little over ten minutes; the other four took under two minutes between them.

Getting the frame wrong is the most common way people make this worse, so the frame comes before anything else. The relationship manager is not an obstacle and is not being difficult. The relationship manager has held the account for four years, keeping the relationship is a substantial part of what their job is measured on, and they were not in the room to be persuaded of anybody's arithmetic. The relationship manager holds an honest interest that runs in a different direction from the analyst's, and both interests belong inside the same organisation on purpose. An objection almost never comes from somebody being unreasonable; it usually comes from somebody doing the other half of a job of which only one half is visible from the desk.

Underneath the sentence sat no view about 1.2 points at all. Underneath it sat an unwillingness to be the person who tells a customer of four years that their line is being cut. The unwillingness is the interestWhat the person raising an objection actually wants, underneath the words they used. Naming it turns an unanswerable exchange into a problem that can be worked on.. Name it out loud and the exchange changes character completely: a problem has appeared where previously there was only a disagreement.

Now the delicate part: how it is said. There is a version of this that is an accusation and a version that is work, and they contain the same insight. Telling the relationship manager that they are worried about losing the relationship names a motive, puts the other person in the position of denying something about themselves, and adds a second argument to the one already being made. The other version names a problem instead: it sounds like the question is who speaks to the customer and what exactly they are told. Nobody has to admit anything, the sentence is about the work rather than about the person, and it is now on the table as something two people can solve. Name the problem rather than the motive: the same insight said as an accusation costs the speaker the room, and said as work it produces an ally.

The outcome was ordinary and unremarkable, and that is the point. Two things got decided: that the relationship manager would speak to the customer rather than the desk, and that the sentence they would use would carry the review date in it. The conversation was then about a line being looked at again in 90 days rather than about a line being cut. Nothing about the analysis moved. The 10.8 per cent stayed at 10.8 per cent. The unwillingness that had been in the way is what moved.

What becomes of dissent that does not change the decision?

The dissent on this assignment did not happen in the room on 12 March at all, and that is itself instructive. Dissent attaches to the decision. Latha Menon recorded that decision on 12 March: the working capital line for Meenakshi Tubes Private Limited falls from Rs 40,00,000 to Rs 25,00,000, reviewed again in 90 days. The risk colleague argued that 90 days was too long and that it should be 30. The decision went to 90 anyway. The disagreement, and the reason for it, went into the decision record.

Recorded dissentA disagreement written into the record of a decision without changing that decision. Recording preserves the reasoning of the person who lost the argument. is not overruled dissent, and the distinction is not a courtesy. Overruled means the disagreement was heard and then disappeared. Recorded means it was heard, did not prevail, and is now a permanent part of what the decision consists of. A decision plus the recorded disagreement with it is a different object from the decision alone. The pair tells a future reader what was known, what was considered, and where the argument actually was.

On the day it is written, the recorded line does nothing at all. The line still falls to Rs 25,00,000. The review is still 90 days away. Nobody's mind has been changed and no outcome has moved by a rupee. Judged by its effect on that day, recording dissent looks like a consolation prize handed to the loser of an argument, and a great many people quietly treat it that way.

The entire point shows up 90 days later. The review does not open with somebody trying to remember what anybody was worried about in March. The review opens with a written sentence saying that one person thought 90 days was too long and why, and the first thing it does is check whether that person was right. The dissenter has not been consoled; they have been handed the opening question of the next meeting. Recorded dissent is worth nothing on the day it is written and is the most useful line in the document on the day the review falls due. Every record worth keeping has exactly that shape.

There is a second thing it does, and it works on the people who have not spoken yet. A room where disagreement gets written down is a room where the next person raises theirs. A room where it evaporates is a room that gets quieter every quarter, and quiet rooms produce decisions that look unanimous and are not. The effect is slow, it runs in both directions, and nobody notices it happening in either.

The seniority problem bites hardest at this point, so somebody with no standing needs one practical note. Recording dissent looks like something only the chair can do. It is not. Whoever is writing the record writes it, and nobody will stop them. When somebody else is writing it, one line sent afterwards does the same work: for the record, the risk colleague argued for a 30 day review rather than 90, on the grounds of the concentration in one order. The message needs no permission and takes forty seconds. The line is simply true and it helps the person it names, so pushing back on it is hard.

Which of these moves needs rank, and what works without it?

Every one of these moves has to work for somebody in their first month who cannot call a meeting, cannot overrule anybody, and will not be believed on their say-so. Working without rank is not a limitation being worked around; it is the natural condition of the whole subject, and the reason is worth showing rather than simply asserting.

The item number does the believing, so answering from the record needs no rank. Naming a day needs no rank either. Nobody has to grant permission for Thursday. Naming an interest as a problem needs no rank, precisely because it is phrased as a problem rather than as a claim about somebody. Recording dissent needs no rank. A recorded dissent is a sentence in a document, not a change to a decision. Everything in this area that requires standing is a way of getting people to accept an assertion, and everything that works without standing replaces the assertion with something the room can go and check for itself.

What should never be said in response to a challenge?

Three responses cost more than the challenge did. None of the three is rudeness and none is incompetence, and that is why they survive. All three feel like the sensible thing at the moment they are said, and two of the three are frequently made of true statements.

The first is repeating the argument at greater length. Somebody pushes, and the response is the same case again with more supporting detail in it. Repeating at greater length does the most damage per second of the three, and the failure below works it through in full.

The second is conceding a point that has not been checked. Under pressure, giving something away feels generous and de-escalating, and it buys about nine seconds. Then one of two things happens. Either the original was right and a wrong correction has just gone into the record for somebody to find; or the original was wrong and the error was found by guessing rather than by looking. Either way the room learns that the speaker's agreement carries no information. The move that costs nothing is to say the point will be checked and confirmed by a named day. The same formula as before works in this case too.

The third is naming the questioner's motive: that the question is only being asked because the relationship matters to the person asking. The accusation may be entirely accurate. Naming a motive converts a disagreement about work into a disagreement about a person, in public, in front of everybody who will be in the next meeting, and that makes it the single most expensive sentence available to somebody with no standing. An accusation is the one response with no recovery path inside the room. The other person now has to defend themselves rather than the position, and nobody present can tell which of the two is right about a motive.

How does a lender, an analyst or a household use this?

Most people will spend more of their working life asking these questions than answering them, and from the other side of the table the same moves read differently. Somebody challenging an analysis is running a cheap test, and knowing what they are testing for shows what to supply.

A challenger almost never wants the number. The challenger wants to know where the work is thin, and then decides how much of their own attention to spend on it. A challenge is a probe for the soft part. Handing over the soft part yourself is effective for that reason, and counterintuitive for the same one: it answers the real question in one sentence and returns the rest of the meeting to whatever was actually worth discussing. All four moves share one practical purpose: moving the room off the arithmetic and on to the judgement, and only one of those two is worth the time of six people.

Notice that the household row is the same move as the other three. The contractor who says the quote excludes the tiling, before anybody asks, has done exactly what naming a load bearing assumption does in a meeting. The exclusion sounds like a weakness being confessed and functions as evidence that nothing else is being hidden. Usually it wins the work.

One last observation from the same assignment closes the loop on the five questions. Of the five, the one that mattered most to how Sharada Iyer was read afterwards was the one with no answer in it. The three answered from the log were competent and forgettable. The objection was handled well and took ten minutes. The unanswered one produced an email on Thursday with an item number in it, sent to six people, two days after a promise made out loud in front of all of them. The email is the only thing from the whole twelve minutes that anybody could verify for themselves, and verifiable is worth more than impressive every single time.

The Kavery research desk, Kavery Capital Services Private Limited, Meenakshi Tubes Private Limited, Sharada Iyer, Prakash Nadar and Latha Menon are invented.
Educational material. Not advice on any investment, tax, budget or market position.