Credibility: How It Is Built and How Fast It Is Lost

Credibility is the expectation that what a person says will hold. The expectation is built by small kept promises rather than by being impressive: the answer arriving first, the review finding accepted without argument, the follow-up landing on the day that was named. Credibility is lost faster than it is built because one unkept promise makes a reader check everything said before it.

One assignment runs through this guide from end to end. The Kavery research desk, four people inside Kavery Capital Services Private Limited, produced an assessment of Meenakshi Tubes Private Limited, a company Kavery Capital Services Private Limited has money out to. Sharada Iyer wrote it, seven months into the role, working alone on an assignment for the third time. Prakash Nadar reviewed it. Latha Menon commissioned it and chaired the meeting on 12 March where it was discussed for twenty minutes in front of six people. The subject is what happened between the people rather than what Meenakshi Tubes Private Limited manufactures.

The mechanism sitting underneath credibility is a question of cost. Nobody who reads an analyst's work can afford to rebuild it. When a different person on the Kavery research desk actually did rebuild one figure from the workpaper, it took twenty five minutes, and that was with the workpaper open in front of them and nothing going wrong. Nobody has twenty five minutes for every number they are handed in a day. So a reader reaches for a cheaper signal instead, and the cheaper signal is this: did the last few things this person said turn out to hold. CredibilityThe expectation that what a person says will hold. It is formed by other people out of that person's past statements, and it belongs to them rather than to the person it describes. is that cheap signal. Every working habit worth having either keeps the cheap signal readable or pushes the reader back onto the expensive one.

What is credibility in finance work?

Credibility is the expectation that what a person says will hold. The definition leaves a great deal out. Credibility says nothing about how much the person knows, nothing about how senior they are, and nothing about how well the last thing they wrote was received. The expectation is a forecast that somebody else makes about the next statement, and like every forecast it is assembled out of a record. Credibility is not a verdict anybody passes on a person; it is a prediction they make about that person's next sentence, and predictions are made out of what has already happened.

The mechanism is easier to see outside the office, where there is no finance in it. A household has a plumber who says he will come on Tuesday morning. The first time, somebody stays back from work to be sure. The second time, somebody still stays back. By the fourth time, nobody does. The household has quietly moved from checking the plumber to expecting him. Nothing about the plumbing changed. The count of stated things that turned out to be true is what changed, and the household has stopped paying the cost of verifying. The saving is what credibility actually is, and it is paid out of somebody else's time.

The version inside a finance desk is the same transaction wearing different clothes. A reader who expects a figure to hold stops re-deriving it and starts arguing with what it means. A reader who does not expect it to hold either re-derives it and loses an afternoon, or waves it through and costs everybody later. Both of those are worse than the first outcome, and neither has anything to do with how good the analysis was.

Is credibility the same as being liked, being senior or being confident?

No. The three near neighbours are worth separating carefully, and an early-career professional can spend a year chasing the wrong one. Being liked is a fact about how people feel in someone's presence. Seniority is a fact about where they sit. Confidence is a fact about how they sound. Credibility is a fact about whether their statements hold, and it is the only one of the four that a reader can test without knowing them at all. All three neighbours travel alongside credibility often enough to be mistaken for it, and none of them is it. Only a statement can be checked, and only checking builds the expectation.

Confidence deserves one extra line. Confidence is the neighbour that actively misleads. A firmly delivered wrong number does more damage than a hesitantly delivered one, for the simple reason that more people acted on it. The point is not to sound unsure. The aim is to make the confidence match the evidence. When the speaker does sound certain, the certainty is then itself information.

What actually built it in this case, and did any of the three need expertise?

Start where Sharada Iyer started. Nobody would choose to start there. The first two assignments produced alone came back from Prakash Nadar for rework. The first came back because the opening did not say what the work had found. The second came back because a chart in it could not be read. Two assignments returned inside seven months is not a small thing to have happen, and it ended nothing at all. Seven months later the third assignment went to a meeting of six people and held.

Three things changed between the second assignment and the third, and none of the three was knowledge. The answer arrived in the first sixty words every time. The three findings from the review pass were accepted without a single argument. And of the five questions asked in the twelve minutes the item actually took, the one that could not be answered was met with a named day, and the confirmation arrived on that day. Not one of the three things that built credibility here required knowing anything Sharada Iyer did not already know in month one. Most people in their first year are holding that luck without knowing it.

Why is putting the answer first a credibility move rather than a writing move?

Putting the answer first is Barbara Minto's rule, set out in The Pyramid Principle, 1978, and it is almost always taught as a writing technique. The rule really is a writing technique, and it is also the cheapest credibility move available to somebody with no track recordThe set of a person's past statements that somebody else can go back and check. It is what credibility is calculated from.. The reason is worth working through slowly.

A reader who has the finding in the first sixty words spends the rest of the document testing it. A reader who does not have it spends the document hunting for it, and arrives at the end with two open questions instead of one: what does this person actually think, and can it be relied on. The second question never gets asked properly. The reader is still busy with the first. Answer first collapses that. Answer first hands the reader something concrete to disagree with on line one, and disagreement is how a reader finds out, quickly, that the writer was right. Answer first is a credibility move because it converts the reader from a searcher into a tester, and only a tester can discover that the statements hold.

There is a second effect that matters more than the first. A reader without the answer does not sit patiently; they build their own. Part way through they have a private view, and the finding, when it finally appears, has to defeat that view rather than simply being heard. Sharada Iyer's first returned assignment failed on exactly this. The work was not wrong. The draft was arranged so that Prakash Nadar had to read nine hundred words to discover what it concluded, and every one of those nine hundred words was read by somebody slightly annoyed.

What does accepting a review finding without arguing actually buy?

The review pass on this assignment ran forty minutes against about nine hours of writing, a little over seven per cent of the time the work took, and it produced three findings. Roughly one finding every thirteen minutes. The ratio says what a review is: a very cheap machine for catching what the writer cannot see, run by somebody who did not write the work and therefore cannot make the writer's mistakes.

Now the part that decides credibility. When a finding lands, there are three things that can be done with it, and only one of them is free. Argue in the moment. Accept it without looking, change the figure and move on. Or check it, then answer with the trace. Arguing costs the writer whether the writer turns out to be right or wrong, and checking first costs nothing at all. The choice between them is arithmetic rather than character.

The two branches run like this. Arguing and turning out to be wrong spends credibility and then hands the position back. No sequence available is more expensive. Arguing and turning out to be right wins a small point and teaches the reviewer that findings will be contested. The price of reviewing that person quietly rises, and the number of findings they get next time falls. Checking first compares well against both. If the reviewer is wrong, the trace comes back twenty minutes later, and the reviewer learns that this person checks. If the reviewer is right, an argument that would have had to be climbed out of never happened. Three of the four outcomes favour checking, and the fourth is a draw.

Accepting a finding is not the same as agreeing with it before looking. Agreeing before looking is worse than arguing. The wrong figure stays in the work, and the reviewer learns that the agreement carries no information. The move is narrow and it is entirely available to somebody in their first week: thanks, then a check, then a return with what the check found.

Why does a follow-up that lands on the day outweigh the answer given in the room?

The item took twelve minutes of a twenty minute slot on 12 March, and five questions were asked inside it. Three of the five were answerable straight from the source log by item number. One was answerable only with a date. One was not a question at all, and handling that kind is set out under handling questions, objections and dissent in a finance meeting. The three answered from the log were good moments. The one answered with a date was the one that mattered.

Here is the asymmetry that most people miss. An answer given in a meeting is almost impossible for the room to check while the meeting is happening. The answer sounds right or it does not. A named day is different: it is a claim about the future, it is unambiguous, and it either happens or it does not, in front of everybody, with no interpretation required. Of everything said in those twelve minutes, the only claim the room could actually test was the date. Hitting a named day is the cheapest credibility available to anybody at any level.

The email went out on Thursday 22 March, two days after the meeting. The email carried the answer and the source log item number, so anybody who wanted to check it could do so without writing back and asking. The whole thing was four sentences long. Nothing in it required seniority, a team, or anybody's permission.

Why is credibility lost so much faster than it is built?

Every kept promiseA stated commitment, done by the time it was promised. The smallest unit credibility is built from. is a small depositOne small action that adds to the expectation that a person's statements will hold. A day named and hit, a finding accepted, a figure that survives being checked.. Every unkept one is a withdrawalOne action that makes a reader go back and re-read what was said before it. Withdrawals are larger than deposits and there is no way to make them smaller. several times its size. The picture of small deposits and one large withdrawal belongs to Stephen Covey, The 7 Habits of Highly Effective People, 1989, who called it an account held between two people, and it is used here because the shape is right even though the arithmetic is not.

The reason for the asymmetry is not that people are unforgiving. A broken promise works backwards through the record. Until it happens, every kept promise looked like process: this person names days and hits them. Afterwards, each of those same kept promises could have been luck, or an easy week, or a small ask. Nothing was subtracted from them individually, and yet all of them are now worth less. The explanation that covered them has been weakened. A kept promise adds one observation to the record. A broken one changes how every observation already in the record gets read, and that is the whole of why the fall is steeper than the climb.

Credibility carries no number. No move is worth so many points, and a quantity that nobody can measure should not be given a scale. The shape of the rise and the fall is real. The size of each step is not. The moment somebody hands over a number for this, they have said that they are guessing.

Two things follow from the shape, and only one of them is the obvious one. The obvious one is that a promise is expensive, so there should be fewer of them and they should be smaller: Thursday named when Wednesday was achievable, one thing to be closed rather than three. The less obvious one is that the fall is survivable. Somebody who believes a single mistake ends them will hide the next one, and a hidden mistake is the only kind that actually does lasting damage.

How to Build Credibility as an Early-Career Finance Professional: What Is Actually Available in the First Month?

Most advice on this subject quietly assumes a person who can call a meeting, decline an assignment, or be believed because they say they checked. In a first month none of those is available, and pretending otherwise wastes the month. One test sorts the advice: does the move need a history behind it. If it does, it can be put aside without regret. If it does not, it is available today.

Four moves survive that test, and they are the whole of the first month list. Being on time, every single time: a named day is the only claim a room can verify. Being specific, meaning a number, a date and a document rather than roughly or shortly. Saying what is not known and attaching a day to it. And closing what was opened, meaning that every thread started is either finished or explicitly handed back to somebody who has agreed to take it. The four moves that build credibility fastest are exactly the four that require no history, and that is luck rather than a consolation.

One more move belongs on the list, and it is the cheapest of all: asking out loud which question the work is meant to answer. On this assignment the request arrived as whether the firm should be worried about Meenakshi Tubes Private Limited. A request in that form has no stopping point in it and cannot be finished. One question unlocked it, asked out loud in a corridor: what would be done differently depending on the answer. Latha Menon said reduce the line or leave it. The exchange in the corridor is where the 11 per cent floor came from, and it took under a minute from somebody with no standing whatsoever.

What happens to credibility after a mistake?

A mistake that has been seen does not end a person's credibility, and the week after one is exactly when that is hardest to believe. A seen mistake suspends part of the credibility, makes the next few weeks heavier than they should be, and is recoverable by ordinary means. The people whose credibility does not come back are almost never the people who made the error, but the ones who went quiet afterwards, or who explained before they disclosed, or who let somebody else find the second one.

RecoveryWhat is done after a mistake so that the reader can stop checking everything. It is a short sequence, and the order of it carries most of the effect. runs in a fixed order of four steps, and the order does more work than the content. Say that there is an error. Say what it changes, naming the conclusion that moves and by how much. Say what is being done about it and by when. Then say what stops it happening again, named specifically enough that it is a check rather than a promise to be careful. The error is stated before anything else about it is said. An explanation that arrives first is heard as a defence, and the reader stops listening before the sentence that mattered ever arrives.

The one thing that makes recovery worse is doing step three first. Explaining first feels helpful. The explanation sounds like the machine dropped a digit when the file was copied across, and everything after that is heard as positioning, including the parts that are simply true. The same applies to reassurance offered too early: telling somebody the conclusion is unaffected before telling them there is a problem reads as managing them, even when it is accurate.

One further point about recovery, and it is the one that costs people the most. Finding one's own error and reporting it is worth more than never being caught. Self-reporting is direct evidence of what the reader actually wants to know: whether anything is checking the work when nobody is watching. Reporting it does not feel like an advantage at the time, and it is still true.

How can a person tell whether they have any?

There is no measurement and no score, so the question has to be answered by observation instead. Two signs are reliable, and both of them are behavioural rather than verbal, so they are worth more than anything anybody says out loud.

The first sign is being brought in earlier. In month one the work arrives after the question has been decided and the shape has been chosen, and the task is to check a figure. Later, the work arrives before the question has been settled, and somebody asks what the question should be. The second sign is that the questions get harder. Easy questions are what people ask when they are not really going to rely on the answer. When somebody asks what would change the conclusion, they are planning to act on it, and they want to know where it breaks. Both signs look like more work rather than like recognition. People frequently read the arrival of credibility as pressure and conclude, precisely backwards, that they are being tested because they are doubted.

How does somebody on the other side of the table actually use this?

The whole thing is worth turning around. Credibility is not something a person holds. Credibility sits with the reader, it is spent by the reader, and the reader uses it to solve a scheduling problem: how much of a limited amount of attention to spend re-deriving other people's numbers. Every practitioner who reads finance work is running that calculation constantly, usually without noticing.

The practical value of credibility is that it changes the meeting's subject: the room spends its time either on the analyst's arithmetic or on the analyst's judgement, and only one of those is worth twenty minutes of six people's attention. The change of subject is also the honest reason to care about credibility. Credibility is not about being respected. It is about whether the hardest part of the work, the part where somebody actually thought about something, ever gets discussed at all.

One last observation from the same assignment shows the signal working when nobody was performing. Sharada Iyer took five days of leave on 24 March and left a handover one sheet long: what is open, what is waiting on somebody else, what comes next, and where the workpaper sits. The handover worked, and the reason it worked had nothing to do with the note. It worked because a different person had already rebuilt the 10.8 per cent from that workbook in twenty five minutes, so the note could point at something rather than having to explain it. Credibility built while nobody was looking is what makes a handover of one sheet enough.

Kavery Capital Services Private Limited, the Kavery research desk, 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.