An INFP man can say better than anyone what the work is for, and the saying is portable — it lifts off him, goes on a wall, and gets used to sell the thing he objected to. His judgment does not lift off. An institution takes the part that travels.
The first time I read one of Amos Kirkbride’s loan files I thought somebody had left a draft of something in the wrong system.
It was a 2009 file, pulled at random out of a batch I was reviewing my second week here, and in the field marked REASON FOR EXCEPTION, where every other lender in this institution had typed a code or the word character, there were four sentences.
They said that the member had a repossession in 2007 from a vehicle her ex-husband had been driving and had stopped paying on, that she had since been at the same employer for nineteen months and had never missed a rent payment there or anywhere, that the loan we were making her was for a transmission and not a want, and that the score in front of me was a picture of a marriage and not of a woman.
The loan was for two thousand eight hundred dollars. It paid as agreed, and closed in 2013, and she has been a member here for nineteen years and has a mortgage with us now.
I watched him do it properly one time, which is one more time than most people in this building have.
Arnetta Threlkeld came in about a debt consolidation in 2019 with a folder and a number in her head that was wrong, and Amos sat with her for fifty minutes, and about thirty of those minutes were not about money.
“How long’s your mother been with you?”
She said since the spring.
“And the car’s how old?”
She told him, and then she told him about a job she was thinking about taking that paid less and was nine minutes from the house, and he asked her whether she would take it if the loan went through, and she said she thought she would.
He restructured the thing around the job she had not taken yet.
I would not have approved that loan. I want that in here plainly, because I have spent a lot of this account being decent about him and that one I would have declined on the debt-to-income, correctly, on the information available to me, which was less information.
Amos came to Ironwood in 1997 as a loan clerk. He was twenty-six and had a degree in literature and had been substitute teaching. He is a tall, stooped, slow-talking man who apologizes for the length of his own emails in the emails. He listens to people in a way that makes them tell him more than they meant to, and he does not use it. He remembers the names of members’ children and he is not doing a technique when he does it. He has, in nineteen years, told me twice that he was wrong about something, both times in writing, at length, unprompted, when nobody had noticed.
Ironwood Community Credit Union is four hundred and six million dollars in assets across three branches in Wheelock and one in the county seat. Ormond Teale ran it from 1988 to 2017 and I replaced him. There is a vault door in the main branch that has been propped open with the same fire extinguisher since 1998, because the hinge is out and the quote to fix it is more than the vault is worth, and I have approved that quote twice and cancelled it twice.
I was hired from outside. Everyone in this building assumed the job was Amos’s, including, I believe, Amos, and the board’s search committee did not put him on the slate at all, and I read the minutes about it in my first month.
Here is what Ormond told the committee, and what two board members have said to me since, and what I have said myself, at a board table, at a moment when saying it decided something.
He’s the best thing about this credit union and he can’t run anything.
The member meeting on the merger is in five weeks. The agenda went to print on the fifteenth. Amos is on it, presenting the combined-organization piece, and I put him there, and I have not told him.
Cleta Marchetti chaired the board through most of it and had a way of closing an item that everybody could do.
“Let’s have Amos write that up.”
The Reason Field
Consumer lending in a place like this is a set of rules and a hole in the rules.
The rules are the policy: score bands, debt-to-income limits, loan-to-value ceilings, time on the job. The hole is the exception. An exception is a loan approved outside policy, and every one of them is logged, and the log goes to the board every month and to the examiner every eighteen, and the whole reason the category exists is that a policy is a general statement about people and every borrower is a specific one.
The system has a field where the lender says why. That is the entire mechanism. In our old platform it was a free-text box with no character limit, and in nineteen years Amos filled it, on something like two thousand loans, with prose.
Not stories. This matters and I got it wrong for eight years. They read like stories because they are in sentences, but every one of them is doing work: here is the fact the score is compressing, here is the reason it will not repeat, here is what specifically I checked. The prose was the underwriting. There was nowhere else for that reasoning to go.
We put in a new loan origination system in 2021. The reason field became a dropdown. Twelve options, one of which is other, and if you pick other a box opens with a two-hundred-character limit, and the vendor’s implementation team explained that this was a control improvement and they were right, because free text is unsearchable and unreportable and an examiner cannot audit a paragraph.
Reza Amini came to us in 2022, out of a bank, and will be very good at this. He asked me in his first quarter why the older files had paragraphs in them.
I said that used to be the field.
“For what, though? There’s a code for everything.”
I told him there wasn’t a code for everything and he looked at the dropdown on his screen and said there were twelve, and I said that was the point I was making, and he did not follow me, and I did not push it, and that exchange has been on my mind since the spring.
The drive-up at the main branch keeps a jar of dog biscuits and has kept one since long before my time, and there is a woman who comes through it once a month with a horse trailer she cannot get around the turn, so she parks across two lanes and walks up, and nobody has ever said anything to her about it, and the drive-up teller has the biscuit ready for a dog that has been dead since 2019 and puts it in the tube anyway.
I sit on the zoning board of appeals in my township, which has cost me two friendships and a holiday card list, and I would not give up the seat.
What It’s For
Amos’s actual faculty is not warmth, and calling it warmth is how this institution avoided looking at it for twenty years.
What he has is a fixed sense of what a thing is for, running underneath everything, all the time, comparing. Not a rule and not a policy. A reading of the purpose — what this credit union said it was, what a loan is supposed to do for a person — held against what is actually happening in front of him, continuously, whether or not anybody asked.
And when the two come apart he can say so, at length, with real force, and this is where he differs from every other person of conscience I have worked with. He does not go quiet. He writes you two pages. He stands up in a management meeting and says the thing so well that people repeat it in the parking lot.
What he cannot do is close it.
The finding does not come out of him as one demand. It comes out as a spray of possibilities — we could do this instead, or we could do it this way, or what if the whole product were built differently, or has anybody asked the members. Five doors, all of them open, none of them costed. Ask him which one and he will tell you he does not know yet and that any of them would be better than what we are doing.
In a room where the question is what are we for, there is nobody in this state I would rather have. In a room where the question is which of these two, he makes the room longer.
Bettina Ochs ran our branch on the north side from 2011 to 2018 and had something close to the same equipment, and when Bettina opened five doors in a meeting it was called scattered, and when Amos did the same thing it was called a valuable perspective, and Bettina left for a bank and is a regional manager now and makes more than I do. What the identical performance earns a man in a room like that, and what it earns a woman, is a different count and it is taken somewhere else.
I want to put down a thing I did in my second year, because it is the smaller version of what came later.
We were arguing about branch hours. Amos had spoken for six or seven minutes about who actually needs a lobby open at half past five and what it means to a person who works second shift to be told that their credit union closes when their shift starts. It was a good six minutes. Then he stopped, and the room waited, and somebody asked what he’d propose, and he said he wasn’t sure but that the current schedule was built for our convenience and not theirs.
“So what do you want to do,” I said.
“I don’t know yet.”
I asked whether he wanted us to survey the members, and he said maybe, and then he said that a survey would ask people what they wanted from a schedule they already knew about, which is a different question entirely.
He was right about that as well.
I said we’d take it under advisement and moved to the next item.
The hours did not change for three years, and when they did change it was because a competitor extended theirs.
Indirect
The indirect auto program started in 2019 and it was mine.
I want to describe it honestly, because the version where it was a betrayal of our mission is not available to me. Our loan book was flat. Our members were buying cars and financing them at the dealership because that is where the paperwork is, and the dealer was selling that paper to somebody else, and we were watching our own members’ auto loans show up on other institutions’ books. Indirect lending puts us on the dealer’s rate sheet. It is how a credit union our size stays in the auto business at all, and I would make the same call now.
What it does, structurally, is move the decision. A dealer’s finance office sends an application at nine at night and wants an answer in eleven minutes, and you cannot write four sentences about a marriage in eleven minutes, and the whole thing runs on tiers and buy rates and a scorecard. In the first full year we booked more auto paper than the branches had done in the previous three.
Amos wrote me a memo about it. Four pages. It is the best-written document I have received in this job and I still have it. Its argument, if I can compress something that resists compression, was that we were about to acquire a large number of borrowers we had never met, on terms set by somebody whose interest is the sale, and that within a few years the composition of our membership would be different in a way nobody had voted on.
He was right about all of that, and I can now put numbers on the part he could not. Of the members we booked through dealers in the first three years, a bit over half have never been inside one of our branches, and a shade under a third carry nothing with us but the car. They are our members. They own this institution. Most of them do not know that, and no amount of regret about it makes it any less the case.
I did not know any of that then. What I knew then was that the memo contained no alternative I could take to a board.
Then in 2021, with the new system, we cut lender exception authority from fifteen thousand to five, and moved anything above it to a committee that meets twice a week.
He came into my office about that one, which he almost never does.
And I did the thing.
There is a sentence on the wall in our main lobby, in eight-inch letters, and it has been there since 2014, and it is his. We lend to the person, not the score. He wrote it for a newsletter in 2013 and Ormond had it painted.
He was standing in my office telling me that a five-thousand-dollar authority meant that the lending decisions that most needed a human being were the only ones a human being would no longer be making. And I said — warmly, and in front of nobody, which somehow makes it worse — that we lend to the person and not the score, and that the committee was full of people, and that nothing about that had changed.
He stopped. He actually stopped, mid-sentence, and looked at the corner of my desk for a second, and then he said, “That’s fair,” and thanked me for my time.
None of that was fair. It was his own sentence, and I reached for it because I knew it would work, and it worked because he could not argue with it without disowning it.
The Reveal
Redmond Sasse did our brand work in 2023. Six months, discovery interviews, a member survey, three workshops. The engagement was a hundred and eighty-six thousand dollars.
The reveal was at the all-staff in the county seat, ninety-some people in a rented hall with box lunches. Redmond is good at what he does. He built the presentation properly — the research, the tension, the competitive landscape — and he held the platform line to the end.
He put it up in white on a dark slide.
We lend to the person, not the score.
He said the phrase had emerged from the discovery work and that it captured something the organization had been circling for years without naming, and that everything downstream — the campaign, the branch signage, the onboarding — would ladder up to it.
Amos was in the fourth row on the aisle. I was standing at the side because I never sit at those.
He applauded.
I watched a man applaud his own sentence being sold back to his employer as a discovery, and he was not being ironic, he was pleased, and I have gone over what was in his face and the only thing I can find in it is that he was glad they had landed on the right one.
I found him at the coffee urns during the break.
I said the platform line had landed well.
“It’s the right one,” he said. He was genuinely happy about it. “It’s better than the one I’d have picked now.”
I asked him what he’d have picked now.
He thought about it for a second and then gave me three, and each of the three was interesting, and by the time he got to the third one Redmond was calling us back in.
Tavia Roeder, who runs our marketing, told me afterward that Redmond’s team had pulled the line out of an old newsletter in week two.
I said that made sense.
I did not say anything at the all-staff and I did not say anything the next day and I have said nothing about it since, and the sign painters did the lobby in the county seat that autumn.
Brock’s Numbers
Brock Yeary has been our CFO since 2012 and does not like me, which has been useful.
He walked into my office last spring with two pages and put them on the desk the way he does, squared, facing me.
“You’re going to want to sit with this.”
The indirect book was charging off at a shade under three per cent. The direct consumer book was under one. That much I knew and it is what everybody expects; indirect is a different animal and the pricing is supposed to carry it.
The second page was the one.
He had gone back and pulled every exception Amos had personally approved between 2005 and 2020 — a bit over two thousand loans, twenty-nine million dollars — and run them against everything else we had booked in the same period.
They charged off at forty basis points.
Not near the book average. Under it. A portfolio of loans that failed our policy, approved by one man on the strength of paragraphs he typed into a text box, performed better than the loans that passed.
I said the sample was self-selected and that he had chosen his own deals.
“He did,” Brock said. “That’s the whole point. That’s what an underwriter is.”
I asked what he expected me to do with it.
“Nothing. I want it on the record that somebody in this building said it once.”
The Archetypes Slide
Redmond’s discovery deck is a hundred and forty pages and I had never opened the appendix until this spring.
There is a section in it called leadership archetypes. His team ran the management group through an instrument in the second workshop — the sixteen boxes — and the appendix has a page for each of us, four letters and a paragraph, and then a mapping table with our letters down one axis and brand voice attributes across the other.
That is what the table is for. It is how the voice was built. They typed the leadership team in order to write the sentences.
Amos’s page says INFP.
The paragraph under it is about idealism and depth and caring about authenticity, It reads like something printed on a placemat, and reading it is how a person decides the whole exercise is astrology and never turns the page. What is worth having sits further in, in the methodology section, where Redmond’s people explain that a set of letters compresses an order of habits — which one carries the weight, which one it comes out through, and which ones sit down at the bottom, reachable only on purpose and never quickly. Where each habit lands for this one is written out on the page that handles the type by itself, pulled from the sixteen boxes that workshop grid came off.
The habit doing most of the work is the one I described a few pages back and could not name for eight years: a standard of what a thing is for, held privately, not derived from anybody, running as a continuous comparison against what is actually going on.
The second one is what it comes out through, and this is the whole of his career. It opens. The finding meets a faculty that generates possibilities, and what leaves the man is not a demand but a field of them — five doors, all live, none of them priced. It is why the memo about indirect lending had no alternative in it. There was no shortage of alternatives. There were too many, all of them still open, and closing them is a different operation performed by a different part of a person.
It also produces language. That is the part nobody in this building understood, including me. A standard about purpose, run through a faculty that keeps finding new ways to put a thing, generates sentences that other people can pick up and carry — and a sentence, unlike a judgment, is completely portable. You can lift it off the man. You can put it on a wall. You can pay a consultant a hundred and eighty-six thousand dollars to hand it back to you.
Down at the bottom of his order sit the two that this job is made of. Running the apparatus — the schedule, the numbers, the org chart, holding people to a standard from the outside. And carrying the institution’s factual history: what we actually did in 2011 and what it cost and how it went, in detail, retrievable. Ormond had both of those and nothing else, and Ormond ran this place for twenty-nine years.
I have said that in a board meeting — that Amos couldn’t run anything — and I should be precise about what was true in it. He could not have done my job. He also had a book of twenty-nine million dollars that outperformed the institution and nobody in the building, including him, ever described that as running something.
The plain part, which had better come from me. If your employer is putting your name and your voice on a communication about a decision you argued against, ask in writing what happens if you decline, and keep the answer. And in a regulated business — which this is — if you are asked to describe a product to members in a way you think is misleading, there is a regulator with a channel for that and in most places using it is protected. Those are the actual instruments. A slide with sixteen boxes on it is none of those things, and this credit union has now paid to have the letters run across its own leadership in order to write copy, which is a use nobody involved would defend if they had to say it in a sentence.
Redmond’s methodology page notes that the instrument is a facilitation aid, that roughly a third of people who take it again a month on land on a different letter somewhere, while the continuous measures behind the boxes hold their shape.
The mapping table is three pages after that note.
I asked Redmond about it at the wrap dinner, mildly, the way you ask about a thing you have not decided to be angry about yet.
“It’s a shortcut into voice,” he said. “You don’t build a platform off it. You use it to know who you’re building it with.”
I asked whether he’d told the team that.
“It’s in the deck,” he said.
The Agenda
The merger closes in the fall if the members approve it, and the member meeting is in five weeks, and Amos is on the agenda.
I put him there because he is the only person at this institution the membership actually trusts, and because a merger vote in a place like this turns on whether people believe the thing they had is still going to be there afterward, and because he can say that better than anyone alive.
I also put him there because if he is on the platform saying it, he is not in the room objecting to it.
I did not think that sentence when I made the decision. I thought it three weeks later in the parking lot and could not get rid of it.
Two things I had settled came unsettled after Brock’s second page.
The 2009 file, first. My second week, four sentences in a text box, and I have told that story at two conferences as a charming detail about a small institution with a lot of heart. What was in that box was an underwriting rationale that a scorecard could not produce, written by the only person who had the information, in the only place the system gave him to put it. I filed it under heart because it was in sentences. If he had put the same content in a table I would have called it analysis and given him a department.
Then 2021, in my office, and that’s fair. I have carried that for five years as him accepting a business reality with good grace. He was not accepting anything. He had been handed his own sentence and there was no move available that did not require him to take it back, and the thing he had built over nineteen years — a way of saying what this place was for that other people could carry — had been turned around and pointed at him by somebody who had learned it from him.
The parts the page did not have room for: he told a member in 2011 that a denial had come from the board when it had come from him, and when she raised it with me I let the story stand; he has not spoken to his older brother since 2016 over something nobody here knows; and he is, in a specific and narrow way, vain — he keeps every note anybody has written him in a drawer in his desk and I know that because he told me, proudly, at the holiday party, and none of it belongs under anything else here.
Reza asked me last week who was doing the member meeting.
I told him.
“Good,” he said. “They’ll listen to him.”
I am going to go and tell Amos about the agenda this week.
I have a version where I ask him rather than tell him. I have written it out. It is a good version and it ends with me saying that if he does not want to do it, he does not have to, which is a sentence I can say and cannot mean, because the agenda is printed and there are nine thousand of them in a warehouse in the county seat.
The people in these essays are composites drawn from long observation. The theory is not.