---
title: "Program 14 Remembers Everything Except Why, In Conversation With a Liquidation Valuer"
canonical: "https://current.tinyfish.ai/issue/53/practitioners-corner/article/57546/program-14-remembers-everything-except-why-in-conversation"
publication: "TinyFish"
publisher: "TinyFish"
language: "en-US"
guide: "https://current.tinyfish.ai/llms.txt"
issue: 53
section: "practitioners-corner"
section_name: "Practitioner's Corner"
published: 2026-10-30
word_count: 1622
source_links: 12
---

# Program 14 Remembers Everything Except Why, In Conversation With a Liquidation Valuer

*October 30, 2026*

[Read on the website](<https://current.tinyfish.ai/issue/53/practitioners-corner/article/57546/program-14-remembers-everything-except-why-in-conversation>)

More small businesses now come to market with a new line on the asset list. It sits somewhere between the walk-in freezer and the delivery van, and it says *automations*. That might mean scripts that reconcile invoices, a spreadsheet that quotes jobs, or an agent that logs into a supplier portal at 6 a.m. and files the reorder before anyone has made coffee. Buyers want to know what it's worth. Brokers call it "the tech stack," in the voice people use for a loft conversion that may or may not have planning permission.

People who price equipment ask one question before they reach for a number. Which world is this asset being priced in, the one where it keeps running or the one where it gets wheeled out the door? Software almost never gets asked. So we put the question to someone who works in rooms after the music has stopped.

Dolores Mangle values commercial laundries and kitchens for sale and liquidation. She exists only on this page. She is a composite assembled from appraisal definitions and patent filings, and her surname is a piece of laundry equipment. She points out that being imaginary puts her in the same position as most of what she prices: perfectly real until someone asks who can operate her. We met at an imagined auction preview. Washer-extractors stood in a row with lot tags zip-tied to their doors, and she was pressing buttons on a control panel like someone trying to recall an old phone number.

## You value dryers and dish machines. Why are people handing you scripts?

**Dolores:** Because nobody knows which column to put them in, and I'm the person you ring when a column is in doubt.

I start every job the same way, tumble dryer or macro. A credible equipment valuation names its **premise** before it prints a number.[^1] *Continued use* means installed, powered, maintained, and part of a business that's still trading. *Liquidation* means it's leaving the building. Same machine, two numbers. Sellers hate that. Buyers pretend to. That's the whole job.

## So what's a script worth in liquidation?

**Dolores:** About what the USB stick cost. Nobody bids on a macro. You can't strap a file to a pallet. I've checked.

So all of its value sits in continued use, and continued use quietly assumes somebody keeps it running. For a dryer, that somebody is any competent technician with the manual. For a script, it's often one particular person, and she's out in the car park holding a cardboard box with a cactus in it.

That's my reasoning, by the way. I haven't found appraisal guidance that treats scripts this way. I'm borrowing my own trade's vocabulary to see whether it takes the weight.

## Does it hold? Your machines run software too.

**Dolores:** Riddled with it, which is what made me take this seriously. A chemical dispenser on a commercial washer has to be programmed with wash formulas, anywhere from one to several dozen depending on the setup.[^2] Read the patents and a formula turns out to be a table of timings. This pump waits so many seconds after the washer signals, then runs for so many more. One system holds twenty of those tables.[^3] One manufacturer's washer-extractor stores ninety-nine programmable wash programs in non-volatile memory, and the manual helpfully includes a section on erasing every one of them at once.[^4] I've met sellers who would find that section very tempting.

So the machine remembers perfectly. Program 14 survives a power cut. What it doesn't remember is *why program 14 exists*. Hotel towels? Care home bedding? Some dosing systems match the chemistry to the load with a classification code, one for shirts, another for towels.[^5] Nothing on the hardware tells you what the codes mean. The meaning lives in the setup, and probably in somebody's head.

*(She taps the panel. It beeps, unhelpfully.)*

Here's the part I love. Somebody once decided that getting a controller's programmed status back out was a problem worth patenting. Their answer was a system that keeps a permanent record of what each controller had been told to do.[^2] Picture it. A room full of dosing boxes, someone staring at them, and slowly the thought arrives: we should probably write this down.

> Writing it down was the invention.

A script is program 14 without the stainless steel. It runs, and it keeps running until the first thing it touches changes. That's when you find out whether anyone left in the building knows what it was for.

## Then the old spreadsheet must be the safe asset. Everyone understands a spreadsheet.

**Dolores:** Everyone has *seen* a spreadsheet. I'd like to see its paperwork.

Raymond Panko pulled together field audits of real company spreadsheets, and the more recent ones, using better methods, found errors in at least 86% of them.[^6] Researchers argue about what counts as a significant error, so the figure wanders depending on who's counting.[^7] I haven't seen anyone find the direction reversed. There's also an ethnographic study in which developers struggled to explain parts of spreadsheets they'd built themselves.[^8] So when a seller tells me "she'll walk you through it before she goes," my question is whether she can walk herself through it.

Then there are the stale ones. A link to a file on a departed colleague's drive keeps showing the last value it pulled in, so the number looks perfectly healthy even though it stopped updating in March. Or a formula covers rows 2 to 500 on a sheet that now runs to 800.[^9] The arithmetic is flawless. It's just adding up the wrong customers.

> A spreadsheet fails politely. It never throws an error. It just keeps handing you a confident number about a world that has moved on.

## And the agent? At least it tells you when it breaks.

**Dolores:** Sometimes. What worries me more is that it *acts*. A wrong spreadsheet sits there being wrong, like a man waiting at a bus stop that was moved last year. A wrong agent goes ahead and submits the form. And anything that drives other people's websites breaks when those websites change. A button that wanders to the other side of the page can be enough.

I'm trying very hard to be fair here. From where I stand, the spreadsheet and the agent are the same kind of asset. Each is somebody's judgment frozen into a file. **What I'm pricing is how many people can thaw it.**

## Which would you rather inherit?

**Dolores:** Whichever one somebody other than its author has already broken and fixed. Preferably twice.

## How does the market actually price "only she understands it"?

**Dolores:** More quietly than people expect. Sale advisers call it **key-person risk**, meaning the slice of future cash flow a buyer can't count on because the person producing it is leaving.[^10] The market rarely writes the business off. It shaves the multiple, holds money back in earnouts and escrow, and keeps the seller around for a longer transition until the business proves it can survive the departure.[^10] One adviser reckons heavy dependency in smaller firms knocks half a turn to a turn and a half off the earnings multiple, with transitions of six to twenty-four months.[^11] That figure comes from a website that would very much like your business, so treat it the way you'd treat a dealer's opinion of his own forklift.

What strikes me is that the market already knows how to price the person who built the script. It's just been calling them "the owner."

## What would earn an automation its full continued-use value?

**Dolores:** A drill. Send the builder on holiday with the phone properly off, and I mean off, not "off except for emergencies." Then watch whether the buyer's staff can recover the thing when it falls over, change a rule someone's arguing about, and explain the scar behind a strange setting. Documentation proves you own the files. It tells you very little about Tuesday, when the supplier redesigns their login page.

I'd also want to see at least one change the buyer made that the builder didn't direct. One is enough. *That's the moment the asset actually changes hands.*

## Doesn't that just move the dependency onto whoever passes the drill?

**Dolores:** *(A long pause. She takes her hand off the panel.)* Yes. Damn.

Advisers say the same thing about managers. Once the business depends on them instead of the owner, they become the key people.[^12] So no, you can't drill dependency out of a business. You can spread it thin enough that losing one person is an inconvenience rather than an auction. I'd know the difference. I'm the auction.

I don't love that answer. Most of what I'd put on a script turns out to be a number for how many people could explain program 14.

## What do you tell a seller who wants a higher number?

**Dolores:** Write down what each controller has been told to do, and why. Someone patented that idea. You're allowed to just do it.

---

[^1]: Machinery and equipment valuation premises: https://ctacquisitions.com/machinery-and-equipment-valuation/
[^2]: US Patent 6,325,286, chemical dispensing system: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/6325286
[^3]: US Patent 8,117,703, modular dispensing system: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/8117703
[^4]: Alliance Laundry Systems washer-extractor programming manual: https://docs.alliancelaundry.com/tech_pdf/Production/D1632EN.pdf
[^5]: US Patent 5,897,671, wash cycle identification: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/5897671
[^6]: Panko, "Spreadsheet Errors: What We Know. What We Think We Can Do": https://arxiv.org/pdf/0802.3457
[^7]: "Correctness is not enough": https://arxiv.org/pdf/0808.2045
[^8]: "Spreadsheet Risk Management in Organisations": https://arxiv.org/pdf/1009.2775
[^9]: On inherited workbooks, stale external links and fixed ranges: https://zelobiz.com/explain-excel-spreadsheet-with-ai/
[^10]: FE International on key-person risk in business sales: https://www.feinternational.com/blog/key-person-risk-business-sale
[^11]: Focus HR, "Key Person Risk: The Quiet Deal Killer": https://focushr.net/key-person-risk-the-quiet-deal-killer/
[^12]: Salt Creek Advisory on reducing owner dependence before a sale: https://saltcreekadvisory.com/articles/reduce-owner-dependence-before-selling-2026

## More

- [Practitioner's Corner — Issue 53](<https://current.tinyfish.ai/issue/53/practitioners-corner/markdown>)
- [Issue 53](<https://current.tinyfish.ai/issue/53/markdown>)
- [Publication guide](<https://current.tinyfish.ai/llms.txt>)