“But the spreadsheet team is in the budget for the Philadelphia office, not the servicing hub.”

“Exactly. That’s how we hit the target.”

“So the ‘cost per contract’ we showed the board yesterday… it’s a lie?”

“It’s not a lie. It’s a measurement. It just doesn’t measure the Philadelphia office.”

I sat three rows behind them, nursing a lukewarm coffee in a Marriott ballroom that smelled of industrial carpet cleaner and ambition. On the stage, a consultant was clicking through a slide deck titled Efficiency Benchmarks for the Modern Equipment Lessor. There it was: the Top Quartile. A bar chart showing that the most efficient firms were servicing forty thousand contracts per full-time employee. The room was nodding. People were taking photos of the slide with their phones.

I looked at the back of the woman’s head in the second row-the one who had just admitted to hiding an entire team of spreadsheet-jockeys in a different cost center. She wasn’t a villain. She was a survivor. She knew that in the world of institutional finance, you are what you measure, and if you can’t make the reality efficient, you make the measurement convenient.

The Micrometer’s Blind Spot

I spent a decade as a machine calibration specialist. My job was to ensure that a millimeter in one factory was exactly the same as a millimeter in another. When you use a micrometer, you’re taught about the “feel” of the ratchet. If you crank it too hard against the metal, the frame of the micrometer itself flexes. You get a reading that looks incredibly precise-three decimal places of certainty-but is fundamentally wrong. You’ve distorted the object by trying to measure it.

Benchmarks in the equipment finance world work exactly like a heavy-handed micrometer. We measure what comes out of the SQL tables easily: delinquency ratios, days to book, and servicing cost per contract. These are the “clean” metrics. They are the extractable truths.

But because they are clean, we optimize for them at the expense of everything messy. We treat the benchmark as a mirror of operational performance, when in reality, it is a mirror of data availability. If the system doesn’t track how many times a human had to manually recalculate a mid-term modification because the software couldn’t handle a partial asset swap, then that “human touch” doesn’t exist in the benchmark. It’s just “noise” or “overhead,” or it’s hidden in the Philadelphia office.

The Tonnage Paradox

In the 18th century, the British government taxed merchant ships based on a formula called “Builder’s Old Measurement.” It was a simple calculation: (Length minus 3/5ths of the beam) multiplied by the beam, multiplied by half the beam, divided by 94.

The formula was designed to estimate the carrying capacity (tonnage) of a ship. But because it used the beam (width) twice in the calculation and ignored the depth of the hull entirely, shipbuilders changed their designs. They built ships that were incredibly deep and dangerously narrow. These “wall-sided” ships could carry massive amounts of cargo while appearing “small” on paper, thus avoiding taxes.

The problem? They were unstable. They capsized in heavy seas. They were slow. They were miserable to sail. By measuring “taxable tonnage” instead of “actual displacement,” the government incentivized the creation of death traps.

We are doing the same thing with portfolio servicing. By measuring “cost per contract,” we incentivize leaders to strip away the “depth” of their servicing capability. We ignore the turnaround time for a complex end-of-term negotiation or the accuracy of a casualty loss calculation, because those things are hard to pull into a standardized survey. We build “narrow” operations that look efficient on a slide but capsize the moment a customer asks for something that isn’t a standard monthly payment.

The Tyranny of the Extractable

Last week, I was comparing the prices of two seemingly identical torque wrenches online. One was $210, the other was $44. On a spreadsheet, they have the same “specifications.” They both claim to measure 10-100 foot-pounds. They both have a half-inch drive. If you are a procurement officer measured on “cost per tool,” you buy the $44 one every time.

But I’ve held both. The $44 wrench has a pivot point that feels like it’s filled with sand. It loses its calibration after three clicks. The $210 wrench is a precision instrument that will still be accurate when my grandkids are using it. The spreadsheet cannot capture the “feel” of the tool, so the spreadsheet favors the junk.

In the back office of a lender, the “junk” is the manual workaround. If your core system is a legacy beast that requires a vendor change request every time a tax jurisdiction changes its rates, you don’t fix the system. That’s a capital expense. Instead, you hire a temp to run a spreadsheet. That temp is an “operating expense.” And if you can hide that expense under “Administrative Support” rather than “Portfolio Servicing,” your benchmark stays in the Top Quartile.

We have reached a point where the unmeasured dimensions of finance are free to degrade indefinitely because they don’t appear in the quarterly review. We see the delinquency rates staying steady, but we don’t see that it’s taking six weeks to process a simple contract assumption. We don’t see the “shadow work” that keeps the engine from seizing.

The Architecture of Honesty

I once made the mistake of calibrating a high-precision lab scale using a set of weights that had been sitting in a humid basement. I assumed the 100-gram weight was 100 grams. It wasn’t. It had oxidized. It was 100.04 grams. Every scale I “fixed” that day was actually being broken by my own standard of truth.

If your standard of truth is a benchmark that ignores the mechanics of life-in-life servicing, you are calibrating your business to fail.

Real efficiency isn’t the absence of cost; it’s the presence of integrated data. When you’re looking at equipment financing software, the metric that actually matters isn’t how many contracts an employee can “oversee.” It’s how many of those contracts can be modified, billed, and reconciled without the data ever leaving the system of record.

If a payment is touched by a human hand, it’s a failure of the architecture. If a contract modification requires a spreadsheet, it’s a failure of the platform. But most industry surveys don’t ask about the “human touch ratio.” They ask about the “headcount.” So, we keep the headcount low and the spreadsheets high.

The Cost of the Invisible

The most dangerous part of steering by extractable data is that you lose the ability to see a better future. When everything is “fine” on the dashboard, there is no urgency to innovate.

I spoke with a COO recently who was proud of his 2.1% delinquency rate. He thought he was winning. But when we dug into the “Philadelphia office” equivalent of his firm, we found that his collections team was spending 40% of their time just trying to figure out which collateral records were up to date. They were doing detective work, not collections.

The “cost per contract” looked great because the collectors were salaried and the “detective work” wasn’t a line item. But the opportunity cost was massive. If those collectors had been using an API-first system where collateral data was synced in real-time with the accounting core, they could have been proactive. They could have been preventing defaults before they happened.

But “prevented defaults” are hard to measure. “Detective work” isn’t in the survey. So he continues to steer his narrow, deep ship into the storm, confident that because his tonnage is low, he must be doing a great job.

The Calibration of Reality

We need to stop asking “What is the industry average?” and start asking “What is the invisible cost of my current success?”

If you are in the top quartile, ask yourself what you had to hide to get there. Is your servicing speed a result of better technology, or is it a result of a staff that has become world-class at “working around” the limitations of a 20-year-old database?

The goal of a system shouldn’t be to produce a pretty report for a consultant. It should be to provide a frictionless environment where the contract, the collateral, and the customer are in a constant, automated state of sync. That kind of efficiency doesn’t just look good on a slide; it feels good in the P&L. It feels like a torque wrench that clicks exactly when it’s supposed to, every single time, without you having to wonder if the frame is flexing.

I left that conference early. I couldn’t stand to watch another slide about “Optimal Ratios.” On the way out, I saw the Operations Director from the second row. She was on her phone, looking stressed, likely talking to someone in Philadelphia about a formula that wouldn’t balance. She was a victim of her own high scores.

We are so busy measuring the height of the waves that we’ve forgotten to check if the ship is actually moving forward. It’s time we stopped optimizing for the extractable and started building for the reality of the work. Because at the end of the day, a benchmark is just a number. But a spreadsheet team in Philadelphia? That’s a symptom of a system that has lost its way.

I’m done with the top quartile. I’d rather be in the quartile that actually knows where its data lives. That’s the only place where real growth-the kind that doesn’t capsize when the wind picks up-actually happens.