Friction Hunting: How to Find What's Actually Slowing Your Business

Nothing looked broken. That's what I remember most about the hardest stretches in the businesses I've run over the past three decades. Orders were coming in. The team was working hard — harder every quarter, in fact. And yet everything took a little longer than it used to. Margins thinned in ways no single line item explained. Nobody could point at the problem, because the problem wasn't a thing that broke. It was friction — dozens of small, quiet costs compounding inside workflows that everyone had stopped noticing because, technically, they "worked."

Growing companies don't struggle from a lack of ambition. They struggle because operational friction compounds faster than teams can scale. Friction can be found, priced, and killed — but only with a repeatable discipline, because it will never announce itself. Over the years I've boiled that discipline down to a method my team calls the Friction Hunt. This post lays out the whole thing. It costs nothing to run and requires nothing but a walk-through and a calculator.

What Is Friction Hunting?

Friction hunting is the discipline of systematically finding the small, recurring costs hiding inside workflows that still function — the repetitive tasks, the bottlenecks, and the quiet inaccuracies that consume hours nobody tracks. The Friction Hunt is a five-step method for doing it: Pick one workflow, Walk it with the person who does the work, Count every touch, Price the friction honestly, and Rank-and-Kill the worst item first. The output is a one-page Friction Ledger — a ranked, priced list of specific problems, which is worth more than any tool you could buy because it tells you what's actually worth fixing.

The Three Faces of Friction

Before hunting, know what you're hunting. Nearly every piece of operational friction wears one of three faces, and on the hunt you'll tag every find as one of them.

Repetition — the most visible face: product descriptions rewritten one at a time, order details retyped between systems, the weekly report assembled by hand from four exports. Each instance feels too small to fix; multiplied by frequency, it's a part-time job nobody hired for.

Bottlenecks — the quiet face: work piling up behind one person, one approval, one system everything routes through. The tell is the phrase "waiting on." Hear it twice in a day about the same step and you've found one.

Inaccuracies — the expensive face, because it charges twice: once when the error happens, again when someone finds and fixes it. Oversold inventory, stale pricing, a catalog that disagrees with the warehouse — friction that burns hours and erodes customer trust at the same time.

The Friction Hunt: Five Steps

Step 1 — PICK One Workflow

Choose a single workflow that touches money: order handling, catalog updates, invoicing, sales follow-up, returns. One workflow per hunt — the discipline dies the day it becomes "audit everything." A hunt should fit inside ninety minutes. If you're torn, pick the workflow behind your most recent end-of-month scramble; the scramble is a map of every process that doesn't run cleanly the other twenty-nine days. The financial impact of the "friction" should be a major consideration as well.

Step 2 — WALK It With the Doer

Walk the workflow end to end with the person who actually does it — not the person who designed it. They know things the org chart doesn't. Two rules govern the walk. First, watch, don't fix: the moment you start solving, people stop showing you the truth. Second, capture every workaround verbatim — the spreadsheet built "just for now" three years ago, the copy-paste ritual between systems that don't talk. Every workaround is a flag your team already planted on a friction point and has been paying for with their time.

Step 3 — COUNT Every Touch

A touch is any human intervention the workflow can't move without: a retype, an export, an approval, a "let me check with…", a correction, a wait. Log each one on a simple tally — what the touch is, which face it wears (Repetition, Bottleneck, or Inaccuracy), minutes it takes, times per week it happens, and who does it. Don't editorialize during the count; a touch that "only takes two minutes" still goes on the sheet. Two minutes, forty times a week, is a work week per year.

Step 4 — PRICE It Honestly

Now the calculator. For each touch: times per week × minutes per touch × loaded hourly rate, annualized. For inaccuracies, add the double charge: the cost of the error itself plus the hours spent finding and fixing it. The total for the workflow is its Friction Score. Write the number down — in my experience the first one is always a shock. It shocks large companies too: MIT's researchers found that even among enterprises spending heavily on AI, 95% of pilots produce zero measurable return — and a core reason is that nobody priced the actual friction before buying a fix.

Step 5 — RANK and KILL

Put every priced item on one page — the Friction Ledger — ranked by Friction Score, worst first. Then kill the top item, in this order: process first (can the step be deleted or combined?), then tool (only if a step must exist and a machine should do it). Before any tool spends a dollar, apply the gate I've used since my eCommerce years — the pain formula: adoption happens only when the pain of the problem exceeds the friction of the new behavior. If a fix demands more new behavior than the pain it removes, it will die in a drawer no matter what it cost. Kill one item, not five; momentum comes from finished kills, not long lists.

Make Friction Hunting a Quarterly Discipline

One hunt fixes one workflow. The discipline compounds when it's on the calendar: one workflow per quarter, same five steps, same one-page ledger. Four hunts a year and you will know more about where your margin goes than most companies twice your size — and your team starts spotting friction on their own, because now it has a name, three faces, and a price tag.

The Bottom Line

Friction hunting isn't a technology exercise — it's a seeing exercise. The costs slowing your business are already visible to the people doing the work; they've just been renamed "how we do things." Pick, Walk, Count, Price, Kill. One workflow, ninety minutes, one page. I've written before about what operational friction quietly costs SMBs — but the pricing only matters once you've done the hunting.

Go find your friction. It's there, it's countable, and it's costing you more than you think.