The Other Wall: Why Success Is What Slows You Down

Useful controls, left alone, become friction

Here is a pattern I find in almost every company I work with once it is past its scrappy stage: decisions keep enforcing themselves long after the conditions that justified them have changed. A discount approval added after one bad deal becomes permanent. A report outlives the manager who asked for it. An acquired business arrives with controls built for a different product and keeps them. Software overlaps. Exceptions pile up without review. Each addition made sense on its own. Nobody looks at them together, and the commercial engine starts leaking revenue and margin out the seams. That is organizational friction: the gap between the hands-on work a routine job needs and the elapsed time it takes, most of it decision latency, time spent waiting for someone with the authority to say yes.

Two problems sit under this, and they have different fixes. One is a process that no longer fits the business. The other is a team that lacks the knowledge or the authority to use the process well. Confusing them wastes a year: more training will not fix an unnecessary approval, and removing an approval will not supply missing technical skill.

Why winning makes a company slow

Marshall Goldsmith wrote a book for executives, What Got You Here Won't Get You There, and the idea travels to companies: the instincts and processes that carried a business this far become load-bearing beliefs, and nobody audits a load-bearing belief. Clausewitz had the mechanical version. "Everything is very simple in war," he wrote, "but the simplest thing is difficult," because small resistances "accumulate and produce a friction, which no man can imagine exactly who has not seen war" (On War, Book 1, Chapter 7). Scale does not remove friction. It manufactures it.

A smaller competitor that has fixed its own system does not beat you by getting bigger. It beats you by carrying less friction: ten experiments run while you schedule the meeting to approve one. The edge is metabolism, not resources. Any company that has cleared the $10M Wall has the same move available: once your own system works, speed is the advantage you hold over the larger, slower incumbent.

Is it the market, or is it you?

Before you fix friction, rule out the market. A few signals separate the two, and none of them needs a consultant:

• What the market did. If your segment grew and you did not, the market is not the story. If the whole segment shrank, part of the stall is not yours to fix, and the rest still is.
• Where you are losing. A market shift tends to hit most segments at once. Friction shows up where your process touches the deal: losses in segments you used to win, to competitors whose products did not get better.
• Why you are losing. Pull the reasons on your last 20 lost quotes. "Price" and "lead time" can point either way. "Took too long to get back to us" and "went with someone easier to buy from" point inside.
• How your speed is trending. If your quote turnaround keeps getting longer while what customers require does not, that is friction, whatever the market is doing.
• What customers say. Ask five recent customers, including two you lost, what it is like to buy from you. If slowness is the problem, it is usually the first thing they mention.

Most stalled companies have some of both. The point is to know which share is yours, because that is the share you can fix.

What a stalled winner looks like

The stall rarely announces itself. Revenue still grows, just slower. Underneath, the pattern is consistent. Ideas have to clear more people, each of whom can say no and none of whom can say yes alone. The installed base and this quarter's number make any disruptive bet look reckless, so the company keeps optimizing the thing that is slowly dying. And the systems that created the win have become sacred, so "this is how we do it" has replaced "is this still the best way?"

Some of this shows up across companies, not only yours. In one large survey, roughly 70% of respondents at organizations with one to three management layers rated their decisions high quality, against roughly 45% at seven layers or more (McKinsey).

Measure it on 20 cases

You can measure it instead of arguing about it. Pick one process, say routine quotes. Pull 20 recent comparable cases, including the stalled and lost ones, and split each into hands-on work, internal waiting, and customer waiting. Find the largest controllable delay and ask what causes it: policy, capacity, information, skill, or authority. Then change one handoff or approval, inside clear limits, and watch speed, margin, errors, and delivery afterward. Twenty cases is a starting sample, not a benchmark, and not all waiting is waste. But "move faster" is useless, while "your engineering review takes two hours and the request sits unassigned for six days" is something a manager can fix on Monday.

For each case, record the request date, every handoff and who it went to, the minutes of hands-on work at each step, the reason for each wait (policy, capacity, information, skill, authority, or the customer), and the done date. The timestamps are usually already in the CRM, the ERP, or the email thread; the wait reasons take a conversation with whoever held the job. Download the 20-case template (CSV).

What puts the energy back

A stalled winner is not stuck. The stall is usually a maintenance problem, and maintenance problems have known fixes.

Give elapsed time an owner. Every step in a process has an owner. The total usually has none, which is how a two-hour review turns into a six-day wait with nobody at fault. Name one person per core process who holds the request-to-done number and has the authority to change a handoff or retire an approval. It is the smallest change on this list, and the one that makes the others stick.

Protect a lane. Carve out a team allowed to move at small-company speed, and give it what permission alone cannot:

• a defined customer segment and a bounded problem
• one accountable decision-maker
• a capable and trained team
• a spending limit with pricing and quality boundaries
• a test window matched to the buying cycle
• a stop, modify, or scale rule
• a path for wins to re-enter the core business, carrying both the process and the knowledge behind it

Subtract. LEGO, near collapse in 2003, cut its product range, roughly halved its count of unique parts, and rebuilt around what sold (David C. Robertson, Brick by Brick; ThinkInsights). Getting faster usually starts by removing what accumulated, not by adding more. Try it on one process: list every approval step and recurring report, and ask who would notice if each one stopped for 30 days. Anything nobody would miss is a candidate to retire.

Redeploy, do not defend. Kodak did not miss digital; its own engineer built the first digital camera in 1975. Its margins, plants, and habits were built on film. It filed for bankruptcy in January 2012 (abc27). Fujifilm faced the same shock, treated film as a declining business, and redeployed its chemistry into cosmetics, diagnostics, and biologics; healthcare is now about a third of its revenue (Fujifilm). One treated its success as a base to defend, the other as capital to redeploy.

Make the truth cheap to say. When Alan Mulally took over a Ford losing $12.7B in 2006, his main instrument was a weekly review where every executive showed the same color-coded chart. The first time one of them showed a red one, Mulally applauded; the next week the room filled with red and yellow, which meant the company could finally see itself (Ford 2006 results; Bryce Hoffman, American Icon). The cheap version: one weekly review, the same one-page chart for every leader, and a rule that a red status gets help, not blame.

I ran this at Powerblanket, the manufacturer I led, after Thermon acquired it in 2022. We kept the unit's marketing budget, people, and decision speed close to the business. Inside its existing budget, the team could launch campaigns and run tests without corporate sign-off, and it owned all of e-commerce, including testing Thermon products through online channels and partners. We protected the habits the team had built: moving fast, testing, letting customers and employees make the call. Thermon contributed what scale improves: industrial reach, resources, and market knowledge.

The part that matters most is what happened after I left the business-unit role. We had built the commercial systems and trained the people to run and keep improving them. I moved into a Thermon corporate role in January 2025, and the unit kept growing on those systems and that team. EBITDA more than doubled between fiscal 2022 and fiscal 2026, from about $3M to roughly $7M, and fiscal 2026, the year ended March 31, 2026 and the first full year after my move, was the highest of the series (internal figures). The capability outlived my day-to-day leadership, which is the whole point. A process that only works while its author is in the room is not a capability. It is a dependency.

The corporate role showed me the other side of the wall. Thermon was growing well through this period; what I found were pockets of unowned friction inside a healthy company, which is where the other wall hides. I ran demand generation, and most of what I fixed was friction nobody owned. Three acquired businesses still ran their own websites years after the deals closed, because folding them in belonged to no one; integrating them and redirecting the old sites lifted search traffic for their keywords. The marketing and sales drives held more than a million files with no naming convention, which is information that exists and cannot be found; I used AI to name, summarize, and categorize them into an asset library people could search. There was no steady content program, so demand arrived in bursts. A webinar series fixed that, at about 150 leads per webinar, and left evergreen content on the site. Then the plumbing: forms, rep portals, trade shows, ads, search. Leads rose 88% year over year after I took it over (internal figures). None of it needed a reorganization. It needed an owner for each piece of elapsed time, and permission to subtract.

When an outside read helps

"You cannot see your own blind spots" is too strong, and it sounds like a sales line. Internal teams diagnose their own problems all the time. An outside read earns its place in narrower conditions: when competing explanations cannot be resolved internally, when the people who run a process cannot independently judge what it costs because they built it and are measured on it, when the problem crosses departments, or when the business lacks the time, authority, or method to investigate. The evidence that outside help can move the number is good if you read it carefully. In a randomized trial with large textile firms in India, plants that received hands-on management consulting (a diagnosis plus months of implementation) raised productivity 17% in the first year through better quality, efficiency, and inventory management (Bloom et al., Quarterly Journal of Economics, 2013). The gain came from the diagnosis and the work of changing things, not from a fresh pair of eyes alone.

Is your company stalling? A five-question read

Five yes or no questions. Answer for the company as it runs today, not as it is supposed to run. This gives you a read, not a grade.

Find the leak before it compounds

Pick one process this week and run the work-versus-wait split on 20 cases. If the largest delay is a control nobody can explain, you have found a leak. A Commercial Engine Diagnostic does this across every core process, in about four weeks of fixed scope: it is built to find where the system has drifted, establish what the drift costs, and end in a readout and a sequenced 90-day plan with clear ownership, skills, and decision rights. Book a 15-minute fit call. Or start with the smaller company's version of this problem: The $10M Wall.

Book a 15-minute fit call →

Nate Evans
Founder, Evans Advisory | Fractional CCO for industrial manufacturers
Led Powerblanket from $5.5M to $21M in sales and through its sale to Thermon for $35M at 12x EBITDA.
evansadvisory.co

Quick answers

What is the other wall?

The other wall is the stall that catches companies after they succeed. The company builds real systems, managers, and software, and slowly gets harder to change. Every step has a reason, and nobody owns the total elapsed time.

How can I tell a market problem from organizational friction?

A market shift tends to hit most segments at once. Friction shows up where your process touches the deal: losses in segments you used to win, "took too long" in your loss reasons, and quote turnaround that lengthens while customer requirements do not.

How do I measure friction?

Measure organizational friction on one routine process: pull 20 recent cases, including stalled and lost ones, and split each into hands-on work, internal waiting, and customer waiting. Find the largest delay you control and ask whether policy, capacity, information, skill, or authority causes it.

How does an established company get faster?

Give the total elapsed time of each core process one owner, then subtract before adding: retire approvals and reports nobody would miss, give one team a protected lane with a budget and clear limits, and make it safe to report problems early.

When does an outside diagnosis help?

When competing explanations cannot be resolved internally, when the people who run a process are measured on it, when the problem crosses departments, or when the business lacks the time, authority, or method to investigate.