Opex creep: the slow-killer most owners don't see

The slow-killer most owners don't see

Pull up your income statement from three years ago. Then pull up last year's. Most small business owners doing this for the first time get an unpleasant surprise: revenue grew nicely, gross margin held up fine, but operating expenses grew faster than revenue. The result is that EBITDA margin is two, three, sometimes five points lower than it was — and the owner is working harder for less.

This is opex creep. It's the most common reason profitable, growing small businesses stop being satisfying to run. The business looks bigger every year. The owner is busier every year. The bank balance grows more slowly than revenue. And nobody has stolen anything; nobody has made a single bad decision; the company just slowly ate its own margin one tolerable line item at a time.

The fix isn't hard once you see it. The hard part is seeing it — because each individual cost increase is defensible, and the aggregate damage only shows up in a ratio that most owners aren't tracking.

What opex creep actually is

Opex creep is the gradual increase in operating expenses as a percentage of revenue. The absolute dollar amount almost always grows when a business grows — that's normal. Opex creep is the percentage growing.

A worked example. A residential plumbing company:

• Three years ago: $1.2M revenue, $440K COGS (37% gross margin: 63%), $480K opex (40% of revenue), $280K EBITDA (23% margin).
• Today: $1.9M revenue, $700K COGS (37% gross margin: 63% — held perfectly), $874K opex (46% of revenue), $326K EBITDA (17% margin).

Revenue grew 58%. EBITDA dollars grew 16%. EBITDA margin dropped 6 points. The owner is now running a business 60% larger and taking home barely more dollars than they did before. They're working more hours, managing more people, dealing with more headaches — for almost the same payoff.

The owner can't point to anything that broke. They got a slightly bigger office because the team grew. They added a dispatcher because they couldn't answer phones anymore. The bookkeeper's rate went up. They started paying for a CRM. Insurance went up. Vehicle costs went up. Phone plans went up. Every single decision was reasonable in isolation. The aggregate destroyed six points of margin.

Why it's structurally invisible

Three reasons opex creep is hard to catch in real time, even for owners who look at their financials regularly:

• Each individual increase is defensible. A new hire because the team is overloaded is the right call. Upgrading from QuickBooks to a real ERP because you've outgrown spreadsheets is the right call. Adding a sales rep when the existing team is at capacity is the right call. None of these decisions, taken individually, look like a problem. The problem is in the aggregate.
• The income statement shows dollars, not percentages. When revenue grows from $1.2M to $1.9M and opex grows from $480K to $874K, both numbers went up. The income statement looks like progress. The ratio — opex as a % of revenue — tells the real story, but most accounting software doesn't display ratios alongside the dollar columns.
• The owner is busy. Growing businesses consume their owner's attention. Three years ago you were watching every expense line because there were 30 of them. Today there are 80, and you can't watch them all. The natural defense is to delegate or batch the review — and reviewing in batches is exactly when creep slips through.

The single ratio to track: opex as % of revenue

This is the diagnostic ratio that catches opex creep before margin is destroyed. The formula is trivial:

Opex % = Total operating expenses / Revenue

What counts as opex: everything below the gross profit line on your income statement, except interest and tax. So payroll for non-production staff, rent, utilities, insurance, software subscriptions, marketing, professional fees, vehicle costs, depreciation, and the rest.

What does not count as opex: cost of goods sold (COGS) — that's a different line, and it has its own ratio (gross margin). The two ratios answer different questions. Gross margin tells you whether your pricing and production economics are healthy. Opex % tells you whether your overhead is in proportion to the business.

More on the difference in the COGS vs. opex guide. Getting the categorization right matters here, because mis-classifying expenses as COGS instead of opex (or vice versa) breaks both ratios at once.

Industry benchmarks for opex %

Healthy ranges vary widely by business model. The headline numbers, for owner-operator small businesses ($250K–$10M revenue):

• Trades and construction: typical opex 22–32% of revenue. Most cost is in COGS (labor, materials, subcontractors), so overhead stays relatively lean.
• Manufacturing: typical opex 18–28%. Similar to trades — heavy COGS, moderate overhead.
• Professional services (agencies, consulting, firms): typical opex 60–75%. Most of the cost is people, and most of those people are opex (since in services the “production” isn't a separate department). Gross margin is typically very high and opex % is correspondingly very high. The two ratios are essentially mirror images.
• Retail (general): typical opex 22–32%. COGS is the dominant cost (60–70% of revenue); overhead fills out the rest.
• Restaurants: typical opex 28–35%. Labor is split between COGS (back-of-house) and opex (front-of-house, management), making the ratio industry-specific in its accounting conventions.
• Ecommerce: typical opex 25–38%, with heavy variability based on whether marketing is treated as opex (most cases) or a customer-acquisition cost tracked separately.

The absolute number matters less than the change over time for your specific business. A SaaS business at 70% opex is fine; the same SaaS business at 70% last year and 78% this year is bleeding. A construction company at 30% opex is fine; the same company at 30% last year and 36% this year has a creep problem.

Why the trend matters more than the level

Most opex-creep diagnoses come from comparing your current ratio to your own historical ratios, not to industry benchmarks. The reason: industry benchmarks are noisy. They mix businesses with different mixes of services, different age profiles, different operating models. Your own ratio three years ago, computed from your own books, is the most comparable benchmark you have.

Compute the ratio for the last 4–6 fiscal years if you can. Plot it. Look at the trend.

• Stable (within 1–2 points year-over-year): the business is scaling cleanly. Revenue and overhead growing in proportion. No action needed beyond routine attention.
• Drifting up 1 point per year: yellow flag. Over five years, that's 5 points off EBITDA margin. At a $2M revenue business, that's $100K of annual EBITDA gone — about $500K of business valuation lost on a 5x EBITDA multiple.
• Drifting up 2+ points per year: red flag. Aggressive creep. Something structural is happening — probably the org chart is growing faster than the revenue can support, or you've added recurring software/service spend that isn't producing measurable value.
• Spiking 3+ points in a single year: not opex creep, just opex. Something specific happened — a major hire, a system migration, a new location. Investigate the specific line that moved; this is different from the slow-creep pattern.

The four most common creep sources

When opex % is rising and you're trying to figure out where the leak is, these four categories are responsible for the great majority of cases I've seen in owner-operator small businesses.

1. Headcount that scaled with revenue but not productivity

A team of 12 generating $2M of revenue per person scales to a team of 18 generating $1.7M per person. Each new hire felt necessary. Nobody is underperforming. But productivity per employee dropped 15%, and the wage bill grew faster than revenue.

The diagnostic question: revenue per employee, last year vs. this year. If it's flat or dropping, headcount is growing faster than the business needs.

The fix is rarely “fire someone.” More often it's “don't replace the next person who leaves naturally,” or “the next role we're tempted to fill, we automate or outsource instead.” Headcount that grew 15% over four years can shrink 5% over two years through normal attrition without anybody being fired.

2. Software stack that no one audits

Three years ago you had QuickBooks, a payroll service, and email. Today you have those plus a CRM, a project management tool, a time tracking app, two scheduling apps for different teams, three different chat platforms because different departments preferred different ones, a marketing automation tool, two analytics tools, a customer support tool, an HR management tool, and a separate document signing tool. Each one was a sensible decision at the time. The monthly bill total is somewhere between $1,500 and $4,000 and nobody has audited the list in two years.

The diagnostic action: pull the credit card statement for last month. Highlight every recurring software charge. Have someone (you, or a competent admin) verify each one is actively used by the team and producing value. Cancel anything that fails the test.

The first time you do this you'll typically find 20-30% of the SaaS spend is dead weight. After that, schedule it quarterly. The audit takes 30 minutes; the savings compound forever.

3. Professional fees that scaled with complexity, not value

The accountant's annual fee is up because there's more to reconcile. The lawyer's retainer crept higher because there are more contracts. The insurance broker renewed at higher premiums because the business is bigger. The consultant who helped with the rebrand is now on a monthly engagement.

Each of these is defensible in isolation. The aggregate question is: are these professional services producing value proportional to their cost? Most owners find, on inspection, that one or two of them are overpriced for what's actually being delivered — usually because the relationship started small and grew without an explicit renegotiation.

4. Marketing spend that's not measured

Sponsorships, branded swag, conference attendance, agency retainers, ad spend on platforms that aren't generating tracked leads. Marketing is the line where “we can't measure it but it's probably working” gets the most leeway, and that leeway compounds into serious dollars.

The diagnostic question for every dollar of marketing spend: what is it producing, and how do we know? Anything that can't be answered concretely is a candidate for either measurement or cancellation. Often you don't need to cancel; you need to either start measuring it or accept that it's a brand-building cost and budget accordingly.

The quarterly review process

Catching opex creep doesn't require a finance team or a big-software platform. It requires 30 minutes a quarter with your income statement and a calculator. Do this:

1. Pull the last four quarters of P&Lsfrom your accounting software. Most software lets you export this in one click.
2. Compute opex % of revenue for each quarter. Total opex (everything below gross profit, excluding interest and tax) divided by revenue for that quarter. Four numbers.
3. Compare to the same quarter last year. Q3 this year vs. Q3 last year is more useful than Q3 this year vs. Q2 this year, because opex has seasonal patterns that distort sequential comparisons.
4. If opex % is up 1+ points YoY, drill in. Pull the line-item P&L. Sort opex lines by absolute growth in dollars. The top 3–5 lines almost always account for the majority of the creep.
5. For each top creep line, ask: what is producing this growth, and is it generating proportional revenue growth? If yes, the spend is healthy — just growth-related. If no, that's a candidate for reduction or cancellation.
6. Make one or two specific changes. The point isn't to slash everything. It's to address the worst leaks as you find them. Three quarters in a row of small adjustments compound.

The valuation impact

For owners who might ever sell the business, opex creep is a much bigger deal than it looks. The mechanics:

Buyers value owner-operator businesses on a multiple of earnings — typically 3–6× for SDE-based valuation in the small-business space. Every dollar of excess opex you can cut translates to a dollar of EBITDA (or SDE) growth, which translates to 3–6 dollars of sale price.

A worked example. A $2M services business with 18% EBITDA margin ($360K EBITDA), valued at 4× EBITDA, is worth $1.44M. The same business with opex trimmed by 4 points (back to where it was three years ago) would have 22% EBITDA margin ($440K EBITDA), valued at 4×, is worth $1.76M. A $320K difference for an owner who spent six months auditing their software stack and renegotiating their professional fees.

More on the valuation mechanics in the owner's salary guide and the business valuation calculator.

What to do this week

Three concrete actions, in order of payoff:

This week: compute your opex % of revenue for the trailing 12 months and the same 12 months from one year ago. Two numbers. The difference is your year-over-year drift. If it's up more than 1 point, you have creep. The Profit Lift calculator can do this in 10 minutes if you have your numbers handy.

This month: if creep is present, pull the line-item opex from both periods and identify the top three contributors. Apply the diagnostic question to each: is this growing because the business is growing, or because the business hasn't pruned?

This quarter: establish the quarterly review as a standing 30-minute appointment on your calendar. Most owners who do this for two consecutive quarters never have a serious creep problem again, because they catch it at 1 point of drift instead of 5.

The bigger point

Opex creep is the most expensive non-mistake in small business. Nothing is broken, no decision was wrong, and the aggregate damage is enormous. Most owners only catch it when they get an external trigger — a buyer's diligence pointing it out, a new investor reviewing the books, or a banker's annual review noticing the trend. By then, several years of margin have been lost and rebuilding takes years more.

The owners who don't experience this aren't smarter or more disciplined. They're the ones who looked at one ratio — opex as a percentage of revenue — once a quarter for ten years. That's the entire defense. Five hours of attention per year stops most of the bleeding before it starts.

For a structured run through opex %, gross margin, EBITDA, and the other ratios that affect operating health, see the Financial Health Check — it computes them all and flags whether your numbers are healthy, drifting, or in trouble.