The Leverage Audit

An owner who decides to grow reaches, almost every time, for the leverage that can be bought: another hire, another truck, another borrowed dollar. The instinct is reasonable and the sequence is wrong, because nothing established that headcount is what holds the business back. The house method runs in the opposite order. Before optimizing anything, audit: inventory where leverage actually lives in the business, find the one place it binds, and fix that node first. The audit tends to return a verdict the owner did not order: the constraint is not in the category he was about to fund.

Leverage, plainly, is anything that multiplies the result of an hour without multiplying the hour. It lives in seven places: labor, capital, code, media, information, relationships, and distribution. Every business runs on some mix of them, whether or not anyone has written the mix down. The audit writes it down. That single page outperforms most strategy documents, because a strategy that does not know where its leverage lives is a wish.

Seven places leverage lives

Labor: other people’s hours, multiplied through delegation and judgment. The oldest form, and the only one that manages you back: every added person consumes some of the attention they were hired to free. The audit question: which skill is actually scarce here, and whose calendar is the queue everyone waits in?

Capital: money commanding assets and people. It compounds, and it converts into every other form on this list. The audit question: where does cash sit idle while the business queues, and what is the queue costing per week?

Code: systems that work while everyone sleeps, at zero marginal cost. The audit question is blunt: count how many times the same fact gets typed into different screens. Every retype is a place where labor is quietly substituting for a system.

Media: published knowledge that compounds into audience. An audience is stored attention and trust, convertible on demand into demand. The audit question: does anything this firm knows compound in public, or does its entire reputation live in one person’s phone?

Information: the record of your own operations — costs, outcomes, choices over time. It is the one dataset no competitor can buy. The audit question: what does this business observe that nobody else can, and is any of it priced in?

Relationships: who returns the call, who vouches, who deals on a handshake. Slow to build, near impossible to copy, and the gate to every closed market. The audit question: which revenue depends on which specific people, and who holds those bonds?

Distribution: the owned route to the customer. Whoever owns the route taxes everyone upstream of it. The audit question: does this firm own its route to the customer, or rent it? Relationships are who takes your call; distribution is who owns the road to the buyer. When the same names answer both questions, write them down twice: the overlap is a finding, not a bookkeeping error.

The permission line

The seven sort along one line, and the line predicts most audit findings. The split has a familiar shorthand: labor and capital require permission — people choose to follow, lenders and investors approve — and both carry a permanent management cost. Code and media ask no one: they replicate at no marginal cost and work unattended. Relationships are the case the split leaves out: neither granted nor copyable, only kept, compounding at the speed of kept promises.

The recurring misallocation follows directly. Firms overbuy the leverage they can purchase and underbuild the leverage they must author. Buying feels like action; authoring feels like risk. A firm that doubled headcount in three years has usually doubled the one form of leverage that requires permission and management, while its operating records, its systems, and its public reputation — the forms that would work unattended — sit exactly where they sat.

One constraint at a time

The inventory is the map. The constraint is the destination. In any sequential business — a pipeline, a shop floor, a hiring funnel — total throughput is set by exactly one binding constraint at a time. Goldratt’s constraint theory made the point for factory floors; it holds anywhere work queues. Effort spent anywhere else is cosmetic: the improved step simply waits faster on the step that binds.

The discipline that follows is strict. Break the flow into small steps, find the largest drop, change one variable at that step, measure, repeat. One variable: a five-front improvement program destroys attribution, and a firm that cannot attribute cannot learn. The test is unforgiving in the way useful tests are: if total output did not move, the thing you improved was not the constraint.

Most businesses do not have a growth problem. They have one constraint and six distractions.

The audit, walked

What follows is the audit run against a composite — a business assembled from the patterns this framework was built to catch, with figures chosen to be typical rather than reported. No client appears here. Call it a mechanical services contractor: twelve million in revenue, seventy people, owner-led, profitable, stuck. The owner’s growth plan is two salespeople and a marketing retainer.

The audit reads differently. Labor: three estimators, every bid queued through them, the owner re-checking every number above fifty thousand dollars. Capital: a credit line untouched on principle while receivables float for fifty-five days. Code: the same job typed into four systems — estimate, proposal, job costing, invoice — by four different people. Media: none; the firm’s reputation lives in the owner’s phone. Information: eighteen years of actual job-cost history in closed folders, the most accurate pricing data in its market, used for nothing. Relationships: three general contractors worth forty percent of revenue, every one of those bonds held personally by the owner. Distribution: the same three names again — the firm reaches its end customers only through those contractors’ bids, and pays for the route in margin. One fact fills two headings, and that is the diagnosis: the firm’s only road to market is held by the same three people its revenue already depends on.

Seven categories, one constraint. Quoting takes nine days, and in a bid market the first credible number frames the job. The two salespeople would pour more demand into a queue that is already the binding node; the marketing retainer would advertise a nine-day response. The fix touches two assets the firm already owns: the job-cost archive becomes estimating templates, estimating becomes review instead of craft, and the quote goes out in two days. In the composite, nothing else in the plan matters until that node moves. The audit does not tell the firm to become a software company. It tells it to stop retyping its own past.

Then the sequence continues, because a constraint removed promotes the next one. With quoting fixed, relationship concentration becomes the live risk: forty percent of revenue held in three personal bonds is a succession problem, not a sales problem. The audit is a loop.

Why owners miss their own constraint

Two reasons. First, the binding constraint is usually attached to the least-loved work. A business tends to be constrained precisely where its owner prefers not to look; the interesting problems get attention on schedule, and the ugly one compounds.

Second, the instrument panel is wrong. Owners audit the P&L, which reports categories of spending, not forms of leverage. The better instrument is the calendar. Read the owner’s calendar as a balance sheet: every recurring meeting is a claim against a system that does not exist yet. Where the owner’s hours pool, the firm is running on its least scalable asset — the owner’s own labor — and the constraint is almost always within walking distance of that pool.

Running the audit

For the owner-operator: run the audit in a week, on paper. Seven headings, one honest page each, and three instruments: the calendar, the retype count, and a table of revenue by named relationship. Then name one constraint, change one variable at it, and let total output — not activity — deliver the verdict. If output does not move, the audit was wrong in a way you can now see; run it again.

For the acquirer or investor: the audit is a diligence lens. Trailing financials price the average; the constraint prices the future. A business whose binding constraint is cheap to remove is systematically underpriced by buyers who value it on history, and a business whose constraint is the owner’s personal bonds is riskier than its margins look. Ask where the leverage lives before asking what the multiple is.

For counsel and advisors: the audit locates what is actually worth protecting. Relationship concentration is a key-person exposure before it is a valuation discount; an operating-data archive is an asset the schedules rarely name; channel agreements decide whether distribution is owned or rented. Papering a business without knowing where its leverage lives secures the assets that are easy to name and leaves the ones that carry it unprotected.