📌 The short version
A business is limited by exactly one stage at a time, and improving any other stage produces no change in output. Worse than that: adding capacity upstream of the real bottleneck actively reduces throughput, because it floods a stage that already cannot keep up. That is why buying more leads so often makes a business perform worse rather than better. Walk six rungs in order — visibility, conversion, sales, delivery, retention, margin — and stop at the first honest weak answer. That rung is your constraint, and until it is fixed, everything else is decoration.
A contractor came to us wanting to double his ad spend. Business was fine, he said, he just needed more volume.
Before quoting anything, we asked what happened to the leads he already had.
He did not know. So we spent an afternoon finding out — and what we found was that roughly four in ten inquiries were never contacted at all. Not contacted late. Never contacted. They arrived by web form during a workday, landed in an inbox nobody owned, and quietly died there.
A $40,000 mistake, narrowly avoided
Here is what would have happened if we had simply taken his money.
He was spending roughly $3,500 a month. Doubling that for a year is another $42,000. Those extra leads would have arrived into the same unowned inbox, handled by the same overloaded person, alongside the leads already being missed.
So he would have paid $42,000 to increase the number of inquiries he ignored.
Instead, we spent about two weeks on something considerably less exciting. One named person owning inbound. An instant alert on every form. A simple log. Nothing clever, nothing expensive.
Revenue rose about 30% on the same ad spend. And now the extra budget is worth spending — because there is finally a system on the other end to catch it.
He did not have a lead problem. He had a follow-up problem wearing a lead problem's clothing, and it would have cost him $42,000 to keep the disguise on.
That story is unusual only in that we caught it before the invoice. The pattern itself is the single most common thing we see.
Why more leads actively made it worse
Most people accept that fixing the wrong thing is wasteful. What is less intuitive — and much more important — is that it can be actively harmful.
Watch what happens in this pipeline. Work flows in from the left. One segment is narrower than the others.
Throughput is set by the narrowest point
Adding more at the front does not widen the middle
Pour twice as much in the left and the right side does not change. The queue in front of the narrow point simply grows — and everything waiting in that queue gets slower, including the good work.
Now translate that back to the contractor. Doubling his leads did not create more capacity to answer them. It created a longer queue in an inbox nobody was watching.
And here is the genuinely damaging part: response time gets worse for everyone, including the high-value inquiries that would otherwise have closed. The queue does not politely sort itself by profitability. So the business does not merely waste the new spend — it degrades the performance of the spend that was already working.
That is why this is not a story about efficiency. It is a story about not actively harming yourself.
This is a forty-year-old idea, and it is not mine
None of the above is a marketing insight. It comes from manufacturing.
In 1984 the physicist Eliyahu Goldratt published The Goal, a business novel about a plant manager trying to save a failing factory. The framework inside it — the Theory of Constraints — has been applied to factories, hospitals, software teams and supply chains ever since.
Its central claim is blunt: a system's output is limited by exactly one constraint at a time. Improving anything else does not increase output. Goldratt was explicit that focusing on a non-constraint is not merely low-value but inherently wasteful.
The framework runs in five steps, and the order is the entire point:
| Step | In a factory | In your business |
|---|---|---|
| 1. Identify | Find the machine with work piled up in front of it | Walk the ladder below and find the first genuinely weak rung |
| 2. Exploit | Stop that machine idling — no lunch breaks, no setup waste | Get more from what you already have. Free, so it comes first |
| 3. Subordinate | Slow the rest of the line to match it | Stop flooding the bottleneck. Sometimes that means pausing spend |
| 4. Elevate | Buy another machine | Hire, buy tools, increase budget. Expensive — hence fourth |
| 5. Repeat | The constraint moves elsewhere | Re-run the ladder. Solving one rung always exposes another |
Look at where elevate sits. Adding capacity — hiring, buying, spending more — is step four of five. Most businesses start there, because it is the step you can buy.
The line worth remembering
Goldratt's point was that adding capacity upstream of a bottleneck does not just waste money — it reduces throughput, because it starves the constraint of the ability to work through what it already has. The system performs worse because of your effort to improve it, not despite it.
The six-rung ladder
Goldratt wrote for factories with physical machines. A business that sells services has a different chain, but it behaves identically.
Here are the six rungs. Walk them in order and stop at the first honest weak answer — the order matters enormously, and I will explain why in a moment.
The constraint ladder
Stop at the first honest "no"
Everything below your weakest rung is currently irrelevant. Not unimportant — irrelevant, until the rung above it is fixed.
Why order matters: an upstream weakness routinely disguises itself as a downstream one. A business with poor visibility often has a terrible close rate too — not because its salespeople are bad, but because the handful of inquiries it does receive are badly matched. Fix visibility and the close rate can improve on its own, without anyone touching the sales process.
Diagnose downward. Fix upward.
Find your constraint in two minutes
Six questions. Answer honestly rather than optimistically — the tool is only as good as the candour, and nobody else sees the result.
Free diagnostic · Nothing to fill in
Which rung is limiting you?
Runs in your browser. No signup, no email, nothing stored or sent anywhere.
Whatever the tool said, read that rung's section below. The detail is where the actual work is.
Rung 1 — Visibility
Do enough of the right people know you exist?
UpstreamSymptoms it is genuinely your constraint
- You could comfortably serve twice the work you currently get.
- Your close rate is decent — the problem is arithmetic, not persuasion.
- You are invisible for searches you should obviously own.
- New customers arrive almost entirely by referral, with nothing else contributing.
What actually fixes it
- Finish your Google Business Profile. Free, and for local businesses it is the highest-return asset that exists. Pick the most specific primary category available — most businesses pick something far too broad.
- Build a review habit. Volume, recency and response rate all matter independently.
- Referral partnerships. Businesses serving the same customer without competing for it.
- Paid search, once you know your close rate and what a customer is worth. Not before.
- Local search work for the compounding version, which takes three to six months and then keeps producing.
The trap
- Assuming visibility is the problem because it is the most comfortable problem to have. It is the only rung where the fix is "spend money," which is precisely why it gets over-diagnosed.
Rung 2 — Conversion
Do the people who find you actually contact you?
Cheapest to fixSymptoms
- Traffic looks healthy. The phone does not ring in proportion.
- People land, scroll and leave without contacting anyone.
- Mobile visitors bounce noticeably harder than desktop.
- You have never actually tested your own site on a phone over cellular.
Where sites leak, roughly in order of frequency
- Speed. Mobile visitors abandon slow pages quickly, and this compounds with every other problem below.
- An unclear offer. A visitor should know what you do, where you do it, and what to do next within about five seconds.
- No visible phone number. Sounds absurd. Happens constantly.
- Missing proof. Reviews, photos of real work, credentials, named humans.
- Forms asking too much. Every additional field costs completions.
✅ Why this rung is the best news you can get
It is the cheapest rung to fix and the fastest to show results. Average site conversion sits around 2–3%; strong ones exceed 10%. Same traffic, three times the leads, no extra ad spend. If the diagnostic put you here, you got lucky. Our conversion work is almost entirely this rung.
Rung 3 — Sales
Do the people who contact you buy?
Most under-measuredSymptoms
- Plenty of inquiries. Quotes go out. Not much comes back.
- You cannot state your close rate without checking — or you have never checked.
- Follow-up happens when someone remembers, which is a polite way of saying rarely.
- People "go quiet" and nobody chases them.
Where deals actually get lost
- Response time. The single biggest factor, and it is free to fix. Minutes matter enormously; hours are fatal. Measure yours honestly for one week before assuming it is fine.
- Follow-up depth. Most sales require five or more touches. Most people stop at two, then conclude the lead was cold.
- Talking rather than asking. Pitching before you understand the problem loses winnable work.
- Price without value. A number in isolation is always too high. A number attached to an outcome is a decision.
- No defined next step. Every conversation should end with something scheduled.
⚠️ The multiplication that should stop you cold
At a 15% close rate, 85 of every 100 leads you pay for produce nothing. Double your ad spend and you have doubled the 85. Lift the close rate to 30% instead and you have doubled revenue without buying a single extra lead.
This is the rung where "we need more leads" is most often flatly wrong — and most expensive to get wrong.
Rung 4 — Delivery
Can you serve them well, consistently, at volume?
The one people hideSymptoms
- You are turning down work, or quietly quoting high to discourage it.
- Lead times are stretching and customers have started mentioning it.
- Quality varies depending on who happens to do the job.
- Complaints have crept up alongside volume.
- You are working evenings to keep the delivered work acceptable.
What fixes it, in order of cost
- Raise prices. Counterintuitive but correct. If demand exceeds capacity, price is the fastest lever — and it improves margin at the same time.
- Document the process. A written checklist turns "they cannot do it like me" into "they can now."
- Remove the low-value work that consumes capacity without paying for it.
- Subcontract the overflow before hiring, because it is reversible.
- Then hire — expensive, slow, and correct only after the first four.
⚠️ This is the dangerous rung
Marketing into a delivery constraint is how good companies acquire bad reputations. You win more work, deliver it worse, collect the reviews that reflect that, and permanently damage the asset that was making rung one work. The damage outlasts the campaign by years.
Rung 5 — Retention
Do they come back, and do they refer?
Most neglectedSymptoms
- Revenue is flat despite steady new customers — you are refilling a leaking bucket.
- Customers seem happy and simply never return.
- Referrals happen by accident rather than by system.
- You have a customer list doing absolutely nothing.
What fixes it
- A reason and a reminder. Seasonal service, annual check-up, replacement cycle. Book it at the point of sale, not later.
- Proactive contact. Call before they need you. An hour a week, startlingly effective.
- Ask for referrals specifically. "Send people my way" produces nothing. "Do you know one person dealing with this right now?" produces names.
- A review system, which serves rung five and rung one simultaneously — see the method here.
Worth knowing: acquiring a new customer costs several times more than keeping an existing one, and most churn is caused by silence rather than dissatisfaction. Your customers did not leave. They forgot, and you never reminded them.
Rung 6 — Margin
Does any of this actually leave you money?
Volume never fixes itSymptoms
- Busy, profitable on paper, no money in the bank.
- Growth has not improved how it feels to run the business.
- You have never calculated profit per job type.
- You suspect some customers lose money and have never checked which.
What fixes it, fastest first
- Raise prices. The entire increase is margin, because delivery cost does not change.
- Cut the cost of delivery — rework, waste, scheduling gaps. Rework is usually the quiet killer.
- Fire unprofitable work. Painful, immediate, and it frees capacity for better jobs.
- Reduce overhead last, because it is the smallest lever and the most demoralising.
⚠️ The multiplier that runs backwards
Volume never fixes a broken margin. It multiplies it. Selling more of something unprofitable loses money faster, with more effort and more risk. If this is your rung, every other improvement makes your situation worse until you fix it.
The hidden seventh rung
There is one more, and it does not appear on the ladder because it does not sit at a stage. It sits across all of them.
Sometimes the constraint is you.
Every quote, every decision, every problem routes through one person. So the business cannot move faster than that person can think — no matter how strong the other six rungs are. The symptoms are distinctive: you are busy but not growing, the team waits on you, and nothing progresses while you are away.
The test is uncomfortably simple. Could the business run for two weeks without you? If not, you do not own a business. You own a demanding job with worse benefits and more risk.
The fix, and why the usual version fails
Delegate decisions, not tasks. Handing over a task while keeping the decision leaves the bottleneck exactly where it was — you are still in the loop, just with extra steps.
Then: track your time for a week in writing, list everything only you can genuinely do, challenge half of it, document before delegating, and accept 80%. Waiting for someone to do it precisely your way guarantees you keep doing it forever.
Worth noting for anyone thinking about an eventual exit: reducing owner dependence is also the single biggest driver of what your business is worth. A company that requires you is difficult to sell at any price.
Want a second opinion on your diagnosis?
Tell us which rung the tool landed on and we will give you an honest read — including when the answer is that you do not need marketing right now, which happens more often than you would expect from an agency.
What to do once you know
Knowing your constraint is most of the value. But the order of what comes next matters just as much, and it is where Goldratt's sequence earns its keep.
Step 0 · Already done
Identify the constraint
This is the step you completed above, and it is the one almost everybody skips. Goldratt put it first for a reason: every subsequent step is wasted effort if it is aimed at the wrong stage. If the diagnostic returned "not sure" on a rung, treat that rung as the constraint until you have measured it for thirty days and proven otherwise.
Step 1 · This week
Exploit before you elevate
Get more out of what you already have, because it is free. If sales is your rung, fix response time and follow-up cadence before hiring anyone. If delivery is your rung, remove the low-value work eating capacity before adding headcount. Most constraints have 20–30% of unused capacity sitting inside them.
Step 2 · This month
Subordinate the rest of the business
Make everything else serve the constraint — including, when necessary, deliberately slowing upstream activity. If delivery is straining, pausing a campaign is not defeat. It is the correct move, and it protects the reviews that make rung one work.
Step 3 · This quarter
Then, and only then, elevate
Now add capacity. Hire, buy the tool, raise the budget. It works now because you are expanding the actual limit rather than feeding a queue. This is the step most businesses start with, which is why it so often disappoints.
Step 4 · Next quarter
Re-run the ladder, because it moved
Fix one rung and the constraint relocates. That is success, not failure. Solve visibility and you will discover a conversion problem that was always there but invisible, because too few people were arriving to reveal it.
Three businesses, three different answers
The same ladder produces completely different conclusions depending on where a business actually sits. Here are three we have walked recently, lightly disguised.
The dental practice that wanted more ads
Plenty of new patient inquiries. Good website traffic. The owner wanted to expand into paid search because growth had flattened.
Rungs one, two and three were all healthy. Rung four was the problem: the schedule was already full six weeks out, and the practice had started quietly discouraging new patients to protect existing appointments. Adding demand to that would have stretched lead times further and generated exactly the reviews you do not want.
The answer was not marketing. It was pricing and scheduling — raising rates on the most in-demand appointment types and restructuring the calendar. Revenue rose without a single additional inquiry, and the paid search conversation became sensible six months later.
The e-commerce store with a traffic obsession
Strong traffic, healthy ad account, obsessive attention to visitor numbers. Conversion sat around 1.1%, which the owner considered normal because he had never benchmarked it.
Rung two, unambiguously. Checkout required account creation, shipping cost appeared only at the final step, and mobile page load was over five seconds on cellular.
Fixing those three things roughly doubled conversion. Same traffic, same spend, considerably more revenue — and critically, it made the existing ad budget profitable rather than marginal, which changed what the business could afford to do next.
The consultant who was the constraint
Every rung looked healthy. Good visibility, strong close rate, delighted clients, excellent margins, plenty of referrals.
And revenue had not moved in three years, because every proposal, every delivery and every decision ran through one person working at full capacity. The seventh rung, exactly as described above.
There was no marketing fix available. The work was documentation, a first hire, and the genuinely difficult habit of delegating decisions rather than tasks. Slower than a campaign, and the only thing that would have worked.
The pattern worth noticing
In all three cases the owner arrived asking for marketing, and in none of them was marketing the answer. That is not an argument against marketing — it is an argument for diagnosing before prescribing. We eventually did paid work for two of these three, and it worked precisely because the constraint had been cleared first.
How to know it actually worked
One test, and it is unforgiving: total throughput rises. Not the local metric — the whole system's output.
If you lift your close rate from 20% to 30% and revenue does not move, the close rate was never your constraint, and something downstream is now absorbing the difference. That feedback arrives fast, usually within weeks, and it is the main practical reason this framework is worth using at all. The system tells you honestly whether you improved the right thing.
Why fixing the wrong rung feels like progress
Here is the psychological trap, and it is worth naming because it catches thoughtful people.
Fixing a non-constraint produces visible improvement in that stage. Your traffic really does go up. Your close rate really does improve. The dashboards move. It feels like progress, and everyone involved has evidence they did good work.
But total output does not change, because the bottleneck is still the bottleneck. So a business can spend a full year genuinely improving things and end the year in exactly the same place — with better metrics and identical revenue.
The most demoralising year a business can have is one where everything improved except the only number that mattered.
There is a second reason it happens: we fix what we enjoy fixing. A marketing-minded owner diagnoses a marketing problem. An operations-minded owner diagnoses an operations problem. The diagnosis follows the diagnostician's comfort, which is precisely why walking the ladder in a fixed order matters more than intuition.
And frankly, agencies are not immune. Ask a firm that only sells traffic what your problem is and you will hear a traffic answer. That is not always dishonesty — it is more often just the limit of what they can see. It is also why we put the answer desk on our own site: some of the answers it gives point people away from hiring us.
The compounding argument for doing this properly
Four modest gains, one large result
Once the constraint is cleared, improvements start multiplying instead of adding — because each stage's output is the next stage's input.
1.10 × 1.10 × 1.10 × 1.10 = 1.4641 — roughly a 46% revenue increase from four ordinary changes.
Chasing a single 46% jump in one stage costs a fortune and usually fails. Four 10% improvements are boring, achievable in a quarter, and mathematically superior — provided the constraint is not eating them first.
Questions we get about this
Because adding capacity upstream of a bottleneck sends more work to a stage that already cannot keep up. In Goldratt's framing, that reduces throughput rather than increasing it.
Practically: if your team already misses half its inquiries, doubling inquiries means missing more of them. Response times worsen for everyone, including the good leads that would have closed. You paid for volume and received a slower, more chaotic version of the same revenue.
Walk the ladder in order and stop at the first honest no. Do enough of the right people know you exist? Do the people who find you contact you? Do the people who contact you buy? Can you serve them well at volume? Do they come back and refer? Does any of it leave you money?
The first genuinely weak stage is your constraint. Order matters, because a downstream problem often disguises itself as an upstream one.
Fix the higher one first. Upstream weakness frequently causes what looks like downstream weakness — a business with poor visibility often has a bad close rate simply because the few inquiries it gets are badly matched.
Fix visibility and the close rate may improve without anyone touching sales. If both remain weak afterwards, the second one was genuine and now you know for certain.
Yes, applied to a small business rather than a factory floor. Eliyahu Goldratt introduced it in his 1984 book The Goal, built around five focusing steps: identify, exploit, subordinate, elevate, repeat.
It was written for manufacturing, but the underlying logic — that a system's output is limited by exactly one constraint at a time, and improving anything else is wasted effort — transfers directly to how a business generates customers.
Every time you fix one. People find that frustrating; it is actually the sign of success.
Solve visibility and you discover a conversion problem you never had before, because previously too few people were arriving to reveal it. Solve conversion and delivery starts straining. Re-run the ladder quarterly, or after any significant change in volume, pricing or headcount.
This is the hidden seventh rung, and it is very common in owner-operated businesses. The test: could the business run for two weeks without you?
If every quote, decision and problem routes through one person, the business cannot move faster than that person can think — regardless of how strong the other rungs are. The fix is documentation and delegated decisions rather than delegated tasks. Handing over a task while keeping the decision leaves the bottleneck exactly where it was.
Not stop — reconsider scaling. If your constraint sits downstream of visibility, increasing spend buys a bigger version of an existing problem.
Maintaining current activity while fixing the real bottleneck is usually right. Cutting marketing entirely creates a visibility problem you did not have, which just relocates the constraint instead of removing it.
Total throughput rises. Not the local metric — the whole system's output.
If close rate goes from 20% to 30% and revenue does not move, close rate was never the constraint and something downstream is absorbing the difference. That feedback is fast and honest, and it is the main reason the framework is worth using.