Theory of Constraints for a Service Business: 5 Checks

The theory of constraints says every system's output is capped by one weakest link, so improving anything else won't grow results. In a service business, that link almost always sits in one of five places: lead flow, conversion, delivery capacity, owner time or cash. Find the one that's binding, break it with the five focusing steps, and wrap it in a 90-day target with one owner and one deadline.

Eliyahu Goldratt wrote the theory for factories. This guide rewrites it for coaches, consultants, agencies, practitioners, real estate and mortgage pros, and local service owners. Every test runs on numbers you already have.

The 30-minute constraint check (copy this)

Pull your last 90 days of numbers, then answer each line yes or no:

1. Lead flow: Does delivery have room for more clients while your sales calendar has open slots?
2. Conversion: Are you getting enough qualified conversations but closing too few, or watching proposals sit for weeks?
3. Delivery capacity: Is the backlog growing, turnaround slipping, or are you turning work away?
4. Owner time: Would sales or delivery stall if you were unreachable for a week?
5. Cash: Is there a spend you're confident would pay back (a hire, ads, a tool) that you can't fund right now?

More than one yes is normal. Only one of them is the binding constraint. The rest of this guide shows you how to tell which, and what to do about it.

The theory of constraints in plain English

Goldratt introduced the idea in his 1984 business novel The Goal. Picture your business as a chain. It's only as strong as its weakest link, and pulling harder on the strong links does nothing.

The method has five focusing steps: identify the constraint, exploit it, subordinate everything else to it, elevate it, and repeat (EBSCO). In 1992, two steps were added in front: identify the goal of the system and decide on the global measures you'll judge it by (The Decision Lab).

That addition matters for owners. Before you hunt for the constraint, decide what the business is trying to produce. For most service firms, the goal is profitable revenue delivered and collected. The measure is something like monthly cash collected or gross profit. Without a goal, every annoyance looks like a constraint.

Constraint vs. bottleneck

People use the words interchangeably, but the distinction is useful. A bottleneck is any step where work backs up. The constraint is the one step limiting the output of the whole business right now.

You can have three bottlenecks and one constraint. Fixing a bottleneck that isn't the constraint feels productive and changes nothing on the P&L.

Why it fits a service business (and why services are harder)

TOC arrived in services late. A 1996 health-care paper noted that no study had yet addressed applying TOC in service organizations (Emerald). Since then, a literature review concluded TOC applies to companies in both production and services (ScienceDirect), researchers have applied it to banking (Springer), and John Ricketts wrote a full book on it for services, Reaching The Goal (IBM Press, 2008) (Business901). Recent research even pairs TOC with queuing theory to analyze bottlenecks in multi-stage service delivery (ScienceDirect, 2025).

Two things make services harder than a factory:

• Capacity is perishable. A factory can build stock in a slow month. You can't stockpile billable hours. An empty Tuesday is gone for good.
• The work is invisible. Factory work-in-progress sits on the floor. Yours sits in inboxes, CRMs, half-written proposals and your head.

So you have to make the pile visible before you can find it.

Translate the factory terms into your numbers

Researchers mapping TOC to banking defined throughput as the money from services rendered, inventory as the money needed to generate that profit, and operating expense as what funds the expenses (Springer). Here's the version for a service owner:

| TOC term | What it means in a service business | Number to track |
|---|---|---|
| Throughput | Revenue from work delivered and collected | Monthly cash collected, gross profit |
| Inventory / WIP | Work started but not finished or not paid | Backlog, open proposals, unbilled work, unanswered leads, items waiting on the owner |
| Operating expense | What it costs to run the business | Payroll, contractors, software, rent, ad spend |
| Constraint | The step that caps throughput | Whatever has the biggest pile in front of it |
| Drum | The pace the constraint sets | Clients or deliverables the constraint can handle per month |
| Buffer | Ready work queued so the constraint never sits idle | Days of prepped work waiting |

The rule of thumb that does most of the work: the constraint is where work piles up. EBSCO's advice for finding it says the same thing: look for areas with "an excess of work in progress" (EBSCO).

The 5-place theory of constraints check for a service business

Before running the tests, pull these numbers for the last 90 days. Rough is fine; guessing is not.

• Qualified leads or sales conversations per month
• Close rate (deals won divided by qualified calls or proposals)
• Average days from first call to signed deal
• Open proposals right now, their age and total value
• Active clients and the maximum you can serve well
• Backlog in weeks (signed work not started, divided by weekly delivery capacity)
• Average turnaround vs. what you promise
• Items waiting on your approval or input
• Cash on hand, monthly operating expense, average days to get paid
• Unbilled work (done but not invoiced)

Each check below has a symptom, the number to look at, and a pass/fail test. The thresholds are my rules of thumb, not industry standards. Adjust them to your model.

1. Lead flow

Symptom: Delivery has room. Your team or calendar could take more clients next month, but there aren't enough qualified conversations to fill it.

Number: Qualified conversations per month vs. the number you'd need to fill open capacity at your current close rate.

Test: Divide open capacity (new clients you could take next month) by your close rate. That's the number of qualified conversations you need. If you're getting less than about 80% of that, lead flow is a candidate.

For example, imagine a financial coach with room for 6 new clients a month and a 30% close rate. She needs 20 qualified calls. If she's getting 9, more leads is the move. If she's getting 25 and still signing only 3, the problem is somewhere else.

2. Conversion

Symptom: Plenty of qualified calls, but few close, or proposals sit for weeks.

Number: Close rate, sales cycle length, count and age of open proposals.

Test: If you have enough qualified conversations to fill capacity (using the math above) but sign fewer clients than delivery can handle, conversion is a candidate. Second flag: open proposals older than twice your normal sales cycle. That's inventory rotting on the shelf.

Watch for a disguised version. A weak or unclear offer looks like a sales problem. If prospects keep saying "let me think about it," the constraint may be the offer, the price or the follow-up, not the pitch.

3. Delivery capacity

Symptom: Backlog growing, deadlines slipping, quality dipping, telling prospects "we can start in six weeks," or turning work away.

Number: Backlog in weeks, turnaround vs. promise, how fully booked your key delivery people are.

Test: If backlog has grown two months in a row, or turnaround is running past your promise, delivery is a candidate. If you're actively turning away work you'd want, it's almost certainly the constraint.

Inside delivery, find the single person or step everything waits on: the senior strategist who reviews every deliverable, the one loan processor, the lead designer. That's usually where the pile is.

4. Owner time

Symptom: Work waits on you. Approvals, client escalations, sales calls, quotes, hiring decisions.

Number: Count the items waiting on your input right now and how long they've waited.

Test: The vacation test. If you disappeared for seven days with no phone, would sales, delivery or collections stop? If yes, owner time is a candidate. It's also often the root cause hiding behind a delivery or conversion constraint.

Two tools help. Small Business Coach suggests a two-color calendar audit that separates admin work from judgment work (Small Business Coach). Do it for one week: one color for hours only you can do, another for everything else. Jeff Meade names three kinds of bottleneck (team inefficiency, systems inefficiency and approval workflows) and draws a useful line between keeping control and becoming the controls everything has to pass through (Meade).

5. Cash

Symptom: You know a spend would pay back, like a hire that frees delivery or an ad budget with proven returns, but you can't fund it. Or you're profitable on paper and short on cash because clients pay slowly.

Number: Months of operating expense in cash, average days to get paid, unbilled work.

Test: If you've found the fix for one of the other four and the only reason you aren't doing it is cash, cash is binding. Second flag: if unbilled work plus receivables older than 30 days exceeds one month of operating expense, collections are throttling you.

Cash shows up in the research too. In the banking study, the most common constraints were administrative (regulations and policies) and lack of technology, both tied to scarce cash flow (Springer).

How to pick the one binding constraint

Most owners will say yes to two or three checks. Here's how to decide.

Work the chain in this order

Run the checks in this sequence and stop at the first clear fail:

1. Delivery first. If delivery is at or over capacity, that's the constraint. More leads or better sales just deepen the backlog.
2. Then owner time. If delivery is stuck because work waits on you, the real constraint is owner time.
3. Then conversion. If delivery has room and you have enough qualified conversations, the leak is in sales.
4. Then lead flow. If delivery has room and conversion is healthy, you need more conversations.
5. Cash underneath. Cash is binding only when it's the reason you can't execute the fix for one of the others.

Why start with delivery? Because it's the one place where solving an upstream problem actively hurts you. Selling more into a full delivery team buys you refunds, burnout and reviews you don't want.

Ask the doubling question

For each candidate, ask: if this doubled next month and nothing else changed, would cash collected rise next quarter? Only the constraint passes. Doubling leads into a full delivery team: no. Doubling delivery capacity with an empty pipeline: no.

Separate symptoms from causes

"We're always behind" is a symptom. "Every deliverable waits two days for my review" is a cause. Keep asking "what is this waiting on?" until you hit a specific step, person or rule.

Look for policy constraints

Often the constraint isn't a resource. It's a rule, usually one you made yourself:

• "I approve everything that goes to a client."
• "We only take referrals."
• "Only I run sales calls."
• "We have a minimum project size."
• "We don't take deposits."

A widely shared answer on TOC in services lists policy constraints alongside capacity, demand swings, information and market constraints (Quora). Policy constraints are the cheapest to break. They cost a decision, not a hire.

The five focusing steps, rewritten for owners

Here are the five steps with a service example for each. The example is hypothetical: imagine a six-person marketing agency where every strategy deck waits on the founder's review, and the backlog has grown for three months.

Step 1: Identify the constraint

Run the check above and name it in one sentence: "Client deliverables wait an average of four days for founder review." Not "operations." Not "we're busy."

Step 2: Exploit it (get the most out of what you already have)

Before spending money, make sure the constraint never wastes time. For the agency: block two 90-minute review windows a day, batch reviews, and require a one-page summary with every submission so a review takes 15 minutes instead of 45. Strip everything from the founder's week that isn't review or sales.

Exploiting means protecting the constraint's time: no meetings it doesn't need, no rework, no waiting on missing inputs.

Step 3: Subordinate everything else to it

Let the constraint set the pace of the whole business. Don't start more work than it can finish. For the agency: sales stops promising two-week turnarounds until reviews catch up, and the team finishes in-flight projects before starting new ones.

This is drum-buffer-rope in plain English:

• Drum: The constraint sets the pace. If the founder can review 12 decks a week, the business delivers 12 decks a week.
• Buffer: Keep a small queue of ready-to-review work in front of the constraint so it's never idle. Remember, its idle hours are gone for good.
• Rope: Tie the start of new work to the constraint's pace. Don't sell or start more than it can handle.

Step 4: Elevate it (add capacity)

Only now do you spend money or make structural changes. For the agency: write a review checklist as an SOP, train a senior strategist to approve standard decks, and keep only high-stakes work for the founder.

Other ways to elevate: hire, raise prices so the same capacity produces more revenue, narrow the offer, or put AI on the repeatable work.

Step 5: Repeat

Once reviews flow, the constraint moves. Maybe the agency can now deliver more than it sells, so lead flow becomes the limit. Go back to step 1, and retire any rules you built around the old constraint so they don't become the new one.

Where AI agents fit

AI is most useful at steps 2 and 4. When you exploit, AI takes prep work off the constraint: first drafts, summaries, follow-ups, call notes. When you elevate, it takes repeatable work off the plate entirely.

Match the agent to the constraint:

• Lead flow: content, SEO and paid ads
• Conversion: sales follow-up and proposal drafts
• Delivery capacity: SOPs and project management
• Owner time: SOPs, project management and drafting
• Cash: financial planning

I built the Two Comma Boss AI agents myself for exactly this job. You can see how an AI business operating system with 8 agents covers content, sales, SOPs, SEO, paid ads, project management, financial planning and wealth strategy. The rule either way: put AI on the repeatable work around the constraint, not on the judgment calls the constraint exists to make.

Turn your constraint into a 90-day target

My coaching method in one line: find the one binding constraint holding the business back, decide on a 90-day target, put AI agents on the repeatable work, and turn the extra revenue into lasting wealth.

Ninety days is long enough to move a real number and short enough to keep urgency. It's also a sensible interval to re-run the check, because a broken constraint tends to move.

The 90-day target template

Copy this and fill in every line:

```
Constraint (one sentence):
Metric that proves it's moving:
Baseline today:
90-day target:
Owner (one name):
Deadline (a date):
Weekly check (day, time, 15 minutes):
First three actions (each with an owner and a date):
```

Rules that make it work:

1. One metric. If you track five, you'll move none.
2. One owner. A name, not a department.
3. One deadline. A date on the calendar.
4. A weekly 15-minute check. Number up, down or flat, and what happens next week.

This is the specific-steps-no-fluff test: if an action can't have an owner and a deadline, it doesn't make the cut.

Example targets by constraint

These numbers are illustrative. Set yours from your own baseline.

| Constraint | Metric | Example 90-day target |
|---|---|---|
| Lead flow | Qualified calls per month | From 9 to 20 |
| Conversion | Close rate; stale proposals | From 15% to 25%; zero proposals older than 30 days |
| Delivery capacity | Backlog in weeks | From 6 weeks to 2 |
| Owner time | Items waiting on the owner | From 40 to under 10; pass the vacation test |
| Cash | Average days to get paid | From 45 to 20; unbilled work under one week |

If you want help running this, here's how I coach service-business owners: find the one binding constraint, set a 90-day target, use AI to scale. For a full do-it-yourself version, my free ten-chapter book The Art of Building an Empire ends in a 90-day blueprint you can fill in.

When the constraint moves (it will)

Breaking a constraint doesn't end the work. It moves the limit somewhere else. That's the point of step 5.

Signs the constraint has moved:

• The pile in front of the old constraint shrinks, and a new pile appears elsewhere.
• The old constraint sometimes sits idle, waiting on work.
• Revenue jumps, then flattens again.

Common handoffs to watch for:

• Fix delivery, and lead flow becomes the limit. You can serve more clients than you're selling.
• Fix lead flow, and delivery or conversion becomes the limit. More calls than you can close or serve.
• Fix owner time in delivery, and owner time shows up in sales. You freed your hours, and now every deal still needs you on the call.
• Fix any of them, and cash can become the limit. Growth needs funding before it pays back.

Put a quarterly reminder on your calendar to re-run the five checks. Each quarter, start fresh from the numbers, not from last quarter's answer.

Mistakes that keep owners stuck

• Fixing everything at once. Five projects means none gets finished. Pick one.
• Buying leads when delivery is the constraint. You pay to make the backlog worse.
• Cutting cost at the constraint. Letting go of the one processor everything waits on saves a salary and caps revenue.
• Expecting growth from improving a non-constraint. A faster step that feeds a full queue just makes a bigger queue.
• Hiring before exploiting. Squeeze waste out of the constraint first, so the hire adds capacity instead of covering for it.
• Fixing the step you enjoy. Owners who love marketing tend to diagnose a marketing problem. Let the numbers decide.
• Treating your own policy like a law. If you made the rule, you can change it this afternoon.
• Diagnosing from memory. Pull the numbers. Gut feel is how owners end up fixing the wrong link.

Where the extra revenue goes

Breaking a constraint should raise revenue. That's the start, not the finish. Revenue isn't wealth. Wealth is ownership: the systematic conversion of today's cash flow into assets that support your life tomorrow.

I break that into six wealth phases: Foundation, Accumulation, Acceleration, Strategic Reserve, Financial Independence and Legacy. Knowing your phase tells you where the extra revenue from this quarter's work should go. Here's how revenue becomes wealth, including the Strategic Reserve and your next money move. None of this is personalized investment or tax advice, and no strategy guarantees returns.

Your next step

Do this in order this week:

1. Pull the ten numbers listed above for the last 90 days.
2. Run the five checks and work the chain from delivery to cash.
3. Name your one binding constraint in a single sentence.
4. Fill in the 90-day template with one metric, one owner and one deadline.
5. Book your first 15-minute weekly check.

Then go deeper. Read the free book, The Art of Building an Empire. It walks from six figures toward $100 million and ends in a 90-day blueprint you can use on the constraint you just found.

If you'd rather pressure-test your diagnosis with someone, apply for a free 30-minute consultation. It's a working session on your numbers, not a pitch.

Frequently asked questions

What is the theory of constraints in simple terms?

It's the idea that every system has one weakest link that limits its total output. Strengthening any other link does nothing. You find that one constraint, get the most out of it, organize everything else around it, add capacity to it, and then look for the next one.

Can the theory of constraints be applied to a service business?

Yes. Research has applied it to health care, banking and multi-stage service delivery, and a literature review concluded it works in both production and services. The difference is that service capacity is perishable and the work-in-progress is invisible, so you track backlog, open proposals, unbilled work and items waiting on the owner instead of parts on a factory floor.

What are the five focusing steps?

Identify the constraint, exploit it (get the most from current capacity), subordinate everything else to it, elevate it (add capacity), and repeat. In 1992 two steps were added in front: define the goal of the system and decide how you'll measure it.

What's the difference between a bottleneck and a constraint?

A bottleneck is any step where work backs up. The constraint is the single step limiting the output of the whole business right now. You can have several bottlenecks but only one binding constraint, and only fixing that one raises revenue.

How do I know if I'm the bottleneck in my own business?

Run the vacation test: if you were unreachable for seven days, would sales, delivery or collections stop? Then count the items waiting on your approval right now. If work routinely waits on you, owner time is likely your constraint, often hiding behind what looks like a delivery or sales problem.

What is a policy constraint?

A rule, usually one the owner created, that limits output more than any resource does. Examples include approving every client deliverable yourself, taking referrals only, or being the only person allowed to run sales calls. Policy constraints are the cheapest to break because they cost a decision, not a hire.

Does drum-buffer-rope work in services?

Yes, in plain terms. The drum is the pace your constraint can handle, the buffer is a small queue of ready work in front of it so it never sits idle, and the rope ties new sales and project starts to that pace so you don't take on more than the constraint can finish.

What happens after you fix a constraint?

The constraint moves somewhere else. Fixing delivery often makes lead flow the limit, and fixing lead flow often exposes delivery or conversion. Re-run the five-place check every quarter and set a new 90-day target on whatever is binding now.