How to pick the one bottleneck worth a 90-day sprint

The idea: your business has one narrowest point

Every business that sells something moves buyers through a series of stages, and the total output is limited by the narrowest one. Widening any other stage barely helps.

Think of it like a road with one single-lane bridge. Adding lanes before or after the bridge doesn't move more cars; widening the bridge does. In a small business, the bridge might be too few enquiries, slow replies, weak quotes, or customers who never return. The job of this exercise is to find the bridge.

Step 1: map your five stages

Write these five stages across the top of a page:

1. Attention. People who could buy hear of you: searches, map views, referrals, walk-bys.
2. Enquiry. Some of them contact you: calls, texts, emails, web enquiries, visits.
3. Reply. You respond, and the conversation actually starts.
4. Offer. You quote, propose or present a price, and some accept.
5. Repeat. Customers come back or send someone else.

Some businesses skip a stage. A shop with walk-in buyers may have no formal quote. That's fine: merge stages so the map matches how you actually sell.

Step 2: put a rough number on each stage

Use the last three to six months. Rough is fine.

If a number doesn't exist, estimate it and mark it with a question mark. Unknown numbers are clues in themselves: a business that can't say how many quotes it won last month usually has a follow-up problem.

Step 3: write a plausible level beside each number

For each stage, write down what a reasonable, well-run version of your business could achieve, not a world-class one. Base it on your own best months, on what your team believes is realistic, or on a careful comparison with similar firms you know.

For example:

• Reply: you answer 55% of enquiries within two hours. Your best month was 80%. Plausible: 80%.
• Offer: you win 20% of quotes. In months when the owner followed up personally, it was 30%. Plausible: 28%.
• Repeat: 15% of revenue is from returning customers. For a business with an annual service cycle, plausible might be 30% or more.

Be strict with yourself here. The plausible level should be something you could defend to a sceptical friend.

Step 4: turn each gap into dollars

This is the step most owners skip, and it's the one that settles arguments. Work out what closing each gap would be worth over a quarter.

A worked example for a business with 80 enquiries a month and an average sale of $900:

• Reply gap (55% to 80%). If unanswered enquiries mostly go elsewhere, answering 25% more of them gives 20 more live conversations a month. At a 20% win rate that's 4 more sales, $3,600 a month.
• Offer gap (20% to 28%). Of the roughly 44 enquiries that currently get a timely reply, 8 points more won means about 3.5 extra sales, roughly $3,150 a month. If the reply gap is fixed too, the gain grows.
• Attention gap. Doubling profile views might add 10 enquiries a month, but only half would get a timely reply today, so the gain is about 1 sale, $900.

In this example the reply stage is the narrowest bridge. Fixing it first also makes every later improvement worth more. That's typical: the closer a gap is to the money, the more it's usually worth.

Step 5: check capacity before you commit

A bottleneck you can't staff isn't a sprint; it's a wish. Before you choose, ask:

• Who will own the target figure? One named person.
• Do they have three or four hours a week? If not, what will they stop doing?
• Can the business serve more customers? If you're already turning work away, a lead flow sprint is the wrong choice; fix price, scheduling or capacity first.

If the highest-value gap fails the capacity check, pick the next one down, or make capacity itself the sprint: moving quoting, scheduling or first replies away from the owner is a perfectly good 90-day project.

Step 6: write it down as a charter

The output of this exercise should fit on one page:

• Constraint: first-reply speed on enquiries.
• Baseline: 55% answered within two hours (last three months).
• Target: 80% within two hours, sustained for the last four weeks of the sprint.
• Leading figures: missed calls per week; median reply time.
• Guardrail: close rate must not fall.
• Owner: the office manager.
• Parked until day 90: website redesign, new ad platforms, social media overhaul.

The parked list matters as much as the target. It's your written permission to say "not this quarter" to good ideas that would dilute the sprint.

Common mistakes

• Picking the stage you enjoy. Owners who like design pick the website; owners who like data pick analytics. Let the dollar figures decide.
• Treating attention as the default. "We need more leads" is the most common diagnosis and often the wrong one.
• Ignoring the repeat stage. Past customers are frequently the cheapest revenue available, and the least measured.
• Choosing two bottlenecks. If you truly can't decide, pick the one closer to the money and schedule the other as the next sprint.
• Skipping the parked list. Without it, week four brings a "quick" side project that eats the sprint.

What to do next

If you've run the exercise and the answer is clear, you can run a sprint yourself using the 13-week scoreboard guide. If the answer is unclear, or the numbers are too messy to trust, that is exactly what our sprint diagnostic is for. Either way, the services page explains the three lanes your bottleneck is likely to fall into, and you can call, text or email to talk your table through with a senior strategist.