How to map your business problems and find the revenue leaks

Most owners can tell you roughly how much revenue the business made last month. Far fewer can say how many people asked for work and never heard back, how many quotes went quiet, or how long a finished job sat before anyone sent the invoice. Money disappears in those gaps, and nobody notices because no single report covers them.

Mapping fixes that. You draw the path a customer takes from first contact to repeat purchase, put numbers on each step, and look for the places where people drop out or time piles up. This guide walks through the method we use, with worksheets you can copy and a worked example with the arithmetic shown. It is written for service businesses such as trades, clinics, studios, agencies, and consultancies, but it applies to anyone who sells work that has to be scheduled, quoted, delivered, and invoiced.

What counts as a leak

A leak is money you earned, or could reasonably have earned, that never reached your bank account because of a gap in how the work moves. It is not about how hard anyone worked. Four kinds are worth separating, because each is measured and fixed differently.

Delayed revenue is a cash flow problem, not a revenue problem, so keep it in its own column. Collecting faster does not make the business earn more. It means you need less of your own money to run it.

Why mapping beats guessing

Owners tend to blame the stage they handle personally. Whoever spends the day on quotes thinks the problem is price. Whoever answers the phone thinks it is lead quality. A map built from timestamps and counts takes opinion out of the discussion.

The idea is not new. Lean manufacturing calls it value stream mapping, a method codified in the 1998 workbook Learning to See by Mike Rother and John Shook, drawing on mapping practice at Toyota. The part worth borrowing is that the map follows the whole flow from first request to delivery, and records two times for each step: how long the work takes, and how long it waits before the next step begins. The second number is the one people rarely write down, and it is where this guide spends most of its time.

The six stages most service businesses share

Your business will have its own names for these. Rename them, add steps, or drop one that does not apply. What matters is that the path is complete, with no stage missing between first contact and repeat sale. We will use inquiry, booking, quote, delivery, invoice, and repeat.

1. Inquiry: from first contact to first useful reply

Where the leaks hide:

• Inquiries that arrive in the evening or on a weekend and wait until the next working day or longer.
• Channels nobody checks regularly, such as a web form that emails a shared inbox, social media messages, or voicemail.
• Replies that acknowledge the message but do not move anything forward.
• No single owner, so each person assumes someone else answered.

What to measure: the median time from inquiry to first human reply, split by channel and by time of day, plus the share of inquiries that never get a reply.

The best published evidence on why this matters is a 2011 Harvard Business Review study by James Oldroyd, Kristina McElheran, and David Elkington. They tested how 2,241 US companies handled web inquiries. Only 37 percent responded within an hour, 16 percent responded within one to 24 hours, 24 percent took longer than a day, and 23 percent never responded at all. Among the companies that did respond within 30 days, the average response took 42 hours. In their analysis of lead data from dozens of firms, companies that contacted a lead within an hour were nearly seven times as likely to qualify it as companies that waited even one hour longer, and more than 60 times as likely as those that waited 24 hours or more.

Two cautions. The data is from 2011 and comes from US sales teams, and qualifying a lead is not the same as winning the job. Read it as strong evidence that speed matters, not as a number to forecast with. The sample you are about to pull from your own records is a better guide for your business.

The fix usually starts with a written response-time target, one named owner, and an immediate reply that tells the person what happens next. Acknowledging the message and routing it can be automated without much risk. The reply that moves the conversation forward is worth keeping human, or at least human-approved. The full guide is speed to lead.

2. Booking: from yes to a confirmed time

• Several rounds of email or phone calls just to find a time.
• No-shows and late cancellations.
• Slots opened by a cancellation that are never refilled.
• Bookings made without the details you need (address, job description, access instructions), which forces a second contact.

What to measure: messages or days needed to book a typical appointment, the no-show rate (no-shows divided by booked appointments), empty slot hours per week, and the share of bookings missing required information.

Reminders have some of the cleanest evidence of any fix in this guide. A Cochrane review by Gurol-Urganci and colleagues pooled eight randomized trials with 6,615 participants and found that text message reminders increased attendance at healthcare appointments compared with no reminder or postal reminders, and performed about as well as phone call reminders. Those were medical appointments, so measure your own no-show rate before and after, but the mechanism is simple enough to carry over: a prompt shortly before the appointment.

The usual fix is a self-serve booking link, an immediate confirmation with all the details, a reminder or two before the appointment, a one-step way to reschedule, and a short waitlist that is offered cancelled slots.

3. Quote: from site visit to signed yes

• Quotes that go out days after the visit, once the customer's attention has moved on.
• Quotes sent and never followed up.
• No record of whether a quote was won or lost, or why.
• Each quote rebuilt from scratch instead of from a template.

What to measure: time from visit to quote sent, the share of quotes with a recorded follow-up, win rate for quotes that were followed up versus those that were not, and the stated reasons for lost jobs.

Be careful with that follow-up comparison. The quotes you chose to chase may have been the stronger ones to begin with, so some of the gap in win rates is not caused by the chasing. The worked example below plans on recovering half of the gap, which is a reasonable way to avoid fooling yourself. For ready-made messages and a schedule, see quote follow-up email templates.

4. Delivery: doing the work and keeping people informed

• Customers asking for updates because nobody sent one.
• Details lost in the handoff between the office and the field.
• Scope changes agreed on site but never recorded or billed.
• Rework caused by missing information at the start.

What to measure: status requests per job (count them in your inbox for a few recent jobs), hours of rework, unbilled extras, and how many days late jobs finish.

5. Invoice: from finished work to money in the bank

• Invoices sent days or weeks after the work is done.
• Invoices sent to the wrong contact or missing a purchase order number.
• Reminders that happen only when someone remembers.
• Payment methods that take effort, such as a cheque or a manual bank transfer.

What to measure: days from job complete to invoice sent, days from invoice sent to paid, the share of invoices paid late, and days sales outstanding. Days sales outstanding is accounts receivable divided by credit sales for the period, multiplied by the number of days in the period, and it gives you one number to track over time. For a full invoicing and reminder playbook, see how to get paid faster.

6. Repeat: the customers you already have

• No review request after a finished job.
• No follow-up once the job is paid.
• No reminder at the point where the customer would naturally need you again.
• No record of who referred whom.

What to measure: repeat rate (customers with a second job within 12 to 24 months divided by customers served), the share of finished jobs that got a review request, and the share of past customers contacted in the last 12 months.

How to run the mapping, step by step

Plan on one working afternoon for the first pass and about a month of tracking afterward.

Step 1: write down your stages and owners

Draw your stages as boxes in a row. Under each box write who is responsible today, which tool the work lives in, and what triggers it. If you cannot name an owner for a stage, you have found your first problem.

Step 2: pull a sample of real customers

Take the last 15 to 20 inquiries, not only the ones that became jobs. Include the lost ones, because they hold most of the information. A sample this size will not prove anything statistically, but it is enough to show patterns and it is what a busy owner can finish in an afternoon. If you handle more volume, sample across a full month so slow weeks and busy weeks are both represented.

Step 3: timestamp every stage

Build a spreadsheet with one row per customer and one column per event. Your inbox, calendar, and accounting software hold nearly all of the dates.

Where a date does not exist, leave the cell blank. A blank is data. It means the step did not happen or nobody recorded it.

Step 4: calculate the gaps and the drop-off

For each stage, work out two things: the median gap between that step and the next, and the share of customers who made it to the next stage. Use the median rather than the average, because one job that sat for three months will drag an average well away from what usually happens.

Step 5: put a dollar figure on the biggest leaks

Pick the two or three largest gaps and price them with these formulas:

• Lost revenue = customers lost at the stage × the share of customers at that stage who normally end up paying × average job value
• Cash tied up by delay = monthly revenue × days of delay ÷ 30
• Wasted labour per year = occurrences per week × minutes each ÷ 60 × hourly cost × 52

A worked example

The numbers below are invented for a made-up home-services company with an average job value of $1,800. They are not client data. The arithmetic is the point.

Leak 1: quotes nobody followed up

Followed-up quotes won 60 percent of the time and the rest won 21.4 percent, a gap of 38.6 points. Applied to the 14 quotes with no follow-up, that is 14 × 0.386 = 5.4 jobs. Because the quotes someone chose to chase were probably stronger, we plan on half of that: 2.7 jobs, or about $4,860 a month ($58,320 a year).

Leak 2: slow replies

The 12 inquiries answered late or never booked at 25 percent, against 50 percent for the group answered within a day. If those 12 had booked at 50 percent, that is 6 appointments instead of 3. Each appointment is worth 0.8 quotes × 0.375 win rate = 0.3 jobs on average, so 3 extra appointments is 0.9 jobs, or about $1,620 a month ($19,440 a year).

Leak 3: invoices that start late

Suppose invoices go out a median of 6 days after the job and are paid 31 days after that. That is 37 days from finished work to cash. Monthly revenue of $16,200 means about 16,200 × 37 ÷ 30 = $19,980 is tied up at any moment. Sending the invoice the day the job finishes saves 5 days, which frees 16,200 × 5 ÷ 30 = $2,700 once. That is a one-time improvement in cash, not extra revenue, and it should not be added to the other two.

Leak 4: retyping inquiry details

If someone retypes the details of 15 inquiries a week at 5 minutes each, that is 1.25 hours a week. At $35 an hour the cost is 1.25 × 35 × 52 = $2,275 a year. Small, but nearly free to fix.

Rank what to fix first

Rank each leak by its estimated annual value, multiplied by your confidence in the estimate, divided by the effort to fix it in days. Confidence is a judgment call: 1.0 when you have clean data, 0.6 when it is a reasonable estimate, 0.3 when it is mostly a guess.

The effort figures here are made up too. The ranking tells this business to start with quote follow-up, and the invoice timing fix is worth doing alongside because it is cheap and helps cash flow. Treat the scores as a way to sort a list, not as a forecast.

Choose the right kind of fix

Fixes come in three kinds. A policy fix is a rule or an owner. A tool fix uses a feature in software you already pay for. An automation fix builds a workflow that does the task for you. Work through them in that order. Automating an unclear process gives you an unclear process that runs faster.

Measure again after 30 days

Write down the baseline numbers before you change anything. After 30 days, pull the same metrics from the same sources and compare. Small samples are noisy, so look for direction over a couple of months rather than celebrating one good week. Keep the changes that moved a number and drop the ones that did not.

Common mistakes

• Mapping the process as it is supposed to work instead of as it works. Use dates from your records, not memory.
• Looking only at jobs you won. The lost ones show you the leaks.
• Using averages, which one outlier can distort. Use medians.
• Counting hours saved as revenue. Freed time is only worth money if it gets used for something that earns.
• Fixing all six stages at once. Fix one properly, confirm it worked, then move on.
• Automating before the process is stable. If people disagree about how a stage works, settle that first.
• Ignoring after-hours. For many service businesses a large share of inquiries arrives outside working hours.

A mapping worksheet to copy

Stage:
Owner today:
What triggers this step:
Where the work lives (tool):
Target time:
Actual median time (from my sample):
Share of customers who drop out here:
Biggest leak I can see:
Evidence (dates, counts, examples):
Estimated cost per month:
Fix type (policy, tool, or automation):
How I will measure improvement:
Date to re-check:

Where to go from here

Once you know which leaks are largest, the next question is which fixes suit software, which suit AI, and which should stay with people. The companion guide, what an AI audit is and how to run one, covers that.

Sources

1. Oldroyd, McElheran, and Elkington, "The Short Life of Online Sales Leads," Harvard Business Review, March 2011
2. Rother and Shook, Learning to See: Value Stream Mapping to Add Value and Eliminate Muda, Lean Enterprise Institute, 1998
3. Gurol-Urganci and colleagues, "Mobile phone messaging reminders for attendance at healthcare appointments," Cochrane Database of Systematic Reviews, 2013