Not sure where AI would pay off? Take the 3-minute opportunity check →
Strategy · 5 min read

How to map a business process before you try to fix it

By the InTheory team · Updated November 2026

Most automation projects that disappoint skipped the map. Process mapping is simply writing down how a piece of work actually moves through your business today, who touches it, where it waits and what it costs. Done in a few hours with the right people, it shows where improvement is worth the effort and where it is not.

Key takeaways
  • Map one workflow with a clear start and end, not the whole business.
  • Map how work actually happens, with the people who do it, not how the manual says it should.
  • Waiting time between steps is usually larger than working time.
  • Finish with a baseline and one agreed problem to fix.

Why map before you change anything

When an owner says 'our onboarding is slow' or 'proposals take forever', there is usually a specific reason buried in the steps. A document waits for one partner's approval. A form asks for information the client does not have yet. Two people each think the other sends the welcome email. Buying software without seeing those details often automates the symptom and leaves the cause.

A process map makes the work visible. It gives everyone the same picture, which is often the first time the people involved have seen the whole sequence. It also gives you a baseline, so that any change can be measured rather than judged on feel.

Step 1: Pick one workflow and define its edges

Choose a workflow that is repeated often, costs noticeable time or money, and has a clear trigger and outcome. Good candidates include new enquiry to booked consultation, signed agreement to first delivered work, job completed to invoice paid, or tenant maintenance request to work order closed.

Write the start and end in one line each. 'Starts when a web enquiry arrives. Ends when a consultation is booked or the lead is closed.' Clear edges stop the session from wandering into every related problem the business has.

If you are unsure which workflow to start with, ask two questions. Where do clients complain about delays? And where do staff say they spend time on work that feels pointless? Workflows that show up in both answers are usually the best first choice, because improving them helps revenue and morale at the same time.

Step 2: Walk it with the people who do it

Book 60 to 90 minutes with the people who actually do the work, not just their manager. For a law firm intake process, that might be the receptionist, an intake paralegal and the attorney who takes consultations. For an HVAC service call, the customer service representative, the dispatcher and a technician.

Ask them to walk through a real recent example, step by step. Use sticky notes on a wall or a simple whiteboard tool such as Miro, Lucidchart, FigJam or a shared slide. For each step, capture who does it, what system they use and what information they need. Ask 'what happens next?' and 'what happens if that goes wrong?' repeatedly. The exceptions are often where the time goes.

Keep the session practical. Ask the manager to listen more than talk, since staff are often more candid about workarounds when the boss is not narrating the official version. If people are nervous, make clear the goal is to fix the process, not to judge anyone's performance. A relaxed session produces a far more accurate map.

  • Step name (a verb: review, send, call, enter)
  • Who does it
  • System or tool used
  • Information needed and where it comes from
  • Typical working time and typical waiting time before the next step

Step 3: Mark the waits, handoffs and rework

Once the steps are up, look for three things. Waits are gaps where work sits idle: in an inbox, awaiting approval, waiting for a client. Handoffs are points where work passes from one person or system to another, which is where things get dropped or retyped. Rework is anything done twice, such as correcting a proposal after review or re-entering data from an email into the CRM.

Mark each with a different color. In most service businesses, the waiting time between steps is far larger than the working time on them. A proposal might take two hours of actual work and six days of elapsed time. That gap is usually the most valuable thing to fix, and it is often fixed by changing a rule or an owner, not by buying software.

Also note decision points: places where someone has to judge what happens next, such as whether a lead is a good fit or whether a quote needs partner approval. These are where rules can often be written down, which speeds the work and makes it easier to train new staff.

Step 4: Put numbers on it

A map without numbers is a picture. Add a few measures, using real data where you can and team estimates where you cannot. How many times a month does this workflow run? What is the elapsed time from start to end? How many staff hours does each instance take? How often does it go wrong or need correcting?

Pull a sample of 20 to 50 recent instances from your CRM, practice management or field service system to check estimates. People often underestimate waiting time and overestimate how often things go smoothly. These figures become the baseline you will measure any improvement against.

Turn the numbers into money where you can. For example (illustrative), if a workflow runs 40 times a month and each instance takes three hours of staff time at a loaded cost of $50 an hour, that is $6,000 a month in labor before counting any lost sales from delays. Rough figures like this help decide how much a fix is worth.

Step 5: Choose what to fix, and what not to

With the map and numbers in front of you, list possible improvements and rank them by impact and effort. Some will be simple: remove a duplicate approval, assign an owner to a shared inbox, turn on a reminder in an existing tool. Some will involve integration or automation. A few might justify AI, such as drafting from structured notes or extracting data from documents.

Pick one to start. Define what success looks like in numbers, who owns it and when you will check. This is how InTheory approaches the first 30 days of an engagement: map, baseline, agree one pilot and its acceptance criteria. Sometimes the map shows the workflow is fine and the real problem is elsewhere, which is a useful result too.

Common mistakes

Mapping the ideal process instead of the real one is the most common error. If the documented procedure says the office manager reviews every quote but in practice nobody does, map what actually happens. Mapping too much is the second: a 60-step diagram of the whole company is impressive and useless. Keep to one workflow and a level of detail that fits on one wall.

Finally, do not leave the people who do the work out of the decision. They know which fixes will stick and which will be worked around within a week.

Common questions

What software do I need to map a process?

None, to start. Sticky notes on a wall or a whiteboard work well. Miro, Lucidchart, FigJam or Visio help if the team is remote or you want to keep the map. The conversation matters more than the tool.

How long does process mapping take?

For one workflow, one 60 to 90 minute session with the people involved, plus a few hours to pull data and tidy the map. Larger workflows may need two sessions.

Should we map processes before buying new software?

Yes. A map shows what the software actually needs to do, which steps it would replace and how you will measure success. It also often reveals that a configuration change in your current tools would solve the problem.

Next step

Bring one expensive workflow. Leave with a clear next step.

In 30 minutes we look at one recurring workflow, what it costs you today and whether there is a credible case to improve it. If there is not, we will say so.

Book a 30-minute workflow review Not ready to talk? Take the 3-minute check

30 minutes with Phil or Martin. No technical brief, no pitch deck. Phil or Martin replies personally, usually within one business day.

Book a 30-minute review