Where does an ordinary request wait before the customer receives what was promised? That is a more useful starting question than “How can we make everyone work faster?” A small service business can lose time in handoffs, repeated data entry, unclear approvals and corrections even while every person is busy. This page helps you find one such interruption and measure a small change. It leads to three detailed guides: mapping a routine from request to completion, removing needless duplicate work, and running a short weekly improvement meeting.

Begin with one completed case

Choose an example you can reconstruct without guessing. A home maintenance firm might trace a customer request from first contact through booking, technician visit, completion check and customer confirmation. The start and finish are explicit. Ask the people who actually handled that case to describe what happened, including calls made to clarify details or time spent waiting for a decision. Write each action in order, then mark every transfer between people and every time work had to be corrected. You do not need specialist terms or a process diagram to learn something from a single case.

Write down the elapsed time as well as the time spent actively working. A request may be in someone's inbox for two days even though the actual response takes two minutes. Mark why it waited: no owner, missing customer detail, unavailable part or an intentional approval. The reason matters. An essential safety inspection might look like a delay on the timeline but must remain. An unowned handoff might be solved with a simple status and acknowledgement. ASQ describes process maps as a way to see actions, decisions and people in a sequence and to locate improvement opportunities.

Do not choose a change solely because a stopwatch says one step is slow. Look at the customer's result and the total burden on the business. Removing a five-minute check that prevents expensive rework can cost more time overall. Requiring two staff to enter the same address adds work but may be a symptom of poor records rather than the root problem. A clear map helps the team distinguish a necessary control from a redundant copy.

Choose the right guide

Map a Workflow in 30 Minutes shows how to define the start and finish, draw a real route and mark handoffs, waits and corrections. Read it when nobody can explain where a request slows down or who owns the next step. It includes a service booking example and measures elapsed time and the number of handoffs or corrections in a small set of cases. Start here if the problem feels vague or different staff describe the route differently.

Find and Remove Double Work focuses on information entered, checked or requested more than once. Read it when a customer repeats details, staff keep parallel lists or a transaction is retyped after each handoff. It distinguishes needless copying from independent quality or security checks that must be retained. The proposed first trial is one shared, verified source for a specific piece of information, with duplicate minutes and correction rate recorded before and after.

Run a 15-Minute Weekly Improvement Meeting helps a team turn recurring observations into one small experiment with an owner and a measure. Read it when people identify problems but nothing changes between busy weeks, or when many ideas compete for attention. The meeting reviews one prior experiment, selects one issue and assigns one next action. It is not a substitute for urgent customer complaints or a complex safety discussion.

Test a modest change

Assume a small home maintenance team has ten similar requests in two weeks. Four require an extra call because access instructions were missing; the team also records the time from first contact to confirmed booking. These figures are illustrative, not a general benchmark. A dispatcher tries asking for access instructions when the request first comes in and recording them in the agreed booking field. For the next ten comparable requests, the team counts follow-up calls and measures the same elapsed interval. It also asks technicians whether the instructions were accurate. If calls fall but inaccurate directions cause failed visits, the change needs revision.

The test is deliberately narrow. It changes one point in the process, measures one main outcome and watches one potential harm. A different firm might choose a different bottleneck, such as an approval that sits unassigned overnight. The Institute for Healthcare Improvement recommends testing a change in its real setting, observing the result and adapting it before broader use. The principle is useful beyond healthcare: a small firm can learn from five or ten cases instead of redesigning all operations on an assumption.

Define the measure before starting. Count corrections using the same rule in the baseline and the test. Record unusual events such as an emergency callout or a staff absence so they do not distort a simple comparison. If the sample includes very different job types, compare similar cases or inspect the range instead of treating the average as precise. A gain in one measure should be checked against customer wait, quality and staff workload. The team may need another cycle to learn whether the result lasts.

Keep responsibility visible

At the end of a case, who tells the next person that it is ready? Who notices if an approval is missing? Who has authority to correct the shared record? A process with clearly assigned ownership can be more reliable without an expensive tool. Make the status available to the people who need it, but give access only as appropriate for the data. If staff maintain separate unofficial notes because the shared process is unreliable, ask what information they lack before banning the workaround.

Sometimes the bottleneck is outside the firm's control: a customer must approve extra work, a supplier must deliver a part or a local rule requires a documented check. Record the wait and plan around it instead of pretending it can be deleted. Seek specialist advice before changing regulated, safety-critical or confidential steps. Employment, security and privacy obligations differ by location. Do not interpret every delay as waste.

The first improvement may be as small as one handoff note, one verified field or one weekly check. Next step: take a completed customer request and list the actions from first contact to completion. Count its waits, handoffs and corrections, then choose the most frequent avoidable one to observe over the next five cases.

Sources