The Desk Is the Document: A Week-One Audit of Tacit Service Rules

Most service operating models are described in documents. New starters encounter them through people. In the first week, the formal induction pack competes with desk-side instruction: which fields are genuinely optional, which customer gets a call back, when to bypass the queue, which handover note is trusted. These tacit rules are not automatically a problem. Often they are the model working under pressure. But if they remain unexamined, they become the operating model in practice. A week-one audit captures what is absorbed before it hardens into habit.

Why the first week deserves an audit

Induction is a controlled environment. The first week is live. New starters are unusually observant and least enculturated. They notice contradictions that established staff no longer see, and they ask questions that experienced colleagues stopped asking years ago. They are also vulnerable to adopting shortcuts without understanding the control they replace.

The audit is not a performance review of new staff or their buddies. Its purpose is to compare the designed model with the received model. Timing matters. Conduct it in the first five working days, before local practice becomes the newcomer's normal.

Build a lightweight evidence set

A useful week-one audit does not require a large sample. Two to five new starters across one or two teams is enough to reveal patterns. Explain the purpose in advance, anonymise the findings and keep customer identifiers out of notes. The evidence can be gathered through five streams.

• Structured observation of shadowing sessions. Note where the trainer departs from the induction pack, and what is demonstrated rather than described.
• A question log. Record every question the new starter asks and who answers it. Repeated questions point to documentation gaps or unclear decision rights.
• A help-seeking map. Note who the newcomer turns to when stuck. This exposes informal authorities and single points of knowledge.
• Handover and case notes. With permission, compare a small number of real handovers with the standard template. Look for extra notes, missing fields and verbal workarounds.
• An end-of-week recall interview. Ask the new starter to explain the process in their own words, including exceptions. Their explanation is the received model.

What to listen for

Absorbed rules often appear as conditional statements. If the system times out, use the back door. If the customer mentions the ombudsman, transfer to a named person. If it is after four o'clock, email rather than log. Capture these as statements, not judgments. Then classify them: routing, authorisation, evidence, customer communication, handover, prioritisation or tool use.

Also note what new starters are told not to do. Silence is data. If no one explains why a step exists, it may be ceremonial. If no one demonstrates how a control is evidenced, the newcomer may learn the appearance of compliance rather than the control itself.

Run the week without disrupting service

The audit should sit inside normal service activity, not alongside it. On day one, observe induction and the first shadowing session, noting where the trainer departs from the pack. On day two, collect the question log and a simple tally of help-seeking. On day three, review five real handovers with permission and compare them with the standard. On day four, hold a twenty-minute recall interview and ask the newcomer to walk through what they would do in two or three common scenarios. On day five, synthesise the findings with the team lead.

The tone matters. The aim is not to catch people out. It is to identify where the operating model is unclear, impractical or out of date. Where possible, use existing team meetings and existing onboarding checklists rather than creating a separate audit bureaucracy.

Compare the received model with the designed model

Create a simple two-column map: designed rule and absorbed rule. Add columns for evidence, risk or benefit, and owner. Some absorbed rules will be legitimate local adaptation. Some will be service debt. Some will be control gaps. Do not assume the official version is always right. If experienced staff consistently teach a shortcut, the designed process may be impractical. If new starters learn a risky workaround in week one, the control is not landing.

This comparison is useful evidence for service improvement because it shows where the operating model is being translated, not simply followed. It also gives team leaders a neutral way to discuss local practice without turning the conversation into a compliance exercise.

Turn findings into operating model changes

Three outputs should follow. First, revise the onboarding checklist. Add the reason for each control, the common exceptions, who to ask and what not to improvise. Second, clarify decision rights. If new starters are told to ask a particular person, make that role visible or remove the dependency. Third, review handover and note standards. If templates are ignored, simplify or redesign them. If extra notes are used, incorporate them into the standard.

Assign owners and review dates. If changes cross teams, take the findings to the relevant service governance forum. The audit should leave behind a better model, not just a report.

Make it repeatable

A week-one audit can become a quarterly service improvement routine. Keep a tacit rules register and track whether new starters ask the same questions. Add a received-model item to operating model reviews. Use the findings in training and process design, not in blame. The objective is not to eliminate local knowledge. It is to make that knowledge explicit enough to govern.

The first week is when the operating model is translated. Auditing what the desk teaches first is one of the most practical ways to find out what the model actually is.