Skip to content
Is there an AI for this?

FDE ACADEMY · MODULE 05 · LESSON 02

Working inside a customer team

The first week, the cadence that keeps trust, how to say no without ending the engagement, and how to avoid becoming the shadow IT department nobody approved.

Effort
One day-one session run properly, plus the write-up.
Skills
change-management · compliance-governance

Free and self-paced. No certification is offered.

Before you startYou have a stakeholder map from lesson 01 and a booked session.


01Be in the room, and be useful in the first week

The advantage of the role is proximity: you see the process, you hear the aside about the spreadsheet, you notice the workaround nobody documents. That only happens if you are physically or persistently present rather than joining a weekly call.

Deliver something small and real in week one — a working search over one folder, a script that does the boring export, a diagram of the current state that makes people say “yes, that is it”. It buys the access you need for everything after.


02Say no, and write down why

You will be asked for things that are outside scope, unwise, or both. Refusing well is a specific skill: name what you will not do, name the reason, and offer the smaller thing you will do instead. Then record it in the decision log, so the same request in six weeks meets an answer rather than an argument.

A decision log is three columns — date, decision, reason — and it is the single highest-value document in the engagement. It is also most of the handover written incrementally.


03Do not become shadow IT

The failure mode of a well-liked forward-deployed engineer is becoming the organisation’s unofficial IT function: fixing printers, holding credentials, being the only person who can restart the thing. It feels like trust and it is a liability for both sides.

Guard against it structurally. Every credential goes into the customer’s own store. Every service has a named owner from day one. Every routine is run by them with you watching before you leave. Module 03 lesson 05 built the routine; this is where you insist somebody else runs it.

  • Credentials in the customer’s vault, never only in your password manager.
  • A named owner per service, recorded in the register from module 04.
  • A decision log kept from week one, in a location the customer controls.

04Do this

PRACTICAL TASK

Run a day-one observation and produce its written record

Run one full observation day in the Sprint’s day-one shape and produce the written record it names, then start the decision log.

What you need

  • A confirmed session with the people who do the work
  • A place to keep the record that the customer controls

Steps

  1. 01

    Read the day-one list and use it as your agenda: process, systems, workflows, data, pain points, security, governance.

    Day 1 — Observe

  2. 02

    Spend the morning watching and the afternoon confirming. Read your own notes back to the people who gave them to you and correct what you got wrong.

  3. 03

    Write the record the same day. A written record of the processes, systems and data you looked at is the first thing the Sprint promises to deliver.

    What you get

  4. 04

    Start the decision log with the first three decisions you have already made, including anything you have declined.

  5. 05

    Store both in the customer’s own system, not yours, and tell the process owner where they are.

You are done when

  • A written observation record the participants have read and corrected.
  • A decision log with at least three entries, one of which is something you refused.
  • Both documents live somewhere the customer controls and can find without you.

05Where these facts live

This lesson does not restate anything that is already recorded with its evidence elsewhere on the site. These are the pages it leans on.

Ticks are stored in your browser only. They are not sent anywhere, they are not attached to an account, and clearing your browser data removes them.