Module 8: Start Here

Overview

The previous modules built an architecture. This one is about whether any of it ships.

What kind of organisation are you actually in — and why would anyone follow a decision you have no authority to enforce?

Those are one question, not two. The same architecture succeeds in one company and fails in another, and the difference is rarely technical. You cannot direct a single engineer, so the informal system is not a soft consideration — it is the mechanism by which anything happens at all.

The module runs in that order: read the organisation, take a position in it, earn the standing to be listened to, then make the decision itself defensible.

This is not soft skills

A warning, because the material invites it.

Everything here has an architectural consequence, and if a topic reads as advice about being nice to colleagues, it has been written badly:

• Culture determines which practices work. An automatic regression gate is worth more where bad news is punished than where people raise concerns freely — because in the first, nobody will carry the bad news and the mechanism has to.
• Structure determines which designs are cheap. A design needing one team to own something end to end costs little in one structure and a quarter of negotiation in another.
• Size determines what the job is. An architect hired at fifty engineers makes choices; one hired at a hundred and forty stops the company colliding with itself.

Read every topic for the decision it changes.

What the second half rests on

Three claims, each with a concrete practice behind it:

• Trust is earned by small checkable accuracies and spent by one confident overclaim. The ratio is unfavourable, which decides how you behave in month one.
• Stating the limits of your responsibility increases trust rather than reducing it. Unbounded claims are not credible.
• Criteria fixed before options are scored is the difference between a justification and a rationalisation — and the fixing has to be visible in the artifact.

None of it works if the underlying decision is unsound. A convincing argument for a bad design is a liability, not a skill.

How the module is built

graph TD
  A["<b>1 · Reading the<br/>organisation</b><br/>structure, Conway, culture,<br/>size, power"] --> B["<b>2 · Positioning<br/>yourself</b><br/>remit, antipatterns,<br/>scar tissue"]
  B --> C["<b>3 · Trust and<br/>translation</b><br/>credibility, boundaries,<br/>audiences, the sceptic"]
  C --> D["<b>4 · Justifying<br/>and selling</b><br/>criteria, rejections,<br/>the pitch"]
  D --> E(["Workbook — Exercise 8"])

  classDef read fill:#5B8FB9,stroke:#4A7A9E,stroke-width:2px,color:#ffffff
  classDef pos fill:#E8A33D,stroke:#C4832A,stroke-width:2px,color:#ffffff
  classDef cap fill:#D9695B,stroke:#B4483A,stroke-width:2px,color:#ffffff
  class A,B read
  class C,D pos
  class E cap

Four sections. The first two establish what is possible here; the last two convert that into a decision people act on.

Section 1 — Reading the organisation

1. Structure and what it permitsWhat the shape of the organisation makes cheap and expensive.

2. Conway's Law as a toolUsing the law to predict which boundaries will erode.

3. Reading culture from meetingsWestrum's typology, and the observable evidence for your reading.

4. What changes with sizeThe thresholds at which coordination breaks, and what the job becomes.

5. Power and who actually decidesFormal authority, informal veto, and the minimum winning coalition.

Section 2 — Positioning yourself

6. Ivory tower or embeddedTwo ways to work, and what each costs in attention.

7. Building a remit with no mandateScope accumulates from precedent unless you choose it deliberately.

8. Organisational antipatternsThe recurring failures, and the mechanism that answers each.

9. Scar tissueHow a past failure constrains a present proposal.

10. Expectations, and when to leaveThe cursed role, and how to tell one from a hard job.

Section 3 — Trust and translation

11. How trust is earned and spentSmall checkable accuracies, and the one overclaim that undoes them.

12. Stating your boundariesWhy naming what you do not decide raises your credibility.

13. Translating across audiencesOne decision, four documents that differ in substance rather than tone.

14. Earning trust from the scepticThe person who has seen this fail before, and what actually answers them.

Section 4 — Justifying and selling

15. The comparison tableCriteria fixed before options are scored, and visibly so.

16. Rejected options and uncertaintyWhat you did not choose, and what you are not sure about.

17. The architecture pitchTwenty minutes with people who will not read the pack.

18. When the decision goes against youDisagree and commit, escalate, or record — and how to choose.

19. Authority beyond the rolePublished reasoning as the only authority available to you.

How to work this module

The longest in the guide — about ten hours, and the only one with four sections.

1. Read the nineteen topics — about 4.5 hours.

2. Optional — Workbook exercise 8 — about 5.5 hours, seven artifacts.

In the workbook

Seven artifacts for Lintel:

• An organisational analysis — structure, culture type, and the observable evidence for your reading
• A stakeholder influence map revising Module 1’s naive version, separating formal authority from informal power
• An antipattern assessment centred on the prior failure and how it shapes the reception of your proposal
• A positioning statement — how you intend to build a remit when you were given none
• A decision justification pack for the Module 4 stack decision, with explicit criteria, scored options and rejected alternatives
• Three one-page framings of that decision for the budget holder, the engineering owner and the compliance owner
• A ten-slide architecture pitch that opens with the business problem and answers the prior-failure objection before anyone raises it

Before you start

Find your Module 1 stakeholder map. You wrote it four weeks into the job, labelled it v0, and were told a later module would correct you. This is that. Do not look at it again until you have read Section 1 — then compare.

You need Module 4’s stack decision and the Module 6 and 7 findings. The second half of this module delivers a real decision to the people the first half identifies, and the pitch’s credibility rests on specifics — the cost target that does not close, the regression trigger that sat inside the noise floor. Generic reassurance will not survive the room.

Check your understanding

Answer now, revisit at the end.

1. In an organisation you know, who could stop a cross-team technical decision without having the authority to?
2. What happens there when someone brings bad news to a senior person? Be specific about an occasion.
3. Think of someone whose technical judgement you trust. What specifically did they do to earn that?
4. You have presented a decision and the room has gone against you. What do you do in the next five minutes?

→ Start with Structure and what it permits.