Exercises: Technology and Systems

Part A recall · Part B application · Part C field work · Part D critique · Part M mixed review · Part E extension.

Starred (★) problems have worked answers in appendices/answers-to-selected.md.


Part A — Recall and Comprehension

A.1 ★ State the chapter's threshold concept. Give the five reasons a written finding produces no change.

A.2 Which two reasons account for most of it, and which is the one nobody admits?

A.3 ★ Name the four components of a system. Which of them is software?

A.4 Give the test of whether you have a system or expertise. What follows if you fail it?

A.5 ★ Define a trigger. Give the four kinds and their reliability.

A.6 What is a threshold trigger, and why is it the one almost nobody has?

A.7 ★ Give six of the ten triggers in the standard set, with what each fires.

A.8 Name the nine stack functions. Which four are non-negotiable, and which is free and universally absent?

A.9 ★ Give the six platform-evaluation questions. Which one matters most and why?

A.10 What is the parallel event, and what does it answer that a demo cannot?

A.11 ★ Give the four document conventions and the folder tree's seven numbered directories.

A.12 State the unit convention. Give three examples and name the four errors it fixes.

A.13 ★ What is a correspondence log, and what three things is it for?

A.14 How does a decision record differ from a log entry? When do you write one?

A.15 ★ What goes in a venue file? Which Chapter 34 finding does it close?

A.16 Name four categories of sensitive data a planner holds, and the five things true almost everywhere.

A.17 ★ Give the five fields of the closing finding log. Which is load-bearing and why?

A.18 Why is a log of unclosed findings worse than no log?

A.19 ★ What should a system absorb, and what can it not? Give the ranked list of what to systematise.

A.20 Name the six places technology makes things worse. Which is the sharpest?

A.21 ★ Give the automation test. What may a tool draft, and what may it decide?

A.22 State the capacity problem. How many changes will a working planner implement in a season?


Part B — Application

B.1 ★ Take your own or a constructed finding log and convert it to the five-field form. Name the artefact for every entry — and note which entries cannot be assigned one.

B.2 Build the trigger set for your practice. Ten triggers, each firing a named artefact, applied from a template at signing.

B.3 ★ Write three threshold triggers, with the threshold and what fires. At least one must involve the guest count.

B.4 Map your stack by function. Nine functions, current product, monthly cost, and whether function 8 exists.

B.5 ★ Evaluate a real planning platform against the six questions. Answer question 5 honestly — what would you stop doing?

B.6 Define your folder tree and naming convention. Then apply it to one real event's files and report how long it took.

B.7 ★ Take five numbers from your own practice and write the unit next to each. At least one should surprise you.

B.8 Start a venue file for a venue you know. Every section, from memory, badly — and note what you cannot fill in.

B.9 ★ Write your data position: what you collect, where it lives, who you share what with, retention, and the paragraph clients see. Under 300 words.

B.10 Write ten correspondence-log entries for a real or reconstructed month.

B.11 ★ Write three decision records for decisions somebody will ask about later.

B.12 List everything you do that a competent substitute could not do from your documents. Then pick the two cheapest to fix.

B.13 ★ Run §39.11's ranking on ten findings. Choose four, defer six with dates, and justify both lists.

B.14 Identify three tasks you could automate and one you must not. Apply the draft/decide test to each.


Part C — Field and Practice

C.1 ★ Ask a working planner to show you where their run sheet template lives. Time how long it takes them to find it.

C.2 Price a full stack in your market, by function. Report the monthly total and what it is as a share of a first-year fixed-cost budget.

C.3 ★ Ask three planners whether they keep a finding log, and if so how many entries became changes. Record both numbers.

C.4 Find out what data-protection regime applies to you and what it says about health data. Report what you found and whether small businesses are exempt.

C.5 ★ Try exporting your data from whatever you currently use. Report the format and whether it is usable.

C.6 Ask a venue coordinator what goes wrong at their venue. Start a venue file with the answer.

C.7 ★ Find a planner who changed platforms. Ask what the migration actually cost and what they stopped doing.


Part D — Critique and Judgment

D.1 §39.1 says a beautifully written finding feels more finished. Is that a real mechanism or a rationalisation? How would you test it?

D.2 ★ The chapter says a system is not software and then spends a section on stacks. Is that coherent?

D.3 ★ §39.11 says a planner will implement three or four changes a season. Argue that this is defeatist and that a determined person could do twelve. Then decide.

D.4 The venue file requires writing down things a venue told you in confidence. Where is the line between institutional knowledge and indiscretion?

D.5 ★ §39.9a says a tool may draft anything and decide nothing. Attack that as too conservative, and name a decision you would let a tool make.

D.6 The chapter is critical of client portals. Build the case for one, including client types for whom it works.

D.7 ★ §39.7 says the fact that a couple gave you data does not remove your obligations. Is a solo planner realistically able to comply with a modern data regime? What is the minimum defensible position?

D.8 Chapter 36 said buy the platform fourth. This chapter's parallel-event test is expensive. Is there a cheaper honest evaluation, or is the cost the point?


Part M — Mixed Review

M.1 ★ (Ch.26) What does this chapter add to the production book? Name the direct closure.

M.2 (Ch.30) Which Chapter 30 rule does §39.8 operationalise, and which field is load-bearing?

M.3 ★ (Ch.36) Why buy a platform fourth? What question does §39.4 add that Chapter 36 did not have?

M.4 (Ch.34) Which finding does the venue file close, and what was the corridor conversation?

M.5 ★ (Ch.28, 32, 34, 30) Name the four errors the unit convention fixes, and state the common form.

M.6 (Ch.5) Where does the master variable get managed in this chapter, and what fires?

M.7 ★ (Ch.27) What does "replaceable on a task, irreplaceable on judgment" have to do with the fifteen-second decision?

M.8 (Ch.29) Which artefact carries the access questions, and what trigger fires it?

M.9 ★ (Ch.9) Which trigger would have caught the expired insurance certificate, and when?

M.10 (Ch.38) Which column does this chapter make possible, and what does it unlock?


Part E — Extension and Synthesis

E.1 ★ Run the full systems audit: every change this book has produced, ranked, four chosen, the rest deferred with dates, plus the stack, the architecture, and the data position. Then diary the deferred list.

E.2 Research how another profession handles institutional knowledge — aviation's occurrence reporting, medicine's morbidity and mortality meetings, software's post-mortems. Report one mechanism this chapter should adopt.

E.3 ★ Write 800 words on whether a solo practitioner can realistically run a system at all, or whether "systems" is advice written for organisations and applied to individuals who cannot use it. Reach a position.

E.4 Take the twenty-two changes from Case Study 39.2 and estimate the implementation time for each. Report the total and what it implies.

E.5 ★ Design the ninety-minute annual systems audit. A template somebody would actually complete, producing four changes and a dated deferred list.

E.6 Interview somebody who has run this business for over a decade about what they wrote down and what they did not. Ask specifically what they wish they had recorded.


📐 Mini-Project

Build the system, minimally, and prove it works.

  1. The finding log, five fields, with your last season's findings entered
  2. The artefact for every entry — and a list of the ones that cannot have one
  3. The trigger set, ten triggers, applied to one live event
  4. The folder tree and naming convention, applied to one event's files
  5. One venue file, started
  6. The unit convention, applied to five numbers
  7. The stack by function, with function 8 named and located
  8. The data position, written
  9. The ranking — four changes for this season, the rest deferred with dates

Then the test that matters, and it is uncomfortable:

Hand your documents to somebody competent who was not involved, and ask them to tell you what happens at 4:15 on your next event.

If they can, you have a system.

If they can only find out by asking you, you have expertise — and a broken ankle is a business failure.