The AI-Powered WordPress Development Side Hustle

What is inside, section by section — all 45 of them, in order: 1. A physiotherapy practice's booking site white-screening at 9:40 on a Tuesday after an overnight plugin auto-update, with no reachable backup; and a charity site with a directly-edited parent theme, an unremovable builder, and a load-bearing donation plugin abandoned in 2019 — neither of which appeared in any brief. 2. The job is keeping something working on ground that moves, which makes procedure rather than skill the variable that decides whether it is a business. 3. Six trigger events, and the strongest is a developer who disappeared — which says more about what they will buy than the brief does. 4. Six services sorted by predictability and emergency exposure, with the paid audit sold first and rescue work taken carefully. 5. They picture a finished thing at a price per page, and they think maintenance is optional — which is what the first broken update corrects. 6. Rule one: core, PHP, themes and plugins move on their own schedule, so something that worked in March fails in September untouched. 7. Rule two: they can install, edit, share access or let an auto-update run — so the state of a site is never known, only checked. 8. Rule three: a directly-edited parent theme, an unremovable builder and a load-bearing plugin abandoned in 2019, none of it in the brief. 9. Rule four: nobody has to report a broken site, everybody sees it at once, and the client can usually name what the hour cost. 10. Not a marketer, not a copywriter, not a compliance adviser — and no fixed price on an unaudited site, ever. 11. Five variables decide margin, and whether the client will accept your stack is the one that predicts the next three years. 12. The brief describes the visible layer; the audit finds the work — and it is a paid product rather than free pre-sales labour. 13. Count how many providers state what maintenance includes and what response time they offer, because most state neither. 14. Four lines and four numbers, with emergency tolerance counted separately from hours because it is the real constraint. 15. Almost nothing is required except staging capability, which is a precondition rather than a purchase you defer. 16. Questions rather than answers, and consequential loss is the clause to ask about before you accept responsibility for uptime. 17. The standard stack defined by function rather than by product name, because anything named goes stale and a stale specific is rule one. 18. The paid audit is the structural fix for quoting blind, and the non-standard modifier prices what a deviation actually costs. 19. Dated screenshots at launch, permission in writing, and no claim about rankings, traffic or revenue the client cannot substantiate. 20. Nine steps ending with the maintenance conversation, and the expectation that content arrives late because it always does. 21. The decision that determines everything afterwards, recorded at the time — because otherwise it gets re-decided every eight months. 22. Ask what happens to the site if it is removed, before installing rather than during the incident, and register every one. 23. An untested backup is a belief, "the site loads" is not a successful restore, and the test date is recorded rather than ticked. 24. Measure before changing anything, and never say a site is secure — because nothing here is a guarantee and none of it is said as one. 25. The highest-risk routine task, with a rollback plan written before you start and never two migrations in one session. 26. Commerce raises every stake, payment and tax configuration goes to somebody qualified, and a rebuild without redirects destroys the site. 27. The most common expensive mistake is shortening the audit because the job looked small — small briefs on old sites are the trap. 28. Exclusions in writing, response times stated honestly including what you do not offer, and no plan too cheap to honour. 29. Projects are a treadmill and plans are a business; the month maintenance covers fixed costs is the month it stops being fragile. 30. Widening while standardisation falls makes the estate worse rather than larger — fix the ratio before adding sites. 31. Five records, and two of the key numbers — restore test dates and baseline states — cannot be reconstructed afterwards. 32. Two non-standard sites consume the margin of six standardised ones, and the fix is a migration rather than a departure. 33. Five ratios including host concentration — the one people miss, because one host's outage reaches every site at once. 34. Update windows compete directly with project blocks, so batch them into one protected session — and never deploy before a holiday. 35. Under pressure you fall to the level of your procedure; if you write only three, write updates, restores and incident triage. 36. Stack standardisation is the one number to compute if you compute only one, and standardised and non-standard sites are never averaged. 37. Two gates, a deliberate break-and-restore, a real migration, and four stop conditions of which two decline silently. 38. A model asked for code produces code that runs — and code that runs and is wrong is worse than code that does not run. 39. Twelve documents for deciding whether to do this: the self-assessment with its two gates and its emergency-tolerance question, the model comparison grid, the four-rules reference, the scope definition worksheet, the stack standard planner, the refusal list, the screening card, the one-page plan and the practice plan built on reps rather than courses. 40. Fourteen documents for setting up: the stack inventory with whose account each sits in, the capacity planner counting emergency exposure, the agreement checklist led by consequential loss, the access and offboarding record, the standard stack specification, the onboarding checklist with its first restore test, the rate card, the development standards, and the maintenance plan specification with its exclusions. 41. Twelve working documents: the site audit question set, the pre-launch checklist including the visibility setting everybody forgets, the backup and restore test log with its eleven-point verification list, the update procedure, the plugin decision matrix, the site log recording environment discovery separately, the feedback question set and the incident sheet. 42. Twelve documents for the arithmetic: the working-hour cost, cost per site-month by profile, revenue per site split by standardisation, the break-even site count, the rate change notice, the no-guarantee checklist, the emergency policy, the profitability sheet with host concentration, the SOP template and the eleven-item quality control checklist. 43. Twelve documents for finding and keeping work: ten named targets with triggers and introducers, outreach scripts leading with the paid audit, the second developer agreement, the post-launch survey, confidentiality and upload rules, the six quarterly client questions, correction templates, and the incident record whose quarterly reading distinguishes a procedure problem from an architecture problem. 44. The KPI scorecard led by standardisation and restore-test coverage, the second developer onboarding checklist with its question-counting readiness test, the thirty-day launch plan including the break-and-restore, the ninety-day growth plan, the one-year review worksheet, and the complete library of all 148 prompts. 45. The four rules restated with the interaction that is the whole risk, the five things to do first, the AI warning that code which runs and is wrong is worse than code that does not run, and the people who never hired you — relying without knowing it on whether you tested a restore in March. Every worksheet, register and checklist is included, and every financial figure in the book is left blank on purpose — you fill them in from your own measured costs rather than someone else's guesses. Instant download. Yours to keep and print as often as you like.