The 90-day roadmap: budget, owners, sequence

A strategy without dates, euros and names is a mood board

You left the last lesson with 4–6 kept initiatives. This lesson answers the only question a decision-maker actually asks about them: so what happens Monday?

Ninety days is the deliberate horizon. Long enough for slow channels to show a pulse. Short enough that you cannot hide. And every item on the roadmap needs three things the strategy document doesn't force on you: a place in time, a euro figure, and a name.

Six fortnights, three phases, nothing "ongoing"

Divide the 90 days into six two-week blocks. Every initiative gets build, launch and iterate phases placed in specific fortnights. The word "ongoing" is banned: it is how work escapes accountability. Even a genuinely continuous habit — a weekly post, a monthly send — gets written as launch in fortnight 1, iterate in fortnights 2–6, so there is a date on which it demonstrably started.

Two sequencing rules do most of the work:

1. Dependencies first. Tracking before spend. Lead magnet before nurture. A defined metric before the campaign that will be judged by it. Money spent before its measurement exists is money spent on not knowing.
2. A quick win in fortnights 1–2. Something visible and cheap that lands early. Not for vanity — for permission. Credibility bought in week 3 is what keeps a slow compounder like search content alive at week 10, when someone asks what all this is achieving.

The budget nobody writes down: hours

Here is where most plans die, quietly. Go back to your Module 0 brief and read the number you wrote next to hours per week. That number is a budget, exactly like the euros, and the roadmap must balance against it.

Tomas wrote 3. Three hours a week, usually Sunday evening. Over 13 weeks, that is 39 hours — total, for everything. His kept initiatives, done the way the modules teach them in full, want something like nine hours a week. The gap between those two numbers is not a motivation problem to be solved with resolve. It is an arithmetic problem, and arithmetic problems are solved by subtraction.

The class is shown Tomas's first draft on purpose: seven initiatives running in parallel, ads live from day one before the tracking exists, and an implicit 9 hours a week belonging to a man who does not have them. It fails three ways at once — dependency order, capacity, and the absence of any checkpoint that could kill anything. A plan built for a person you are not is a plan for no one. Whatever doesn't fit your real hours moves to the Not-Doing List with a revisit date. It does not move to evenings that don't exist.

What survives contact with 39 hours:

The search test launches in fortnight 4, not fortnight 1 — not because ads are bad, but because the thing that counts its results (the Tally form and the tagged link) has to be verified first. And note the second name. Tomas is solo, but "solo" rarely means alone: his receptionist already types the email addresses in weekly, so she owns the review-ask cards and the collection habit — about an hour a week that never touches his Sunday three. One owner per initiative. "Everyone" owns nothing. If a row has no honest name, write HIRE or OUTSOURCE in the owner column — that word appearing is itself a finding, and it belongs in your pitch.

Every euro maps to an initiative

Tomas's Module 0 budget: €150 a month. Over the quarter, €450. The whole table:

Four lines. No "misc" — every euro points at an initiative, and the contingency is a named line of 10–15%, not slack hiding in round numbers. The tools column is €0 beyond print because everything else runs on free tiers, which is not an accident; it is what a €150/month constraint selects for. At €1.70 a click — the planning figure from his Module 9 draft — €200 buys roughly 118 clicks, and at his flagged-as-a-guess 5% booking rate, about six Check-up bookings, ≈ €33 each against his ≈ €45 target. If the guess holds. The checkpoint below exists because it might not.

Checkpoints that can kill things

Days 30, 60 and 90 each get a measurable state and a pre-agreed decision rule: continue, adjust, or kill. "Going well" is not a state. Numbers are states.

But here is the trap that ruins honest roadmaps: judging slow channels by outcomes they cannot yet have produced. A search page will not rank by day 30. An email list of 300 will not transform revenue by day 30. If your day-30 checkpoint demands bookings from them, you will kill your best initiatives precisely on schedule. The fix is the leading-indicator discipline from Module 10: each channel is judged, at each checkpoint, by the earliest number that channel can fairly show.

Tomas's three gates, in full:

Day 30. Self-check email sent to the full list (~300 by now at its usual ~30/month growth); ≥25 new addresses collected; cost page live and indexed with impressions above zero; GBP posted all four weeks; reviews 38 → 42. Rule: if the cost page isn't indexed, fortnight 3 becomes a technical fix and the first care page slips — content on pages Google can't see is theatre.

Day 60. Search test: cost per Check-up booking ≤ €45 after at least €130 of the €200 spent — above it, pause and diagnose before another euro. Reel live with its tagged link; ≥100 tagged sessions (a leading indicator — no booking target on it yet). Subscriber → Check-up rate recorded for the first time, whatever it is. Rule: the ads line is the only one with a kill trigger this early, because it is the only one spending cash.

Day 90. Online bookings — Check-up form bookings and enquiries, counted together under the definition below — at ≥15 a month.

Day 90 meets the goal you wrote in Module 0

One piece of housekeeping first, because it decides whether the reckoning is honest. In fortnight 1, Tomas writes down — once, with a date — what "online bookings" means: Tally Check-up bookings plus contact-form enquiries, counted weekly, together. In the fortnight before the roadmap the two routes totalled about 12 a month; his Module 0 baseline of "about 6" counted only the old form. He records that as a definition change, not growth — the counting changed as much as the number, and pretending otherwise would poison every comparison after it. Module 10's rule, applied at the moment it matters: never change a metric's definition without a footnote.

Now the reckoning. Tomas's Module 0 goal was 25 online bookings a month by 1 March 2027. His roadmap's day 90 lands in the same week as that date — so his day-90 checkpoint is not an interim reading. It is the morning the Program has been pointing at since Lesson 0.3, and he pre-commits, now, to what each result means:

≥ 25: the goal is met. Say so, once, and set the next one. 15–24: the goal is missed. Say that plainly too — then show the trajectory, and write a new date. A goal moved in writing, with reasons, is a plan. A goal quietly forgotten is the old life. < 15: the miss is not about pace. Re-diagnose before recommitting a single hour — something in the strategy, not the schedule, is wrong.

Doubling honest bookings in 90 days on €450 and 39 hours is ambitious, and the roadmap says so out loud. That is what the bands are for: he has decided what failure means before it can happen, which is the only time deciding is cheap.

Your goal date may not land this neatly on day 90. State where it falls relative to your roadmap, and give the checkpoint nearest it a number that keeps the goal honest — a trajectory the goal can still be reached from, or a written revision. What you may not do, in this Program or after it, is let the one metric, one number and one date from Module 0 dissolve into "we're doing lots of things".

Now build yours.

---