A 90-day plan is the most useful document a small business can own, and the easiest one to ruin. The usual way to ruin it is to make it complete. Every idea from the planning meeting goes in, each gets a coloured cell in a spreadsheet, and by week three nobody opens the file.
My view: a good 90-day plan fits on two pages, and most of its value comes from what it refuses to include. Here's what earns a place on those two pages.
The seven parts
1. Where you are, in a paragraph
Three or four sentences built from numbers you already have. Monthly enquiries, close rate, average job or order value, what marketing costs you now.
If you don't know a figure, write "unknown" and make finding it a task. That's allowed. Guessing isn't.
2. One goal for the quarter
One. Something the business actually feels (booked jobs, qualified demos, repeat orders) with a number and a date attached.
"Grow brand awareness" isn't a 90-day goal. "Go from about 30 to 45 booked estimates a month by the end of December" is.
Two goals means two plans. Pick one.
3. Three priorities, ranked
These are the moves you believe will get you to the goal. Rank them, because when week six gets busy something will slip, and everyone should already know which thing.
Why three? A small team can carry about that much alongside the day job. Some quarters it's two. Five almost never survives a busy month.
4. Actions, each with an owner and a date
People skip this part, and it's the part that makes the plan real. Every action gets:
- one owner, meaning a person, not "marketing" or "the team"
- a due date
- a "done when" line that someone else could check
"Improve the website" fails all three. "Add a booking form to the furnace-repair page. Owner: Sam. Due October 23. Done when a test submission lands in the shared inbox" passes.
5. Measures, and where each one comes from
Two to four numbers you'll check every two weeks. Mix one or two that move fast (enquiries, calls answered) with the slower one your goal is built on (booked jobs, revenue).
Name the source next to each number: "GA4, form submissions" or "the booking sheet, source column." A number without a source turns into an argument at the first review.
6. What you're deliberately not doing
My favourite section. List the reasonable ideas you're parking this quarter: the TikTok account, the rebrand, the second location's launch page.
Writing them down stops them sneaking back in at every meeting. It also gives your team permission to say no.
7. The review rhythm
When you'll look at it, with dates in the calendar. I suggest a 30-minute check every two weeks, plus a longer review at day 45 where you're allowed to swap out a priority that clearly isn't working. At day 90 you write the next plan, starting from what this one taught you.
A skeleton you can copy
| Section | What goes in | Length |
|---|---|---|
| Starting point | Current numbers, with unknowns flagged | 1 paragraph |
| Goal | One outcome, one number, one date | 1 line |
| Priorities | Three, ranked | 3 lines |
| Actions | Owner, due date, "done when" | 10–20 rows |
| Measures | 2–4 numbers, each with its source | 1 small table |
| Not this quarter | Parked ideas | A short list |
| Reviews | Fortnightly check dates, plus day 45 and day 90 | 1 line |
Want to see one filled in? I've published an illustrative 90-day plan for a fictional HVAC company, week by week, with owners and dates.
What to leave out
- Channel strategy essays. If an action needs a page of justification, write it separately and link to it.
- Mission statements, brand values, SWOT grids. Useful somewhere else. Clutter here.
- Anything under an hour. That belongs on someone's to-do list, not in the plan.
- Anything with no owner. If nobody will own it, it isn't a priority, whatever the meeting decided.
- Money you haven't approved. A plan built on budget that doesn't exist becomes fiction by week two.
Three tests before you commit
- The Monday test. Could each owner tell you, without opening the file, what they're doing on the plan this week?
- The hours test. Add up rough hours per action and set them against the time your people really have spare. First drafts are commonly about twice too big (that's my rule of thumb, not a statistic). Cut now, not in week eight.
- The swap test. If priority three were cancelled tomorrow, would anyone object? If not, it probably doesn't belong.
Check the calendar first
Plans often break on dates nobody looked at. Statutory holidays differ by province. Your busy season eats everyone's spare hours. December is short.
An Ontario HVAC company planning October to December has Thanksgiving, the first cold snap and the holidays inside one quarter. A Calgary retailer planning July has Stampede. Put the calendar in first, then fit the actions around it.
Running it
A working plan should look used. Strike through what's done, note what slipped and why. By day 90 the marked-up copy is worth more than the clean one, because it's the only honest record of how much your team can get through in a quarter. That's the number the next plan should be built on.
If you'd like help with the first one, a 90-day plan comes with the marketing audit. It's also the month-one deliverable on the Working Partner retainer, where I update it with you every month.