The 30-60-90 plan is asked for in final rounds across marketing, sales and operations, and most of the ones handed over are variations on the same document: learn in month one, contribute in month two, lead in month three.
It's a reasonable arc. The problem is that every candidate submits it, so it distinguishes nobody — and worse, the version most people write quietly promises results in month one, which is the fastest way to signal that you haven't joined a company recently.
Below is a plan written for a specific role, with the reasoning, and the three questions a panel will use to test whether you'd actually follow it.
Why most of them get ignored
Three failures, in the order they show up.
Month one promises output. "Audit the channel mix, rebuild the reporting dashboard, launch a nurture sequence." Anyone who has started a job knows that month one is access requests, meeting the team, and discovering that the data is worse than advertised. A plan that ignores this reads as someone describing a job rather than remembering one.
Nothing is conditional. A plan written before you have any information should be full of "if". A plan with no branches is a plan you intend to follow regardless of what you find, which is the opposite of the judgement they're hiring for.
Month three is vague. By day 90 the plan usually dissolves into "own the strategy" and "drive growth". That's where the panel was looking hardest, because it's the only part that describes you doing the actual job.
The role this plan is for
Specificity is most of the value, so:
Marketing Manager, B2B SaaS, ~80 people. Reports to the Head of Marketing. Team of two — one content, one part-time paid. The hiring manager mentioned twice that "we don't really know which channels work" and that the pipeline target went up 40% this year.
Both of those sentences are in the plan on purpose. A 30-60-90 that could have been written for any company is the one that gets skimmed.
Days 1–30: find out what's true
The one thing. Establish which channels are actually producing pipeline, well enough to say it in a sentence to the CEO.
- Get access to the CRM, the analytics, the ad accounts and last year's spend in week one. Flag on day two if any of it takes longer, because everything below depends on it.
- Rebuild last year's pipeline by source myself, once, by hand. Not to produce a dashboard — to find out where the data is untrustworthy. Every company has one source that's secretly a catch-all.
- Talk to six people: both direct reports, two salespeople, the Head of Marketing, and one recent customer if I can get one.
- Ask the two salespeople the same question: "Where do the deals you like actually come from?" Their answer and the attribution data usually disagree, and the gap is the most useful thing I'll learn this month.
What I'd deliver by day 30: one page saying what we know about channel performance, what we don't, and which of the two is bigger. Not recommendations yet.
What I would not do: change any spend. Anything I changed in month one would be changed on the previous person's information, and I'd have no way to tell whether the result was mine.
That last paragraph is the single strongest thing you can put in a 30-60-90. It's a commitment to not act, with a reason, and it directly contradicts the instinct to look productive.
Refusing to touch spend in month one is what makes the rest of the plan credible. Every hiring manager has watched a new marketer arrive, reallocate budget in week three on a hunch, and produce a number nobody can interpret because two things changed at once. A candidate who names that trap before being asked has demonstrated the thing the plan is supposed to demonstrate — that you'd be safe with the budget.
Days 31–60: change one thing, deliberately
The one thing. Make one change big enough to learn from, small enough to reverse.
- Pick the single clearest finding from month one and act on it. If attribution says paid search produces pipeline and sales says those deals never close, that's the first thing to resolve — and it's a data question before it's a budget question.
- Fix the measurement that made month one hard. Whatever the catch-all source was, it gets split properly now, because every decision after this depends on it.
- Give the content person one clear priority instead of five. Most two-person teams are working across too many things because nobody has said what not to do.
Conditional branches, stated:
If month one shows… Then month two is… Attribution is broken Fixing measurement first; no reallocation until we can read the result Attribution is fine, one channel is clearly weak A reallocation test with a defined size and a date to judge it Everything works but volume is too low A capacity question, not a channel question — that's a headcount or budget conversation with you What I'd deliver by day 60: one change made, with what I expect it to do and when we'll know.
The branch table is the part that gets read twice. It shows that the plan is a way of thinking rather than a schedule.
Days 61–90: commit to a number
The one thing. Put a plan against the 40% pipeline increase and say what it needs.
- Take the target apart: what share comes from channels that already work at current efficiency, what needs new capacity, and what would need something we haven't tried.
- Say plainly whether 40% is reachable with the current budget and team. If it isn't, bring the gap with three options rather than a problem.
- Pick the one channel we'd test next quarter, with a size and a kill criterion decided in advance.
What I'd deliver by day 90: a quarter plan with a forecast I'm willing to be measured on, and an explicit statement of what would make me wrong.
Learn, contribute, lead
“30: onboard and audit. 60: optimise campaigns and improve reporting. 90: own the strategy and drive pipeline growth.”
Nothing in it is false and nothing in it is specific to this company, this team, or this target. It could be pasted into any application. The panel's read: this person wrote a plan, not this plan.
One thing per phase, with branches
“Month one: find out which channels really produce pipeline, change nothing. Month two: one reversible change. Month three: a forecast I'll be measured on.”
Three commitments, one per phase, each with a stated reason and a condition attached. It's short enough to remember and specific enough to argue with — and the argument is the conversation you want.
The three questions they'll test it with
"What if you find the channels are fine and the problem is the product?" — they're checking whether you'd say something inconvenient. "Then I'd say so in month one, with the evidence, and the plan changes. Month two becomes working with product on what's losing deals rather than reallocating budget. I'd rather have that conversation in week four than defend a marketing plan for a problem marketing can't fix."
"Thirty days with no changes feels slow. Can you move faster?" — the honest answer, which holds the line while conceding the fair part: "I can move faster on the measurement fixes and the team's priorities — those I'd start in week two. What I'd want to protect is spend. If there's something obviously broken I'd act on it immediately, but 'obviously broken' has to mean obvious, not just suspicious."
"What do you need from me?" — never answer "nothing". "Access on day one, and fifteen minutes in week four when I bring you the channel picture, because some of it will probably contradict what's been reported and I'd rather you heard it from me first."
Presenting it
One page. Read it out in under four minutes, phase by phase, one sentence of reasoning each. Then stop and let them take it apart — the plan is a conversation opener, not a deliverable, and candidates who present it for fifteen minutes lose the part where it does its work.
If the plan is for a marketing role, the questions around it — budget allocation, what you cut, the quarter you miss — are here: Marketing Manager Interview Questions. If it's a growth role, the measurement expectations are different and stricter: Growth Manager Interview Questions.
Practical target: write your month-one section and then delete every activity that produces an output. What's left should be things you'd find out. If nothing is left, you've written a plan for a job you already have — and the panel is hiring for the one where you don't yet know where anything is.
