Product launch plan: six phases, one owner each, a gate before every step

A product launch plan is the one-page sequence that says who does what, which week it is due, and what must be true before the next phase starts. It is not the list of assets and it is not the content calendar. This chapter gives six phases, a copyable table, and the gates that stop a launch from turning into a list of tasks nobody owns.

What a product launch plan is, and what it is not

The plan answers four questions in one place: what has to exist, who makes it, when it is due, and what has to be true before the next step starts. A checklist answers only the first. A content calendar answers only the third, and only for editorial work. Teams that skip the plan usually own all three documents and still miss the launch, because no single page says which gate is currently closed.

The distinction matters when something slips. With a plan, a missed asset blocks a named channel and everyone can see it. Without one, the missed asset is discovered on launch morning, when the only remaining option is to publish something weaker and call it done.

One boundary before the mechanics: a product launch plan coordinates work you control. It cannot approve your directory submission, decide whether a community moderator accepts your post, or make an AI engine cite the page. Those outcomes belong to the platform. The plan can only make sure you arrive with the material and the sequence in place.

The six phases, and the gate between them

Six phases are enough for a small product launch, and each one ends with a gate rather than a date. The gate is a yes or no question. If the answer is no, the next phase waits. Dates still exist, but they describe the intended order instead of pretending to be a promise.

For a one-person product, the owner column repeats the same name six times. Write it anyway. The column is there to expose the week when you are the bottleneck in three phases at once, not to assign blame later.

Set the goal and one success number first

Phase one produces one sentence and one number. The sentence names the audience and the action: for example, a developer who has never heard of the product should be able to explain it after reading the listing. The number is the single measure you will look at on week +2. It can be sign-ups, activated accounts, or reply rate from one community. It cannot be five numbers, because five numbers have no owner.

Pick the number before you choose the channels. If the number is qualified sign-ups, a broad directory with no filtering may add noise. If the number is feedback from experienced users, a smaller community is worth more than a large listing. The choice of channel changes with the goal, so choosing channels first quietly decides the goal for you.

Write the stop date too. A launch plan without a stop date keeps absorbing small improvements and never ships. The stop date is the day you accept the current assets as good enough and move to the channel step. If the assets are not ready by then, the honest move is to reduce the scope of the launch, not to move the date for the fourth time.

Build the asset gate before you book a channel

The asset phase ends when the same facts appear in every file. Product name, category, pricing, platform, language and link should match across the profile, the one pager, the screenshots and the submission form. This sounds trivial until a directory displays last month's price or the old product name, and the listing looks careless before anyone reads the description.

Use the product launch checklist as the working document for this phase. The checklist tells you which fields to fill; the plan tells you when the fields must stop changing. Freeze the facts one week before the first submission and treat later edits as corrections, not as normal work.

Two assets deserve an explicit gate. The first is the demo link: check that it opens without a login, that it loads on a phone, and that the first screen explains the product without narration. The second is the screenshot set: use real product screens, not a stock image, and make sure the text in each screenshot is legible at listing size. A directory card is small, and a screenshot that only works at full width is not an asset.

Sequence the channels instead of submitting everywhere at once

The channel phase produces two lists: a shortlist you will submit to, and a deliberate cut list you will not touch this cycle. The cut list is the important one. Without it, every directory feels possible, and the launch turns into a week of form filling with no time to answer anyone who replies.

Score each candidate on audience match, language, cost, moderation rules and the work required after submission. The scoring method is in how to choose startup directories. Then order the shortlist by the amount of attention each channel needs, not by the size of the audience. A community post needs you present for the replies; a directory form does not. Put the high-attention channel on a day when you can stay with it.

Do not submit to every channel on launch day. A launch is easier to read when the same product appears in different places over two weeks, and corrections are easier when you still remember which text went where. The content side of that schedule belongs in the editorial calendar and content calendar, which should be a subset of this plan, not a separate plan.

QueryWin keeps the distribution side in one place: a product profile is written once and reused when you prepare channel submissions, while the submission itself stays with you. See how the distribution workflow works if the hand-off between profile and form is the part you keep redoing.

The deliverable: a one-page launch plan

Copy the block below into a document and fill it in. If a field cannot be filled with a name, a date or a link, the plan is not ready. That is the whole test. A plan with empty fields is a wish list with headings.

PRODUCT LAUNCH PLAN
Product:
Launch goal (one sentence):
Success number (one number, measured on week +2):
Stop date:

PHASE 1 GOAL
Owner:
Gate: goal written, number chosen

PHASE 2 ASSETS
Owner:
Profile link:
One pager link:
Screenshot folder link:
Demo link:
Gate: facts match across all four

PHASE 3 CHANNELS
Owner:
Shortlist (channel, why it fits, effort):
Cut list (channel, why not this cycle):
Gate: every shortlisted channel has a reason

PHASE 4 COPY
Owner:
Draft link per channel:
Gate: each draft names the audience and the action

PHASE 5 LAUNCH
Owner:
Submission log (channel, time, URL, status):
Gate: every submission logged on the day it happens

PHASE 6 FOLLOW-UP
Owner:
Week +1 check:
Week +2 check:
Gate: one measured result, and one decision for next cycle

The plan stays on one page on purpose. A longer plan becomes a second product to maintain, and the maintenance work lands on the same person who is supposed to be answering community replies. One page is enough to expose an unowned task.

What goes wrong in the last two weeks

The most common failure is a channel added after the cut list was made. Someone suggests a launch platform on day -3, the team submits without a channel-specific description, and the listing reads like the generic profile. The result is not a disaster, but it is a wasted submission. When a new channel appears late, add it to the next cycle unless it comes with a person who will own the copy and the follow-up.

The second failure is a gate that is treated as a date. A date can slip quietly. A gate cannot: either the facts match or they do not. Make the gate a visible checklist item and refuse to start the next phase while one box is open. This is slower for two days and faster for the remaining four weeks.

The third is measuring too many things. On week +2, pick the one number from phase one and write down what it was. Everything else is context. If the number moved, ask which channel plausibly contributed; if it did not move, ask which channel was never live. Both answers are more useful than a dashboard with twelve charts and no decision.

Where this plan stops

This plan does not guarantee acceptance, coverage, citations or traffic. We do not know your approval rate at any directory, and we cannot tell you whether a moderator will read your post as promotion. The plan only makes the work visible and gives you a clean place to record what happened.

It also assumes a product that someone can already use. If the product is still a waitlist or a prototype, the same phases apply, but the asset gate changes: the demo link may become a recorded walkthrough, and the success number may be qualified waitlist sign-ups. Do not copy the launch-week steps from a public launch and expect the same result.

If you have less than two weeks, skip the six-phase version. Run the pre-launch asset checklist, pick one channel, and write down the result. A small launch with a recorded outcome is worth more than a large plan that never leaves the document.

Common questions

How long should a product launch plan be?

One page. If it needs a second page, the extra page is usually a task list for one phase, and that task list can live next to the work instead of inside the plan.

Who owns the plan when there is no marketing team?

The founder owns it, even when other people do the work. The owner is the person who closes the gate, not the person who writes every line.

Should the plan include paid channels?

Only if you have already decided to pay and have a budget line. QueryWin does not buy traffic or run ads, and a launch plan that assumes paid reach is a different plan with a different risk profile.

What if the launch date moves?

Move the date, not the gates. The order of the phases is what protects the assets. If the date moves, check which gate is still open and fix that before rescheduling the submission.

A launch plan is not a promise that the launch will work. It is the page that tells you which part is not ready.