A Seven-Day Sprint From Idea to Launch
Planning · 11 min read ·
A day-by-day plan for going from a rough idea to a live, listed product in one week: scope, build, test, publish and listen, with a checklist for each day.
A week is a good unit for a launch. It is short enough to hold in your head, long enough to build something real and long enough to learn from the first reactions. A sprint is, in agile practice, a fixed period of time in which a defined amount of work is completed, and a timebox, the broader idea, is a fixed period that is not extended even if the work is incomplete. Combine them and you get a powerful discipline: this week, we ship.
This guide lays out a seven-day plan, with what to do each day, what to produce and what to cut when time runs out.
Principles of the sprint
Fixed time, flexible scope. The date is fixed; the scope bends. If you run out of time, remove features, not testing.
One user, one job. Pick one kind of user and one task. Resist the urge to serve everyone.
Real, not complete. At the end of the week, a stranger must be able to complete the core task. Nothing else is required.
Daily output. Every day ends with something you can show. A day with no visible output is a warning.
Parkinson's law. The observation that work expands to fill the time available is the reason to set a short deadline. A week concentrates the mind.
Day one: decide
Goal: a written scope and a plan.
- Write the one-sentence scope: "A [user] can [task] in [time]."
- List everything you could build, then sort into must, should and later.
- Pick the simplest technical approach that could work. Do not choose tools to learn them.
- Sketch the main screen on paper.
- Write your definition of done: three or four testable statements.
- Decide how you will listen: a feedback address, basic analytics, error alerts.
Output: a one-page plan you could hand to a stranger.
Cut if tight: the sketch can be a bullet list.
Day two: build the core, part one
Goal: the core action works on a single happy path.
- Build the path from input to result with real data.
- Do not style. Use plain defaults.
- Commit your work often and keep a note of known problems.
- Run it yourself at the end of the day, from scratch.
Output: a thing that does the core action, however ugly.
Cut if tight: anything not on the happy path.
Day three: build the core, part two
Goal: the core action survives ordinary misuse.
- Handle empty, large and unusual inputs.
- Show clear errors instead of blank screens.
- Add the minimum of what you need for people to use it: a way in, such as a simple link or sign-up, if the product needs one.
- Set up deployment so you can publish in minutes.
Output: a hosted version at a real address that works for you.
Cut if tight: accounts. If you can offer the core without sign-up, do.
Day four: harden
Goal: it is safe and reliable enough to show strangers.
- Review security basics: no secrets in code, passwords handled by a trusted library, connections encrypted, dependencies up to date.
- Add a privacy notice if you collect personal data, and make sure you collect only what you need.
- Add error logging and a simple alert.
- Set up backups for anything users create.
- Test on a phone and in more than one browser.
Output: a short checklist, all ticked.
Cut if tight: nothing from this list.
Day five: write the page
Goal: a page that explains the product and invites a first try.
- Write the headline and supporting paragraph.
- Take a real screenshot or record a short clip.
- Add one call to action and a contact route.
- Add an honest line about the stage of the product.
- Check speed and accessibility basics: image sizes, contrast, headings, alternative text.
- Write the one-line description and category for your listing.
Output: a live page, not yet announced.
Cut if tight: the clip. A screenshot is enough.
Day six: test with strangers
Goal: find the problems that stop people, while you still have a day to fix them.
- Ask three to five people who have not seen it to try it, while you watch or while they share their screen.
- Do not explain. Note where they stop, ask or misread.
- Fix the top two problems.
- Retest the main path yourself on a phone.
- Prepare your announcement text and your listing.
Output: a fixed list and an announcement draft.
Cut if tight: secondary problems. Fix only what blocks the core.
Day seven: ship and listen
Goal: it is public, listed and being watched.
- Do a final check of the core path, the page and the feedback route.
- Publish the announcement, honestly describing the stage.
- Submit your listing, with the description and category you prepared.
- Reply to every message and comment you receive.
- Watch errors and fix anything broken the same day.
- At the end of the day, write down what happened, what you learned and what comes next.
Output: a live product, a list of what people said and a plan for week two.
Cut if tight: social posts. The product, the page and the listing come first.
What to do if you are behind
Behind schedule is normal. On the evening of day three, ask: does the core work? If yes, you are fine. If not, decide: reduce the core to something that works, or move the date. Do not skip hardening, testing or the feedback route to hit a date. A product that launches broken costs more than one that launches a day late.
Working with a team
If you have collaborators, split by role.
- Builder: the core action and its hardening.
- Writer: the page, the announcement and the listing text.
- Tester: strangers, phone checks, accessibility checks.
- Listener: feedback, replies and error watching from day seven.
Hold a ten-minute stand-up each morning and a five-minute review each evening: what is done, what is next, what is blocked.
A worked example
A solo developer decides to launch a tool that checks whether a website's contact form actually sends email. On day one he writes the scope: "A site owner can enter a page address and receive a pass or fail on their contact form, in under two minutes." Must: the checker, a result page, an error message, a feedback link. Later: monitoring, history, accounts, alerts.
Days two and three bring a working checker and clear errors; he deploys to a real address. Day four he reviews security, because the tool submits forms on behalf of strangers: he adds limits so it cannot be used to spam, and writes a short statement about what it does. Day five he writes the page and takes a screenshot. Day six four strangers try it; two are confused by the word "probe", so he changes it to "test". Day seven he publishes, lists the product and replies to everyone who writes.
Over the first weekend, sixty people use it. Fifteen leave feedback; the most common request is monitoring. He writes it at the top of the plan for week two.
After the sprint
Take a day to reflect. What went to plan? What did not? Which of the decisions you made in a hurry were wrong? Update the debt list. Choose the next week's scope from what users said, not from what you imagined. Repeat the sprint with a different goal: improving the main path, adding the most requested feature or fixing the biggest drop-off.
Daily rituals that keep the week honest
Morning: read the plan and the definition of done, pick the single most important task for the day and write it at the top of your notes.
Midday: run the product from scratch as a new user. This five-minute check catches regressions early.
Evening: write three lines: what is done, what is next, what is at risk. Commit your work. Stop at a reasonable hour.
These small rituals take little time and make the difference between a week that ends with a product and a week that ends with a pile of half-finished parts.
Questions teams ask
Should we skip a day for planning if the idea is not clear? If the idea is unclear after day one, the sprint is telling you something. Spend the week on a smaller question: build a prototype or a page that tests whether people want it, rather than the product itself.
What if the technology is new to us? Choose the tools you know. A sprint is not the moment to learn a framework.
What about design? Use a plain layout and a single typeface. You can polish later.
How do we avoid burnout? Keep hours reasonable, take at least one proper break each day and treat the weekend as a buffer, not as extra build time.
Should we tell people we are sprinting? Telling a few friends or an online group that you will launch on a particular day creates helpful pressure and a ready audience. Keep the promise modest.
Summary
A week is enough to launch a small, real product. Fix the date, bend the scope, build the core in two days, harden it, write the page, test with strangers and ship on day seven with a way to listen. When time runs short, cut features, not testing, security or honesty. At the end you will have a live product and, more valuably, real evidence about what to build next.
Questions and answers
- Is one week long enough to launch a product?
- For a narrow first version, yes. The sprint is designed to deliver something small, real and testable, not a finished product.
- Can a team of one do this?
- Yes. The plan is built around one person's time, with optional help noted for teams.
- What if I fall behind?
- Cut scope, not testing. Move items to the later list and keep the launch date, or move the date if the core still does not work.
- What should I do on the weekend?
- Plan for rest. A sprint that runs on exhaustion produces mistakes. The plan works across any seven consecutive days.