The 30-day
playbook.

• Don't skip week one. It's the least satisfying week and it's the one that makes the other three measurable. Every founder wants to start at week three.
• Don't add a channel mid-month. The failure mode this plan is designed against is channel-hopping - trying five things for a week each and learning nothing about any of them.
• Write things down by hand. At this scale a text file beats a dashboard, because the dashboard can't tell you which sentence a customer used before they said yes.

You cannot diagnose anything while nobody arrives and nothing is recorded. This week is plumbing, and it's the week that makes weeks two to four mean something.

Day 1–2 · Instrument it

• Verify Search Console. Four minutes, and search data cannot be backfilled - every day you wait is a day of your launch curve you will never be able to see. Submit a sitemap while you're there.
• Install analytics you'll actually open. Something lightweight. The bar is that you'll look at it weekly, not that it has every feature.
• Define one activation event and check it fires. Not signup - the moment the product has genuinely done its job once. Write the definition down in a sentence. Most people cannot state theirs, which is why most people can't tell what's broken.

Day 3 · Check machines can read you

• View your page with JavaScript disabled. Is the content there? If not, nothing else in this plan matters until it is - crawlers and AI models may be seeing an empty shell. This is common in AI-built sites and completely invisible until you look.
• Read your robots.txt. Decide deliberately whether AI crawlers are allowed. Either answer is fine; not knowing isn't.
• Search Google for your exact domain. If nothing comes back after a few days live, you have an indexing problem, not a marketing problem.

Day 4–5 · Make it legible in four seconds

• Show your homepage to five people who resemble your buyer, for four seconds each. Take it away. Ask what it does and who it's for.
• If three of five can't answer, rewrite the top of the page around what you replace - the spreadsheet, the agency, the four tools taped together. People understand new things by comparison, not by adjective.
• Write one plain sentence naming your category. Models and humans both need it, and "we help teams move faster" gives neither anything to work with.

Not ten users. Ten conversations with people who have the problem, whether or not they ever use your product. This is the week that produces your copy, your channel list and your roadmap - and it's the week everyone tries to skip because it doesn't scale.

Find where the problem is already discussed

• Search the complaint, not the category. Not "invoicing software" - the sentence someone types at 11pm when they're annoyed. The pages that come back are the conversation.
• Follow threads, not platforms. One live thread with forty people arguing about your problem beats a subreddit of four hundred thousand who mostly post memes.
• Look where competitors get complained about. Review sites, support forums, "alternatives to X" discussions. Dense with people already shopping and already articulate about what's missing.

Then actually talk to people

• Answer ten questions properly, with no pitch. If your answer is only useful to someone who buys your product, it isn't an answer, it's an advert - and both the community and you can tell.
• Ask five people what they do about the problem today, and what they tried before. You're listening for the words, not the feature requests.
• Write down the phrases that repeat. Three people using the same phrase is your headline. You will not invent it yourself; you're too close.

Two, not five. One that compounds slowly and one that answers you this week - so you're neither flying blind for a quarter nor starting from zero every morning.

Pick the pair

• The compounding one is usually search: one page answering one specific question your buyer types, written properly. It won't pay this month. It will still be paying next year.
• The fast one is usually direct: the communities from week two, or messaging specific people who have the problem. It tells you within days whether anyone wants this.
• Write down the number you'll judge each by before you start, and the date you'll judge it. For the slow channel the number is leading - impressions and indexed pages, not revenue. Deciding this in advance is the only thing that stops you moving the goalposts in week four.

Use the words from week two

Rewrite your homepage headline, your store listing, or your first paragraph using the phrases that repeated in your conversations. This single change routinely does more than a month of channel work, and it costs an afternoon.

Thirty days is not long enough to judge a channel, but it's long enough to tell which of the five gaps you actually have. That's the decision this week exists to make.

Then do exactly one thing in month two

Not five. The single move that matches the gap you just identified, run for another thirty days with a number attached. The compounding channel keeps running underneath - you don't stop it, you just stop expecting it to have paid off yet.

• Not sure which gap is yours? The diagnostic asks five questions and names it, with three moves in order.
• Stuck on which two channels? The channel picker takes a minute.
• Buyers asking ChatGPT instead of Google? How to get cited covers the newer surface, and it's the least crowded one right now.
• Want someone to look at it properly? The Launch Roast is $99 and comes back in 48 hours against a published rubric.

This page will be updated as the data changes. It's unlisted rather than secret - bookmark it, and share it with anyone who needs it.