Most product launches fail as announcements before they fail as products. The work happens in the wrong order. The press release gets written the night before, the reporters get emailed the morning of, and the customers who would actually have cared find out from a tweet.
Here's a six-week countdown that fixes the ordering problem. Compress it to three weeks if you must, but don't compress the sequence.
First, the uncomfortable question
Is the launch news? A new product from an unknown company is not automatically interesting. Ask what specifically changed for someone who isn't you:
- Does it do something that previously required a person, a spreadsheet, or a competitor's product?
- Is there a number (price, speed, scale, a benchmark) that is meaningfully different?
- Is there a named customer already using it with a result?
If the answer to all three is no, you can still distribute a release, which puts the launch on the record and gives you something to link to, but plan for zero earned coverage and budget accordingly. Being honest about this in week six saves you from being disappointed in week zero.
Week 6: decide what the story is
Write one sentence, under 25 words, stating what launched and why it matters. Not a tagline. A news sentence. If you can't write it, you don't have the story yet and no amount of distribution fixes that.
Then lock:
- Launch date and time. Tuesday, Wednesday, or Thursday. 6:00–9:00am ET for the wire. Avoid Fridays, holiday weeks, and the day of a major industry event you aren't part of.
- What's actually available on that date. Not "coming soon." If people can't use it, the news is the announcement of a future thing, which is weaker.
- Who owns the launch. One person, with authority to make calls on launch morning.
Week 5: gather the raw material
This is the week that determines whether the release is good, because it's when you collect the things you can't invent later.
- Numbers. Pricing, availability, performance figures, beta results. Every claim needs a source you could show someone.
- A real quote from a real executive. Get on a call, ask "what's the thing about this you'd defend in an argument?", record it, transcribe it. Do not write it for them and ask for approval, because you'll get corporate sludge. Our guide to press release quotes covers how to run that conversation.
- A customer or beta user willing to be named. This is the single most valuable asset in a launch announcement. One named customer with a specific result beats three paragraphs of feature description.
- Assets. Screenshots, a short demo video, a product image, headshots. Hosted at a stable URL, not attached to emails.
Week 4: write the release
Draft the release now, not later. 400–600 words, standard structure, news in the first sentence. Our product launch template gives you the skeleton and explains each block.
Then write the three shorter things you'll need, because they are not the same document:
- The pitch email. 120 words maximum, why this specific reporter's readers care, no attachments.
- The customer email. Written to a user, not a market. What's new, what to do, what it costs.
- The social copy. One post, the news plus one number.
Also book your distribution now rather than launch morning. Wires have review queues and scheduling cutoffs, and buying at 5am under pressure is how you end up on the wrong package.
Week 3: build the media list
Twenty to forty reporters, not four hundred. The qualification test for each name: have they written about this exact topic in the last 90 days? Not the industry, but the topic. If not, they're not a fit, and adding them costs you future credibility.
For each one, note the article you found them through. You'll reference it in the pitch, and that reference is the difference between a pitch that gets read and one that doesn't.
Week 2: pre-brief and offer early access
Now you reach out, before the news is out. Two plays:
- Hands-on access. For any product a reporter would want to try, offer an account a week early. Reviews need review time. This is the most common reason good products get bad launch coverage: nobody could actually use it before deadline.
- [An embargo](/blog/what-is-an-embargo), if warranted. Ask first, state the exact lift time with a timezone, send the full package after they agree. Skip it if you're pitching cold, because it adds friction to a relationship that doesn't exist yet.
Also this week: brief everyone internally. Sales, support, your investors, your own executives. Every accidental leak I've seen came from someone who didn't know there was a date.
Week 1: final checks
- Every link in the release clicked in a private window
- Pricing page live and correct
- Support team briefed, help docs published
- Proper nouns spell-checked against official sources
- Named contact briefed and free for 48 hours after launch
- A second person has read the release aloud
Wires cannot recall a distributed release. A wrong number lives on syndicated sites permanently.
Launch day: the order matters
Tell people in this sequence. Reversing any two steps creates an avoidable problem.
- 6:00am ET: wire crosses. The release goes out. This creates the public record and the link everything else points to.
- 6:15am: your own site updates. Product page, pricing, homepage. Nothing should be live before the wire, or reporters holding an embargo get scooped by your own marketing site.
- 6:30am: existing customers. In-app notice and email. Your customers should never learn about your product from a news site.
- 7:00am: reporters you pre-briefed. Short note: it's live, here's the link, I'm available all day.
- 8:00am: the rest of the media list. Individual emails, full release pasted in the body. No BCC blast and no attachments. See how to send a press release.
- 9:00am: social and community. Your accounts, relevant communities where you're an actual participant, not a drive-by poster.
- All day: be reachable. The named contact answers the phone. A reporter who can't reach you at 2pm writes around you or drops the story.
- Day 2–5: one follow-up, once. A single short note to non-responders with one new fact: an early adoption number, a customer reaction. Not "just checking in." Then stop.
What to expect
Set expectations before launch day so you're measuring the right thing. Distribution guarantees your release appears on a network of sites and in search results. It does not guarantee a single reporter writes anything. For a first launch from a company with no existing press relationships, a realistic outcome is: the release indexed and syndicated, one or two trade-publication pickups if the news is solid, and a permanent link you can point customers and investors to.
That's a reasonable return. It compounds, because the second launch is easier once you have a record and a few reporters who've heard of you. What doesn't work is treating any single launch as the moment you get discovered.
Build your launch release from a structure that already works, and [see how the builder works](/how-it-works).