AI Made Product Teams Ship Weekly. Marketing's Launch Calendar Broke.

Rippling’s marketing team now plans launches on a 6 to 8 week cycle while product can release 1 or 2 things a week.

That gap is where the old launch calendar breaks. Product ships when the work is ready. Marketing still needs to decide which release deserves a campaign, who needs to hear about it and whether there is a story worth interrupting customers for.

Kyle Poyar’s new interviews with marketing leaders at Rippling, Notion and Profound show the operating change in progress. AI coding tools have made releases faster. Marketing needs a separate decision for the public moment.

Trevor Pyle, Co-founder at Profound, put the distinction cleanly:

“We get things to customers lightning fast, then we pick the market moment.”

Those are 2 different decisions. Treat them as one and marketing becomes a bottleneck. Treat every release as a public event and customers stop paying attention.

The tier is now the work

Notion’s own launch framework makes the basic point. Major product launches get a campaign and cross-functional coordination. A smaller feature might need an announcement, sales material and support documentation. A minor update can be release notes plus internal awareness.

That distinction used to be a way to manage scarce marketing capacity. It is becoming a way to manage attention.

At Rippling, Poyar reports, launches are tiered S through XL. A small release gets a customer email and LinkedIn post. An XL launch can mean 100 marketing assets. The discipline is deciding early what level of work the customer consequence earns.

Ben Pollack, Senior Director of Marketing at Rippling, described where the extra time should go:

“The positioning statement, key message, claim architecture, problem-solution-payoff, and reasons to believe - that is the surface on which we directly spar.”

The details after that can move fast. The core claim can’t. Profound’s FactCheck launch is a good reminder that stale product claims become an AI-search problem as soon as the product changes. If the product team ships at 11pm, AI can help turn a settled message into a launch email, a page update, sales snippets and social drafts by morning. It can’t decide whether the feature matters, what promise the company can defend or whether the announcement should wait for a customer proof point.

Build before the splash

The old model tied the public announcement to release day. That made sense when release day was rare and the company needed one coordinated moment.

The teams Poyar interviewed are using the gap differently. Notion has product people build in public before a formal launch. Profound has released product to customers, kept talking about it and chosen a later market moment. The customer gets the capability when it’s ready. Marketing gets time to find the story.

This is a case for treating launches as a content system, not a calendar invite. Early access can create customer feedback, a clearer demo and a reason to announce the same product more than once without pretending it is new each time.

That matters most to small teams. They don’t have enough people to do a full campaign for every release. Product velocity can turn the roadmap into a feed of unprioritized announcements. The handoff problem already showed up in adjacent product teams. AI agents also create more briefs, not fewer.

The calendar needs a gate

There is one important caveat. This is a 3-company B2B software case study, not a representative survey of every marketing organization. It also comes from a newsletter whose job is to surface operating playbooks.

The pattern is still useful because each company describes a decision marketing teams can make now: classify the release, lock the message before production starts, release the product when it is ready and spend the campaign when there is something to say.

The gate should be simple. Before the team starts making assets, answer 3 questions.

What changed for the customer? What proof makes the claim credible? Why does this need a public moment now rather than a release note?

If the answer to the third question is “because it shipped,” the feature probably belongs in release notes. Shipping is no longer enough reason to speak.