Carry the reason for the work all the way into delivery.
Keep the reason for the work attached while you move from a reviewed spec to delivery.
Try saying
“Turn this approved spec into the smallest set of tickets the team can ship.”
Start with the spec
Make the decision and acceptance criteria clear.
Shape the work
Split it into reviewable pieces.
Track the flow
See what is moving and what is blocked.
Feature Marketing
What this is for
Help customers understand what changed, why it matters to them, how to use it, and what result they should expect—without letting the launch story drift away from the product truth.
When this helps
- Release notes explain implementation but not the customer job.
- Marketing writes for everyone because the target segment is unclear.
- Landing pages, emails, social posts, sales copy, and in-product announcements contradict one another.
- Teams celebrate launch activity without agreeing on the customer behaviour that should change.
- AI-generated marketing copy invents proof or capabilities when the evidence is weak.
How it works
Feature Marketing combines one product item, a selected ICP segment, the PM’s release context, and recent Product Graph evidence. It generates one persistent launch brief with strategy, expected input and output, benefits, use cases, setup steps, proof, channel copy, SEO, FAQs, and measurement. Claims supported by workspace evidence are kept separate from assumptions that still need validation. The generated brief remains editable, and nothing is published or sent automatically.
Before you start
- An existing feature, opportunity, roadmap item, experiment, ticket, or strategy card—or a clear name for a new feature.
- The ICP segment this release should help. If no segment exists, the brief explicitly treats audience claims as assumptions.
- A plain-language explanation of what changed, why it matters now, setup requirements, and limitations.
- Real customer and product evidence in the active workspace when available.
What you should see
- Audience, problem, positioning, value proposition, why-now, and differentiation.
- A detailed explanation of what the feature does, expected inputs and outputs, benefits, examples, and setup steps.
- A proof check separating supported claims, open assumptions, and real evidence links.
- Landing-page opening, FAQs, release notes, customer email, in-app announcement, social posts, and sales talk track.
- SEO title, description, keywords, and a search-result preview.
- A launch goal, primary metric, guardrails, and review window.
Try it now
- Open Market → Feature marketing and choose Create launch brief.
- Select a product item, or describe a feature that is not yet represented in the Product Graph.
- Choose the ICP segment. Use one audience per brief so the problem and promise stay specific.
- Explain what changed and why the audience should care. Include limitations so generated copy does not overpromise.
- Choose Create evidence-backed brief. Specky reads the product item, ICP, and real workspace signals before drafting.
- Review Product story first. Inspect the proof section and move unsupported customer claims into assumptions.
- Review Launch kit, SEO, and Measurement. Copy only the assets that fit the channel and agreed release stage.
- Choose Edit copy to make the story sound like your team, then save the reviewed version.
- After launch, open Experiments or Outcome Intelligence and compare the agreed primary metric with the customer response.
A real example
A growth PM launches a workflow that answers “Why did signups change?” The selected ICP is a growth PM at a scaling SaaS company. Specky drafts a page that explains which live sources are checked, what the investigation returns, and when it will say the data is insufficient. It creates release notes, an email, an in-app announcement, SEO questions, and a metric based on reviewed investigations—not an unsupported promise that AI always finds the cause.
Your next move
Use Validation Pages when the promise still needs testing, and Outcome Intelligence when the feature has shipped.
Ready to apply this?
Start with your own product and keep the first read grounded. You can create an account after you see the result.