An opportunity report has three acts. After "Is it real?", the last two turn the gap into a plan: "What to build" describes the product, and "How to ship" says how to test it and reach people.
When these acts appear
Both acts belong to a "Report v2" report, where AI writes them from the stored complaints. A section with nothing to show is left out: "Pricing & ways to monetize" needs a suggested price or a way to monetize, and "Names" needs at least one name. "Build & launch" is on every "Report v2" report.
An older report without that chip has no "What to build" act. Its "How to ship" act can still hold a "Validation checklist" and "Names" from its first version, with a domain for each name and a reminder to check it.
What to build
Every section here carries the "AI" label, with a line saying when it was written and from how many quotes. Many statements end with "cites" and a number of quotes: the complaints, still in the evidence, that they rest on.
- "Who it's for": a short portrait of the people with this problem, a few traits, and a "Not for:" note when there is one.
- "MVP": the smallest first version worth building. Its features are ranked by how many complaints each one answers, and "Answers:" names the complaint theme behind a feature. "The wedge" is the angle to win with, and "Left out of the first version" lists what to skip for now.
- "Pricing & ways to monetize": a "Suggested price (AI)" a month or a year, in US dollars, with its reasoning, then "Ways to monetize". AI suggests it; it is not a measured price.
- "Risks": each risk with a level ("High", "Medium" or "Low"), a "Mitigation" (how to reduce it) and a "Kill threshold": the sign that tells you to stop.
How to ship
- "Validation plan" (AI): a short summary and numbered steps to test the idea with real people. Each step has a "Metric:" to watch and, where it fits, a "Stop if:" line.
- "First customers" (AI): channels where people with this problem gather, each with a first step to reach them. A channel links to the place only when its address is already in TrendGap's data for this idea.
- "Build & launch" (Standard): a playbook for the idea's track, with topics such as the stack and payments. It is the same across the track and never written by AI. Click "Standard playbook for …" to unfold it.
- "Names" (AI): name ideas for the product, most with a short reason. The report does not check their domains.
Rate, flag or re-run a section
On paid plans, each of these sections ends with "Was this useful?": vote, or click "This is wrong" to flag it. On Pro and above, "Re-run this section" asks AI to rewrite a section it wrote, and you see the price before anything runs. "Build & launch" takes votes and flags, but no re-run. See Rate, flag or re-run a report section.
What the Free plan shows
On today's free idea and the free previews, Free sees both acts in full, without the "Was this useful?" row. Other reports open as a "Member report", which lists these sections under "In the full report", each with a lock. See The free plan.
Where to go next
What you see here depends on your plan and your account.
- "Build this with AI" turns this plan into a prompt for AI coding tools and app builders. See Build this with AI.
- "Start a project" keeps your work on this idea in one place. Projects are part of the paid plans. See Projects.
- In Build tools, the "Name and domain generator" checks whether a name's domain looks free, and on paid plans the "First customers finder" searches for real communities. See Build and launch tools.
- Before you create a "Share link", untick "Names" or "Pricing & ways to monetize" under "Sections to show" to keep them to yourself. See Share links, PDF and export.