A topic is a subject you want to own in AI answers (for example, "time tracking for agencies"). Stride represents each topic as five archetype prompts — the five ways real buyers ask about a subject — and scores ownership per engine.
What it is
- Definition — "what is …"
- Comparison — "X vs Y …"
- Alternatives — "alternatives to …"
- Use case — "best … for a specific situation"
- Buying — "how to choose / where to buy …"
Stride scans those five prompts across your entitled AI engines and rolls them into one Topic Authority picture: whether you own the conversation or a rival does.
Where to find it
Dashboard → Visibility → Topics. Open a topic for coverage by question shape, rival standing, trend, and leadership flips. Analytics also shows a compact Topic ownership card.
How to use it
- Create a topic with a plain-language name (or track one suggested from your prompts).
- Edit the five archetype prompts so they sound like your customers.
- Scores fill from recent scans (or run a one-credit baseline).
- Prioritize At risk and Rival-owned topics, then Attack windows.
- Pause topics you are not working so their slots free up for active ones.
From insight to action
When a topic has a durable At risk, Attack window, Rival-owned, or missing-question-shape signal, Stride creates at most one Topic Authority action for that topic. It combines the affected buyer-question shapes and engines into one coordinated recommendation rather than adding five unrelated prompt tasks.
The signal first appears in Actions → Watching after a trusted full scan. It enters the active queue only after the same condition appears in a second comparable trusted full scan. Daily priority scans and incomplete evidence do not create or advance a Topic Authority action.
On the Topic detail page, you can explicitly review that first trusted signal as one Topic coverage project without waiting for queue promotion. This does not claim the signal was repeated or improved; it records your decision to coordinate the evidenced work now. Review the proposed buyer-question workstreams and choose the ones that belong in this push. Stride creates only those selected workbenches and preserves the source prompt, result, and scan evidence on each one. In every workbench, you explicitly choose whether to strengthen an existing page or create complementary coverage. Stride does not select a target, draft, or publish content automatically.
Project progress and project proof are separate. Finishing every selected workbench means the execution is complete; it does not claim the Topic improved. Stride waits for a later comparable scan with the same five-prompt benchmark, then reports the measured from-to change. If the prompts or scan configuration changed, the project is labeled Not comparable instead of claiming a win.
If an older Topic contains fewer than five benchmark prompts, Stride labels the unconfigured question shape separately. It is excluded from measured coverage and project work rather than being presented as an answer where your brand was absent.
Definitions
Ownership — you own a topic on an engine when you appear in at least 4 of 5 archetype prompts and lead the next brand by 5 points or more of mention-share. "Appear" means the engine either names your brand or cites your site.
| State | Meaning |
|---|---|
| Owned | You meet the ownership bar |
| You lead | Ahead, but not by the full ownership margin |
| Rival leads | A competitor is ahead |
| Rival-owned | A competitor meets the ownership bar |
| Unsettled | No brand has a clear lead |
| Insufficient | Not enough scan data yet |
Attack window — the leader's grip is weak, so a focused push can flip it.
Plan & limits
Topics are pooled across your workspace. Active topic caps:
| Plan | Active topics |
|---|---|
| Starter | 3 |
| Grow | 10 |
| Scale | 25 |
| Agency | 50 |
The +5 Topics add-on ($29/mo) raises the cap. Paused and archived topics do not count.
TODO — needs review
Prompt Explorer needs documentation here or in a dedicated Prompt Tracking article. It suggests trackable buyer questions from your Search Console queries, a seed phrase, or one of your page URLs, and generation is metered; tracking a suggestion uses the same flow (and plan caps) as adding a prompt manually on the Prompts page.
FAQ
Why does a topic say Insufficient? Not enough of the five prompts have been measured yet — wait for the next full sweep or run a baseline.
Does this replace Prompts? No. Prompts track individual questions; Topics answer whether you are the default for a whole category.
How do I rename a Topic? Use Rename beside the Topic status controls. Renaming changes the label only: the five benchmark prompts, status, measured history, and project links stay with the Topic. If the name is already used, open that existing Topic instead of creating a duplicate.
How do I replace a benchmark prompt? Open Tracked buyer questions on the Topic detail page and choose Edit beside the prompt. Stride keeps the old prompt and its measurements in history, then starts a new measurement baseline for the replacement. Current standing and new coverage projects wait until a completed full scan measures all five current prompts. Replacing a prompt does not start a scan or spend a scan credit automatically. Restore an archived Topic before replacing one of its prompts.
Why isn't a wording variant shown as a Topic opportunity? Opportunities are compared with Topics you already created before they are ranked. Narrow wording variants such as “tracker” and “tracking” resolve to the same category only when their remaining scope also matches. “Time Tracker Mac” is therefore covered by “Time Tracking Mac,” while “Asana time tracking” remains a separate category.
Why was a similar Topic blocked or flagged? Exact duplicate names and exact five-prompt benchmark sets are blocked. The automatic AI check flags only likely merge candidates that cover the same buyer ground and the same scope despite wording differences such as synonyms, abbreviations, or spelling variants. A shared category term is not enough: Topics with distinct buyer jobs, audiences, platforms, integrations, locales, or product scopes remain separate. Topic boundaries to review lets a human merge a true duplicate or record a boundary if needed. Merging archives and links the Topic you do not keep; it never rewrites historical scans, and the AI check never merges or archives a Topic on its own.
What does “question shape is not tracked” mean? The Topic does not have a benchmark prompt for that buyer-question shape. Use Add prompt in the recommendation, review the proposed wording, and save it. The prompt joins future full scans. Existing measured findings remain visible, while coverage project creation waits for a completed trusted full scan to measure all five prompts. A running scan or daily priority scan does not clear this state.
Related
- Scheduled scans — when measurements refresh
- Competitive intelligence — rivals in answers
- Visibility Index & dashboard — measured outcomes and the secondary diagnostic index