Organizing for Scale and Experimentation
Key Concepts
Tags vs. Topics - Topics are buckets that contain prompts. Tags cut across topics. A single tag can apply to prompts in any topic, which is what makes them a cross-sectional filter rather than a folder.
Intent Tags - Tags that classify a prompt by its search behavior type (commercial vs. informational). Informational prompts behave differently in AI responses, so mixing them into commercial tracking distorts your visibility score.
Aggregate Visibility Score - A rolled-up metric across your tracked prompts. Mixing prompt types without intent tags can make this number misleading.
Quick Reference
Tagging prompts within a topic
- Go to Prompt Designer and select a topic
- Select one or more prompts
- Click Edit Tags, type a tag name, and click Add Tag
- Apply before navigating away - switching topics clears your selection
Tagging across topics (scalable method)
- Select All Topics in the topic dropdown
- Use the keyword search to find prompts by pattern (e.g.,
best, what is, pricing) - Select matching prompts, click Edit Tags, and apply in bulk
Tagging an entire topic
- Open a topic in Prompt Designer
- Select all prompts
- Apply a campaign or experiment tag in bulk
- Repeat for each additional topic that belongs to the same initiative
Filtering with AND / OR
- AND - prompt must match all selected tags (narrower, e.g.,
commercial AND Q1 2026) - OR - prompt matches any selected tag (broader net)
Setting up a watch list
- Navigate to the Citations tab
- Scroll to Watched Pages
- Click the star icon on any citation to add it to your watch list
- You can also add URLs for pages not yet published
- Export your watch list as a CSV for tracking over time
Experiment tagging workflow
- Identify the prompts your experiment is designed to move
- Tag them before you ship (e.g.,
exp-landing-page-test) - After launch, filter by that tag to read rank and citation impact in isolation
- Your baseline prompt set stays clean and unaffected in reporting
Recommended tag palette (start small)
- Intent:
commercial, informational - Initiative:
campaign, strategy - Experiment:
exp-[descriptor] - Time/cohort:
Q1-2026, Q2-2026