Documentation

Set up tracking

Tags

Group prompts with tags (one per prompt) and filter every dashboard figure by them, to compare funnel stages, product lines or markets inside one domain.

A tag is a named, colored label you put on prompts. Its value is in filtering: pick a tag at the top of the Dashboard and every KPI, chart and table recalculates for just those prompts. That turns one blended visibility score into answers to sharper questions, like “are we visible for comparison prompts?” or “how does product B do on its own?”.

How tags work#

  • One tag per prompt. A prompt has at most one tag, or none. Design your tags as one dimension (for example funnel stage) rather than mixing several.
  • Per domain. Tags belong to a domain. Each domain has its own set, and a tag cannot be applied to prompts of another domain.
  • Name and color. A tag has a name of up to 50 characters, unique within the domain, and a color used wherever the tag appears.
  • Plan limits. Tags per domain: Trial 5, Starter 20, Pro 50, Premium and Enterprise unlimited.

Create and apply tags#

  1. While adding prompts

    In the Add prompts dialog, open the Tag picker. Choose an existing tag, or type a new name and pick Create “{name}”. The tag is applied to every prompt in the batch. See adding prompts.

  2. On existing prompts

    On the Prompts page, tick the prompts to tag and use Tag: in the bulk bar. To change one prompt, open Edit prompt and pick a different tag, or No tag.

  3. Rename, recolor or delete

    Open Manage tags from the tag picker. Each row has a color button, an editable name and a delete action. Add tag creates a new one. Deleting a tag leaves its prompts untagged; their runs and results are kept.

Tags also appear in the setup flows for competitors: when you add a competitor’s ranking keywords as prompts, you can tag them in the same dialog, which is a clean way to keep “competitor keyword” prompts separate from your own.

Filter by tag#

The tag filter sits in the header of the Dashboard, Prompts and Sources pages, next to the engine, country and date filters. The default is “All tags”. With a tag selected:

  • Visibility score becomes the share of that tag’s prompts where you were named.
  • Share of voice, Avg rank and the charts only count answers to that tag’s prompts.
  • Sources only lists domains and URLs cited for that tag’s prompts.
  • The change indicators compare the same tag against the previous period.

Worked example#

You track 60 prompts: 20 tagged category, 25 tagged comparison and 15 tagged branded. Your overall visibility score is 45.0% (27 of 60 prompts). Filtering by tag shows:

TagPrompts where you were namedVisibility
category3 of 2015.0%
comparison10 of 2540.0%
branded14 of 1593.3%

The blended 45.0% hides the real story: branded prompts carry the average, and you are nearly invisible on the category questions where buyers form their shortlist. That is where to focus.

Tagging schemes that work#

Pick one axis and apply it to every prompt:

  • Funnel stage: problem, category, comparison, alternatives, branded. The most useful default. See choosing the right prompts.
  • Product line: one tag per product or service you sell, when a single domain covers several.
  • Market: us, de, fr when you track translated prompts in several countries. The country filter already splits by country, so use this only when prompt texts differ by market.
  • Source: competitor-keywords, gsc, sales-calls, to judge which source of prompt ideas pays off.

Keep branded prompts in their own tag

Branded prompts almost always mention you, which inflates the overall score. Tagging them branded lets you check your visibility on everything else in one click.