First prompt run

Run trigger

The first prompt run starts the evidence loop. It can be triggered by the initial workspace release or by a prompt action once setup is ready.

The goal is to create a baseline.

Run a buyer-style prompt.

Keep the answer as evidence.

Review mentions, competitors, and sources.

Create one useful next step when the signal is clear.

The first prompt set should include buyer-style questions. It should be specific enough to reveal brand mentions, competitors, and sources.

Avoid trying to cover every market in the first run.

Prompt results can vary by model and market. Keep the first setup simple enough that you can understand what changed.

Add more coverage after the first results are reviewed.

Keep the first run simple enough that mentions, competitors, and sources are easy to inspect.

The answer is saved as a chatlog. This gives CLEA evidence to inspect later.

Chatlog capture is what makes follow-up analysis possible.

Sources show which pages or domains influenced the answer.

Review sources before deciding what content or authority work to do.

Keep the full answer.

Review pages and domains behind the answer.

Let mentions, competitors, and sources update.

Start with prompt cleanup, readiness, or one card.

After the run, analytics can start showing prompt evidence, mentions, competitors, and sources.

Some panels still depend on connected integrations.

The best first actions are prompt cleanup, readiness review, competitor review, source inspection, or one suggestion card.

Do not turn one first run into a huge strategy. Look for the first useful correction.

Create a prompt run

First data arrival