Quickstart

Start onboarding with your public website address. Searchly uses the website to suggest a brand name and project context. Check the name before continuing, especially if your domain and brand name differ.

Choose the region and language you want the project to use. Then select topics that match your products, services, or customer questions. You can keep a suggested topic or add your own.

A project holds one brand's tracking configuration and results. Use the project switcher in the sidebar to check which project you are working in.

Read the suggested prompts before accepting them. Edit questions that use the wrong product name, target the wrong audience, or cover a service you do not offer. Remove questions you would not use to judge your visibility.

Keep a mix of discovery, comparison, and decision questions where those fit your business. A prompt that names your brand tests a different situation from a prompt that asks for recommendations without naming you.

You can add, archive, or restore prompts later in Prompts. See Manage prompts for the steps.

Review the suggested competitors and keep the brands you want to benchmark. Add another brand if the suggestions miss a direct competitor.

Review the plan and model coverage shown during setup. The models and prompt capacity available to your project depend on your workspace's plan and configuration. After setup, open Settings and choose Models (LLMs) to review coverage and prompt allocation.

You can change the comparison set in Knowledge Base, under Competitors. Compare competitors explains how that affects the reports you read.

Open AEO Insights when the first report is ready. Searchly shows a progress or incomplete state while the report is still being prepared. An incomplete report is not a measured zero.

1. Check the project name and date range.
2. Open Visibility and review your brand alongside the tracked competitors.
3. Filter by a topic or model to inspect one part of the result.
4. Open Mentions & Citations and read an individual answer.
5. Open Sources to see which domains and pages appear in the collected answers.

If the selected period has no observations, widen the date range or clear the topic and model filters. Understand your metrics explains the difference between missing data and a measured result.