Quick start

This walks through first sign-in to first results. Budget about fifteen minutes, most of it spent reviewing prompts.

1. Create your account

Sign up at /auth/sign-up with an email and password, or continue with Google. Email sign-ups receive a confirmation link — click it before signing in.

Two-factor authentication is available from day one and is worth enabling now; see Account security.

2. Enter your website

Geonimo runs in team workspaces, and a workspace is created from the site you want to track: enter your domain and the workspace is set up around it.

Even a solo operator gets one, because the workspace carries the brand, the prompt set, the members and the subscription. Its name becomes the slug in every URL (/home/your-workspace/...).

3. Build your prompt set

Prompts are the measurement instrument, and this step decides the quality of everything downstream.

Open Workspace → Prompts and add prompts. The generator audits your website, works out your category and the products you sell, and proposes a set to a mix you choose:

• Brand Visibility — category questions where you want to be named. These may not contain your brand name; a prompt naming you guarantees an answer naming you, which measures nothing.
• Brand Perception — what engines say about you by name.
• Products — chosen by count. Each tracked product gets 4 visibility and 2 perception prompts, which is the minimum that produces a usable confidence interval.

The dialog quotes the totals before you commit, including scrapes per day — the number that recurs on the bill. Review the proposals, drop the vanity ones, and paste in the questions your sales team hears weekly. Those are usually the best prompts in the set.

New prompts are dispatched on the next hourly pass rather than waiting for the nightly run, and the table polls while rows are analyzing.

4. Check your competitors

Workspace → Competitors. Keep the brands you actually lose deals to, and reject the rest — three to six is the useful range. Share of Voice divides among tracked brands, so a bloated set flattens every comparison.

Watch the suggestions over the first fortnight: brands that keep appearing in your answers are surfaced here, and they are often not the ones on your slide.

5. Read your first results

Collection runs daily from 01:00 UTC against ChatGPT, Google AI Overviews, Google AI Mode and Perplexity, with hourly passes picking up anything new.

Start on Monitoring → Overview: Visibility, Brand Rank, Average Position, Sentiment and Grounded Answers, a visibility trend against your competitors with a confidence band, top sources, and recent answers you can open and read verbatim.

Two habits to build immediately:

• Read the noise floor next to each number. A change smaller than the floor is not a change. Day one is a sample; the fortnight is the signal.
• Open the answers. Five minutes of verbatim text explains more than an hour of charts.

6. Invite your team

Workspace → Members → invite by email with a role. Owners manage billing and members; members do the work. See Teams and roles.

A sensible first fortnight

Some screens are still previews on sample data while the rest of the engine is connected — see Product status for exactly which.