Client Creation & GEO Analysis Scenarios — Full Customization

Overview — Two Ways to Create a Client

AILabsAudit offers two methods to create a client. Both methods lead to the same result: a fully configured client profile with GEO-optimized analysis scenarios ready for your first audit.

Regardless of the method, once the client is saved, AILabsAudit automatically generates a complete set of GEO/AEO-optimized scenarios customized to the client’s profile.

Automatic website URL validation

Before any client is saved, AI Labs Audit checks that the website URL really resolves. Auditing a client whose site is dead would burn credits for nothing — so the validation runs live, directly in the form, the moment you finish typing the URL.

• Real-time check. A HEAD request is sent with a 6-second timeout. The result is shown next to the field: green tick (ok), orange warning, or red error.
• "Save" is disabled if the site returns an error, times out, or the URL is malformed. No more clients created on a dead domain by mistake.
• Auto-prefix. If you type example.com, the form silently completes it to https://example.com before validating.
• REST endpoint. The check is exposed at GET /api/client/check-url?url=... — useful if you create clients programmatically via the API or MCP.
• Daily re-check. A scheduled task daily_check_client_websites runs every day at 04:10 server time and re-tests every client website. Yesterday it passed, today it returns 503? You will know it without running an audit.
• Visible status. The client card displays the latest status: ok, http_error, timeout, unreachable or invalid_url.
• "Site KO" badge. After 7 consecutive days of failure, a red Site KO badge appears on the client card so the issue cannot be missed.

AI-Assisted Client Creation (Recommended)

The fastest way to create a client. Perplexity Sonar, a powerful LLM with real-time web access, analyzes the website and extracts all relevant information automatically.

What the AI extracts

• Sector / Industry: the business category (e.g., "Boulangerie-patisserie", "SaaS B2B", "Cabinet d’avocats")
• Products / Services: main offerings listed on the website
• Competitors: top 5 competitors with their website URLs
• City: the business location (headquarters or main office)
• Description: a concise summary of what the company does and its differentiators
• Social media: Facebook, Instagram, LinkedIn, Twitter/X, YouTube, TikTok URLs when available

Manual Client Creation

If you prefer full control or if the client does not have a website, you can fill the form manually. This method costs no credits.

Required fields

• Company name: the official name of the business
• Website: the main URL (optional for manual creation)
• Sector: select from the dropdown or enter a custom sector
• City: the business location for local GEO relevance

Recommended fields

• Products/Services: comma-separated list of main offerings — used as {produit} variable in scenarios
• Competitors: add each competitor with their name and website URL — used as {concurrent_1} to {concurrent_5}
• Description: a concise paragraph about the company and what makes it unique
• Differentiators: what sets this client apart from competitors (used for E-E-A-T signals)

When to use manual creation

• The client does not have a website yet
• You are creating an internal or test client
• The business operates in a very specific niche that AI might not categorize correctly
• You want to save the 20 credits from AI research

Auto-Generated GEO Analysis Scenarios

After client creation, AILabsAudit automatically generates a complete set of customized analysis scenarios covering 8 GEO/AEO themes. These scenarios are the questions that will be sent to AI models during audits.

8 GEO/AEO themes covered

Variable system

Scenarios use variables that are automatically replaced with client data during audits:

• {entreprise} — the company name
• {produit} — the main product/service
• {secteur} — the business sector
• {ville} — the city
• {site_web} — the website URL
• {concurrent_1} to {concurrent_5} — competitor names

Using Base Analysis Scenarios

Base analysis scenarios are your reusable library of scenarios that can be applied across all clients. Instead of recreating the same scenarios for each client, create a base scenario once and assign it to any client.

Why use base analysis scenarios?

• Standard audit questions: questions you ask for every client (e.g., "What is {entreprise} known for?")
• Industry-specific scenarios: scenarios tailored to a sector (e.g., restaurant-specific questions)
• Branded scenarios: if you run a white-label agency, create scenarios consistent with your methodology
• Compliance scenarios: standard E-E-A-T verification questions

How to create a base analysis scenario

Manual Analysis Scenario Creation

Create analysis scenarios from scratch for specific needs that go beyond the auto-generated set. Use the variable system to keep your scenarios dynamic and reusable.

Available variables

Writing effective analysis scenarios

• Be specific: "What are the best {produit} shops in {ville}?" is better than "Tell me about {entreprise}"
• Include competitor context: "Compare {entreprise} and {concurrent_1} for {produit} in {ville}"
• Test with the preview: see exactly how the scenario will look with the client’s data before saving

Categories and weights

Each scenario has a category (visibility, authority, sentiment, etc.) and a weight (1-10). The weight determines the priority of the scenario in the audit — higher weight scenarios contribute more to the overall score.

AI Model Configuration

Select which AI models to query for this client. Model configuration is per-client, meaning different clients can use different model combinations.

Native mode vs. Web search mode

Each model can be queried in two modes:

• Native mode: queries the model’s training data only. Results reflect what the model "knows" from its training cutoff. Best for measuring how well a brand is embedded in the AI’s knowledge base.
• Web search mode: adds the :online suffix, enabling real-time web search results. Best for measuring current visibility and how the brand appears in live AI-powered search.

Recommended model selection

For a comprehensive GEO audit, we recommend selecting at least one model from each major provider:

Creating Clients via MCP

The MCP (Model Context Protocol) server allows you to create clients directly from AI-powered IDEs and assistants like Claude Desktop, Cursor, or Windsurf.

How it works

Importing and Bulk Operations

Import clients from CSV

If you manage a large portfolio, you can import clients from a CSV file. The CSV format supports the following columns: company_name, website, sector, city, products, description.

Duplicate a client’s analysis scenario set

When you create a new client in a similar industry, you can duplicate the scenario configuration from an existing client. This copies all custom scenarios and their weights — saving you from recreating the same setup.

Export client data (GDPR)

For GDPR compliance, you can export all data associated with a client: profile information, audit history, scenarios, scores, and reports. The export is available in JSON format from the client’s detail page.

Frequently Asked Questions