Your First Project in 10 Minutes — BrandWise Step-by-Step Guide

What you'll need to get started

Before you create a BrandWise project and run your first brand evaluation in ChatGPT, Claude, Gemini, and other models, prepare the following:

• Email address — for registration and receiving evaluation completion notifications.
• Brand positioning clarity — your value proposition, key differentiators, and target audience. The more specific your descriptions, the more accurate the analysis.
• 3–5 competitors — brands you're most often compared against. For each competitor, it helps to know alternate spellings and variations.
• 5–10 questions real users ask AI models — for example, "What's the best tool for [task]?", "Compare [your brand] and [competitor]", "Recommend a solution for [need]".

On your first top-up you'll receive a welcome bonus of 1,000 AI credits — more than enough to run several evaluations and explore the reports.

Step 1. Sign up and log in to BrandWise

How to start with BrandWise:

1. Go to brand-wise.ru and click "Get Started".
2. Enter your email address.
3. A 6-digit access code will arrive in your inbox — enter it on screen.
4. If no account exists for that email, one is created automatically on first login. There's no separate sign-up form — login and registration are a single step.

Alternative login method: email + password. A password is set when accepting a project invitation or in account settings.

After logging in you'll land on the "Projects" page — this is where your LLM brand monitoring setup begins.

Step 2. Create a project

A project is a workspace that brings together brands, scenarios, reports, team members, and billing. To create a BrandWise project:

1. On the "Projects" page, click "Create project".
2. Enter a name — for example, your brand name or client name.
3. Click "Create".

What happens automatically:

• You are assigned the Admin role — full access to settings, billing, and team management.
• A billing link is bootstrapped — the project is immediately ready for payment setup.

Tip for agencies: create a separate project for each client. This ensures data isolation, independent billing, and the ability to invite client team members with restricted permissions.

Step 3. Set up your brand profile

The brand profile is the foundation for analysis. All report metrics are built on this data. The more detailed the profile, the more accurately BrandWise evaluates how AI models represent your brand.

Go to "Brands" in the project sidebar and click "Add brand".

Primary name

The canonical spelling of your brand as used in official communications.

Example: BrandWise

Name variations

Alternative spellings — transliterations, abbreviations, names in other languages. At least 1 variation is required.

The system searches for all specified variations in model responses. If your brand is mentioned under a different spelling that you haven't listed — the mention won't be counted.

Examples: Brand Wise, brandwise.ru, Брендвайз

Positioning (10–2,000 characters)

Describe your brand's value proposition: how you differ from competitors, what problem you solve, who your audience is.

Positioning directly affects the Positioning Match metric — the system compares what's written here with how the model describes your brand. Specifics matter more than abstractions.

Tone of voice (10–1,000 characters)

Describe your brand's communication style. Used to evaluate the tone of model responses.

Examples:

• "Expert, calm, with specific numbers and examples"
• "Friendly, informal, with humor"

Desired attributes

Key characteristics your brand wants to be associated with. Each attribute gets a weight from 1 to 5:

Examples:

• "Transparent analytics" — weight 5
• "Cross-model comparison" — weight 4
• "Ease of use" — weight 3

Attributes directly affect the Positioning Match formula: the system checks whether each attribute is supported, contradicted, or unmentioned in the model's response. The higher the weight, the stronger the impact on the final score.

Competitors

List 3–5 main competitors. For each, provide a name and at least one naming variation.

Competitors feed into the Top of Mind metric (who the model recalls first) and Consideration (whether your brand makes the shortlist), as well as the competitive analysis section of the report.

Example:

Learn more about the brand profile structure in What is BrandWise.

Step 4. Create a scenario and choose AI models

A scenario is a set of questions (intents) that BrandWise sends to selected AI models on behalf of real users. The responses are analyzed to produce an AI brand visibility report.

Go to "Scenarios" and click "Create scenario".

Basic settings

1. Name — a descriptive title, for example "Brand position in CRM queries".
2. Brand — select the profile created in step 3.
3. Mode — for your first time, choose Basic.

Intents (1–20)

Intents are the questions sent to models. Each intent is 10–1,000 characters long. Guidelines for writing effective intents:

Phrase them as real user queries. Not "Tell me about BrandWise", but "What's the best tool for monitoring my brand in ChatGPT?"

Mix question types:

• General: "What's the best CRM for small business?"
• Comparative: "Compare [brand] and [competitor] for [task]"
• Recommendation: "Recommend a service for [need]"
• Contextual: "I need a [solution] with a budget under [amount], what do you suggest?"

3–5 intents is enough for your first evaluation. Start with the core questions your customers ask most frequently.

Choosing models

Select at least 1 model, but for comprehensive analysis we recommend 3–5 popular ones:

• OpenAI: GPT-4o, o3-mini
• Anthropic: Claude
• Google: Gemini
• Yandex: YandexGPT
• Others: DeepSeek, GigaChat, Mistral

Models are grouped by vendor in the interface — it's easy to select several from one provider or one from each for breadth.

Step 5. Connect payment and top up your balance

Before running a brand evaluation in AI you need to top up your AI credits balance. Go to the "Billing" section of your project.

What are AI credits

AI credits are BrandWise's internal currency. They are consumed when processing each request to an AI model. Cost depends on the model and response volume (number of tokens).

Payment methods

Topping up

Preset amounts: 1,000 / 2,500 / 5,000 RUB or a custom amount from 100 to 100,000 RUB.

Welcome bonus: on your first top-up you receive 1,000 AI credits free. The bonus is one-time — per project and per user.

Minimum balance to run: 100 AI credits. If your balance is lower, the system won't allow you to start a scenario.

Tip: for a first evaluation with 3 intents and 3 models, the minimum top-up is sufficient. AI credits are consumed efficiently — a typical request costs single-digit credits.

Step 6. Run your first brand evaluation in AI

Everything is ready. Return to the scenario page and click "Run". If you haven't saved the scenario yet, use the "Create and run" button.

What happens after launch

The platform sends each intent to every selected model, collects responses, and analyzes them across 6 metrics. Track progress through these statuses:

Execution time: from a few seconds to several minutes — depending on the number of intents, models, and their response speed. For 3 intents and 3 models, typical time is 1–3 minutes.

Email notification: when the evaluation completes, you'll receive an email with a link to the report.

If the evaluation stops

A "Stopped by billing limit" status means the run exceeded the project's AI credit limit. What to do:

1. Go to "Billing" → "Limits".
2. Increase the per-run credit limit or disable it entirely.
3. Return to the scenario and click "Continue" — the evaluation picks up where it left off. Already-processed items are not recalculated or re-charged.

Step 7. How to read your brand visibility report

Once the evaluation completes, open the report — available from the scenario page or the "Reports" section of your project.

Overall Score

An aggregate brand score on a 0–100 scale, calculated as the weighted average of 6 metrics.

Hero cards

Key indicators at a glance:

• Confidence — how confidently the model recommends your brand
• Mention Rate — share of responses that mention your brand
• Top-1 / Top-3 Rate — how often your brand ranks first or in the top three
• Avg Position — average brand position in recommendation lists

Model breakdown table

A heatmap showing 6 metrics for each model. Color coding makes strengths and weaknesses instantly visible:

• Green cells — high scores, brand is well represented
• Yellow cells — moderate scores, room for improvement
• Red cells — low scores, the model doesn't know the brand well or describes it inaccurately

Compare results across models: if your brand scores well in Claude but poorly in GPT-4o, that signals your content strategy may need adapting for different AI training pipelines.

Dialogues tab

Full transcripts of each model's response to each intent. For every dialogue you get:

• Prompt — the sent intent
• Model response — full text
• Analysis — metric scores with evidence quotes (verbatim excerpts from the response that justify each score)
• Sources — if the model referenced specific resources

Dialogues are the raw material for understanding why a model described your brand the way it did. Use them to spot patterns: what phrasing the model uses, what it gets wrong, what it doesn't know.

Competitors tab

A Competitor × Model matrix with positions and key metrics. Shows:

• Who models recommend instead of you
• Where each competitor ranks
• What competitor advantages models highlight

Learn more about each metric in What is BrandWise: 6 metrics for evaluating brand presence in AI responses.

What's next

Your first report is a starting point. Here's how to use BrandWise for systematic AI brand visibility management:

• Re-run the same scenario in a week — compare results and track trends. AI models are regularly updated, and brand positions can shift.
• Try History mode with a persona — multi-turn conversations reveal how models describe your brand when users ask follow-up questions, which is closer to real user behavior.
• Invite team members — add colleagues with the "Admin" role (full access) or "Member" role (run evaluations and view reports). Go to "Project Users" in the sidebar.
• Create a custom report — combine metrics, filters, and groupings into a reusable view for regular monitoring or executive reporting.
• Expand your intent set — add questions covering different topics, audiences, and use cases. Maximum 20 intents per scenario.

Create an account and get started