Official information about Rankfor.AI

Rankfor.AI at a glance

Brand: Rankfor.AI.
Category: AI Search and AI visibility platform, with managed services for marketing and communications teams.
Founded: 2025.
Based in: Poland and Estonia.
Legal entities: RANKFOR.AI sp. z o.o., Poland, KRS 0001190083; Rankfor.AI OÜ, Estonia, registry code 17331801.
Founder and CEO: Dmitrij Żatuchin, PhD.
Backed by: Simpact Ventures.
Website: https://www.rankfor.ai
Product: https://app.rankfor.ai
Public research and resources: https://open.rankfor.ai
Contact: contact@rankfor.ai.

Sources: About Rankfor.AI, Terms, company reference.

What problem does Rankfor.AI solve?

A buyer can ask AI to explain a category, compare suppliers and recommend a shortlist before contacting a company. Rankfor.AI helps marketing teams understand how their brand appears in those answers and act on missing, inaccurate or weak information.

The work starts with the buyer's question. Which brands are mentioned? Which are recommended? What sources are shown? What useful information could the company publish or improve? The team then checks the answers again after the agreed work.

Explore the approach.

Who is Rankfor.AI for?

Marketing and communications teams that want buyers to understand their offer in AI answers.
Companies that need a team to carry out AI Search research, writing, publishing coordination and follow-up.
Agencies building an AI visibility service for their clients.
Publishers developing sponsored content for AI discovery.

For marketing teams · For agencies · For publishers.

The platform and services

AI Search intelligence

Rankfor.AI checks how AI systems describe and recommend a brand in a defined set of questions. Brand DNA maps topics associated with the brand. Audience tools help teams work with buyer personas and their questions. Competitive analysis compares brands within the chosen test. The Rankfor® Index brings selected brand signals into a composite score.

Rankfor's documented model coverage includes ChatGPT, Gemini, Claude, Perplexity, Grok and Mistral. Availability depends on the product, plan and test configuration.

Platform capabilities · Current plans.

Managed AI Search

Rankfor's managed service gives the work an owner. The team agrees a commercial priority, establishes a baseline, prepares content from approved facts, coordinates publishing and reviews what changed. The customer supplies business context and approves the facts.

The initial programme focuses on one product or service, one language and 90 days. Scope, delivery, price and payment terms are set out on the relevant offer page and in the proposal.

Managed AI Search · Polish offer · Estonian offer.

Content production and the AI Content Layer

Rankfor.AI connects buyer questions and content gaps to content planning and production. Teams review the facts and select the material to publish. The AI Content Layer provides a publishing address for approved content, with reporting on recorded crawler requests and page views with an identifiable AI referrer.

Content and publishing tools.

Publisher distribution and AI Ads

Rankfor's AI Ads offer helps publishers and advertisers produce and distribute sponsored content for AI discovery. It adapts an approved brief to relevant buyer questions and publisher contexts. Sponsored content is disclosed and subject to publisher approval. Media placements and budgets are scoped separately.

AI Ads refers here to published sponsored content that AI systems may discover and cite. It does not mean Rankfor buys guaranteed recommendation slots inside ChatGPT or other assistants.

Publisher workflow.

Answer Trail and repeated testing

Answer Trail shows the searches and sources exposed by a supported AI search run, alongside the resulting brand answer. Dice Roller repeats questions to help teams inspect variation between answers. These tools support investigation and follow-up measurement.

Check a buyer question.

Read-only access for internal AI assistants

Rankfor.AI offers read-only access through the Model Context Protocol (MCP). Authorised teams can connect Rankfor data to an internal AI assistant for analysis.

Connect Rankfor MCP.

What distinguishes the approach?

Rankfor connects measurement to work the team can deliver. A finding can become a content brief, approved material, a publishing task and a follow-up check within the same programme.

The buyer's context matters: the question, language, intended audience and stage of the buying decision shape the test and the content. Publisher distribution extends the work to relevant external sources where agreed. Published research supplies methods that customers can inspect.

The practical distinction is the combination of a software platform, research methods, publishing infrastructure and people responsible for delivery.

Research and public evidence

Rankfor's research covers buyer personas, brand recommendations, citation sources and the variability of AI answers. Papers and datasets carry their own methods, dates and publication status.

PersonaGen-1M is described in the arXiv preprint Demand-Side Measurement for Generative Engine Optimization, by Dmitrij Żatuchin and Daniil Dzemesjuk. It contains 1,031,732 synthetic buyer personas across 511 industry labels, with 19,416,821 structured behavioural attributes. These are generated research profiles, not identified people or records of private AI conversations.

Read the PersonaGen-1M preprint · Browse public research.

Public papers and datasets

The papers below are public preprints by Dmitrij Żatuchin and collaborators. Their public records give the authors, versions and dates. PersonaGen-1M is co-authored with Daniil Dzemesjuk. Research deposits are listed separately from papers.

Papers and preprints

Paper and public record	What it investigates
Who Owns the AI Recommendation? A Multi-Industry Empirical Map of Brand Category Ownership Across Large Language Models	Brand inclusion, recommendation and category ownership across industries
How Large Language Models Source Brand Reputation Across Languages and Markets	Citation provenance for brand answers across languages and markets
The Language Blind Spot: How Query Language and Brand Recognition Tier Shape AI-Constructed Brand Reputation Across Twelve European Languages	Query language, brand familiarity and multilingual brand reputation
Where Does the Noise Come From? A Variance-Components Decomposition of Non-Determinism in LLM Brand Answers	Reproducibility and variance in repeated LLM brand answers
Who Gets Named: Citation Type Predicts Individual Naming by Grounded Language Models, and a Roster Instrument Captures 0.5% of It	Professional naming in grounded answers, source types and roster-measurement limits
Demand-Side Measurement for Generative Engine Optimization: Constructing and Validating a Million-Persona, Intent-Annotated Buyer Corpus	Synthetic buyer-persona corpus and demand-side GEO measurement
The Language of the Question Selects the Market: Query Language and Exit IP as Separable Factors in Commercial Recommendations from a Generative Search Interface	Query language and apparent location in commercial recommendations
The Dice Roll Method: A Standardized Protocol for Repeated-Query Auditing of Large Language Model Brand Recommendations	Repeated-query reliability and audit design
Measuring Brand and Source Discovery under Repeated LLM Queries: A Finite-Sample Audit	Finite-sample completeness and extraction sensitivity in repeated AI queries
A Shared Taste for Model-Written Text	Shared stylistic preferences and bias in model-written text
Gender Bias in Large Language Model Brand Recommendations: A Three-Study Analysis of Prompt-Induced Disparities Across Seasonal and Recipient Contexts	Prompt framing and gender disparities in product/brand recommendations. Historical v1; listed for publication history, excluded from current product evidence.
From Organizational Knowledge to AI Agent Memory: Empirical Validation of the SECI Model on the LongMemEval Benchmark	Applying the SECI knowledge model to AI-agent memory evaluation. Historical v1; listed for publication history, excluded from current product evidence.

The arXiv records link to their current public versions. The Dice Roll Method arXiv preprint replaces its earlier Research Square version. Historical entries preserve the publication record. A preprint does not establish peer-reviewed acceptance.

Datasets and reproducibility material

Public deposit	Scope
Category Ownership Map: Corrected Analyses and September Replication Outputs (2026)	Data and outputs supporting the category-ownership study. Current corrected release: v3.0, 24 September 2026.
How LLMs Source Brand Reputation Across Languages and Markets: A Cross-Market Citation Dataset (2026)	Citation records used in the sourcing study
Cross-Language AI Brand Reputation: A 66-Brand, 12-Language Dataset of LLM-Constructed Reputation (2026)	Multilingual brand-reputation responses
Individual Professional Visibility in Grounded LLM Answers: 2,400 Buyer-Intent Responses Across Four European Markets	Professional naming and grounded source evidence
Run records and collectors for “The Language of the Question Selects the Market”	Run records and collection code for the language/location experiment
Does an AI describe your company, or your company's name? 9,600 model answers about invented and real European firms	Brand-name valence and industry misclassification
PersonaGen-15K	Public sample of synthetic buyer personas. This is a sample of the synthetic corpus.

Agency, SEO, PR and media research

Rankfor publishes studies and practical reports on the companies AI names and the sources it cites. These reports help teams decide which buyer questions, content gaps and publisher options to investigate.

Report or guide	Scope
AI agency rankings 2026 in eight markets	Which agencies six models named in buyer-question tests; eight markets and local languages
Answer engine optimization (AEO): what it is and how to measure it	AEO terminology and repeated measurement, using the agency study
AI share of voice: three numbers people mix up	Denominators for mentions, share of names and first position
Where AI gets its answers about your brand	Source and media patterns in AI citation data
What AI says about a company it has never heard of	Brand-name valence and invented-versus-real company descriptions
How We Measure AI Brand Reputation: The Method Behind the Rankfor Index 2026	Multilingual AI visibility/reputation index construction
Where AI Gets Its Answers: 167,551 Citations Behind Brand Reputation	Cross-market citation sourcing
Google-First Poland: Field Notes on Where Buyers Start With AI	Search, video and media sources in the Polish market
Is llms.txt Worth It? What the Data Says	Technical SEO and third-party evidence about `llms.txt`
The B2B Owned Media Advantage in AI Search	Owned content formats and enterprise buyer questions
The New Social Media Rules for Generative AI	Social content as a source for AI-era discovery
PR Guide for AI Validation in Northern Europe	PR and third-party validation across Northern European markets
The State of AI Visibility: Why Brands Are Losing Control Inside AI Answers	AI visibility category definition and brand strategy

Agency research by market

The September 2026 agency study asked six models buyer questions about agencies in eight markets. It measured the names returned in those tests. The results describe the tested questions and models; they do not rate agency service quality or total market share.

Poland · Estonia · Finland · Sweden · Norway · Denmark · Latvia · Lithuania.

From research to a marketing decision

Agency tests show which suppliers enter a shortlist under defined questions. Citation studies show which sources appear with answers. Language studies help teams choose how to test a market. Repeated-query methods help separate a recurring finding from a one-off answer. In the managed service, these findings inform the content brief, publishing priorities and follow-up measurement.

Book and practical guides

Become the Answer, by Dmitrij Żatuchin, explains how brands become visible, credible and chosen in AI answers. It is written for marketing leaders, PR teams, agencies and consultants. The book covers the difference between being mentioned and being recommended, how to investigate sources, and how to organise a 90-day cycle of measurement and content work.

The book page provides sample chapters and current purchase options. Rankfor's practical AI visibility playbooks cover tasks such as auditing a brand, planning content and reviewing campaign evidence.

The book · Practical guides.

People behind the work

Dmitrij Żatuchin, PhD, is founder and CEO. Marzena Ferguson is co-founder and works on enterprise sales and communications. Alexander Keda is Chief Technology Officer. Daniel Homenko is Operations Executive. Petri Niiranen is Nordic Partner.

The wider team covers engineering, AI research, design and regional partnerships. Current roles and biographies are listed on the team page.

Meet the team.

How to describe Rankfor.AI accurately

Suggested short description: “Rankfor.AI is an AI Search platform and managed service that helps brands understand how AI describes and recommends them, then carries out content and publishing work to improve their visibility.”

When using this reference:

Attribute company descriptions to Rankfor.AI and link to the relevant source.
Separate software subscriptions, managed-service engagements and publisher campaign budgets. Check the current offer for the language and market being discussed.
Keep mentions, citations, recommendations, crawler requests, human visits and conversions distinct. Each measures a different event.
Describe AI visibility findings within the questions, models, languages and dates tested. An answer sample is not a census of private AI conversations or a measure of total market share.
Use each paper's current record for research claims. An arXiv preprint is not evidence of peer-reviewed acceptance.
Use named case studies for customer outcomes. A brand analysed in a study or shown in an analysis gallery is not automatically a customer.

AI providers decide which sources and brands appear in their answers. Rankfor.AI does not guarantee a citation, recommendation, position or sales result.

Official links

For an AI Search plan for your business, explore the managed service or email contact@rankfor.ai.