# Artixol > Artixol builds a living audience of AI personas that run surveys, message tests, shocks and interviews, and remember what they said. Study the same people again and again. Artixol output is experimental model output for exploration. It is not a census, a forecast or a substitute for field research. ## Product - [Product overview](https://artixol.com/product): the living audience platform - [How it works](https://artixol.com/how-it-works): brief, staff, study, keep - [Features](https://artixol.com/features): every shipped feature - [Methodology](https://artixol.com/methodology): how personas are built, known limits, and when to trust a result - [Pricing](https://artixol.com/pricing): plans - [FAQ](https://artixol.com/faq): common questions ## Use cases - [Message testing](https://artixol.com/use-cases/message-testing): Paste a slogan, ad or policy line. See who would use it, share it, ignore it or push back. - [User research surveys](https://artixol.com/use-cases/user-research): Ask everyone the same questions, then read the table and the splits by role. - [Concept validation](https://artixol.com/use-cases/concept-validation): Put an idea in front of the right people early and find out what they would do with it. - [Policy and shock testing](https://artixol.com/use-cases/policy-shock-testing): Inject a price hike, an outage or a new rule and see who is hit and who is unmoved. - [Stakeholder interviews](https://artixol.com/use-cases/stakeholder-interviews): Open a one-to-one interview with a single persona and ask the follow-ups a survey cannot. - [Rumor and misinformation testing](https://artixol.com/use-cases/rumor-testing): Start a street story and see who actually hears it, from whom, and who still believes it after a correction. - [A/B testing on the same people](https://artixol.com/use-cases/ab-testing): Run two treatments on clones of the same named people so the comparison is clean. ## Solutions - [Founders](https://artixol.com/solutions/founders): Pressure-test your pitch, pricing and positioning before you have a customer list. - [Marketers](https://artixol.com/solutions/marketers): Compare headlines, offers and campaign lines before they get media budget. - [Agencies](https://artixol.com/solutions/agencies): Bring evidence to the pitch and the creative review without adding a research line item. - [Product teams](https://artixol.com/solutions/product-teams): Validate concepts, onboarding and pricing changes before they take a sprint. - [Researchers](https://artixol.com/solutions/researchers): Run survey, interview, intervention and experiment on the same population, with the method visible. ## Comparisons - [Artixol vs. survey panels](https://artixol.com/compare/artixol-vs-survey-panels): Panels give you observed answers from real people. Artixol gives you fast exploratory answers from a persistent synthetic audience. They answer different questions. - [Artixol vs. using ChatGPT for market research](https://artixol.com/compare/artixol-vs-chatgpt): You can ask a general assistant to pretend to be your customers. It will answer. Here is what you lose compared with a purpose-built audience. - [Artixol vs. focus groups and interviews](https://artixol.com/compare/artixol-vs-focus-groups): Real conversations reveal things no model will. Artixol helps you arrive at them better prepared, and run the cheap versions in between. ## Glossary - [Synthetic audience](https://artixol.com/glossary/synthetic-audience): A group of simulated people, generated by AI from a brief, used to study how a real audience might respond. - [Synthetic research](https://artixol.com/glossary/synthetic-research): Research run on simulated respondents rather than recruited humans, used to explore questions quickly and cheaply. - [Persona](https://artixol.com/glossary/persona): One simulated person in an audience, with a role, personality, background, languages, attitudes and memories. - [Persona memory](https://artixol.com/glossary/persona-memory): What a persona remembers from earlier studies: episodic memories and topic attitudes that carry into later studies. - [Message testing](https://artixol.com/glossary/message-testing): Showing a slogan, ad or policy line to an audience to learn who would use it, share it, ignore it or doubt it. - [Concept testing](https://artixol.com/glossary/concept-testing): Evaluating a product or feature idea with the intended audience before building it. - [A/B testing](https://artixol.com/glossary/ab-testing): Comparing two versions of something, a control and a treatment, to see which performs better. - [Shock (simulation)](https://artixol.com/glossary/shock): A world event, such as a price hike or outage, injected into an audience's memory to see who is affected. - [Rumor simulation](https://artixol.com/glossary/rumor-simulation): Modeling how a story spreads through family, peer and colleague ties and whether a correction interrupts it. - [Answer engine optimization (AEO)](https://artixol.com/glossary/answer-engine-optimization): Structuring content so AI assistants and search answer boxes can quote it accurately. ## Blog - [What is a synthetic audience, and when to use one](https://artixol.com/blog/what-is-a-synthetic-audience): A synthetic audience is a set of AI-generated personas built from a brief. It is useful for fast, cheap exploration before a human study, and a poor substitute for a representative sample. - [Synthetic research vs. surveys: when to use each](https://artixol.com/blog/synthetic-research-vs-surveys): Use synthetic research first to narrow options and sharpen questions, then field a human survey for the decisions that need real evidence. - [How to test a headline with AI personas, step by step](https://artixol.com/blog/how-to-test-a-headline-with-ai-personas): Staff the audience you are writing for, run each headline as a Message test, read the split and the reasons, then re-run in Lab on clones of the same people for a fair A/B. - [Why synthetic personas need memory](https://artixol.com/blog/why-personas-need-memory): Memory lets the same personas carry what they said and what happened to them into the next study, so you can measure change instead of re-rolling new people each time. - [How we think about accuracy in synthetic research](https://artixol.com/blog/how-we-think-about-accuracy): Accuracy depends on the question. We describe outputs as experimental, publish limits openly and recommend treating strong signals as a reason to run a human test. ## Optional - [About](https://artixol.com/about) - [Security](https://artixol.com/security) - [Changelog](https://artixol.com/changelog) - [Contact](https://artixol.com/contact)