SYNTHETIC RESPONDENTS
Alice: offers synthetic audience pretests for concepts, messages, interfaces and research questionnaires.
PRODUCT PROFILE
ADS Alice documents a synthetic-respondent service for pretesting research materials before fieldwork. The supplier presents the service as a way to identify risks and improve the setup for a later real-audience study, rather than as a replacement for that fieldwork.
Information from the company's product materials.
Alice models reactions from a specified synthetic target audience to research and marketing materials.
The documented use cases include hypotheses, product concepts, messages, UX and customer journeys, surveys and interview guides.
The supplier explicitly describes Alice as a preliminary tool intended to make subsequent research with real respondents more focused.
The English-language page lists a EUR 120 single pretest and a 20-pretest team package available by supplier contact. Current payment and package terms should be confirmed with ADS.
Methods described by the company.
The documented process begins with an uploaded hypothesis, concept, advertisement, survey, guide or customer scenario, together with the research objective and evaluation criteria.
ADS Alice says it models reactions from a specified synthetic target audience to assess clarity, interest, trust, barriers, user reactions and question quality.
The supplier documents output in the form of strengths, risks, questions to validate and recommendations for the subsequent real study.
Data and settings the company says teams can choose.
The supplier asks users to provide the material under review, define a target audience and explain the desired learning objective. Those inputs define the documented pretest context.
ADS Alice describes the service as being built on real respondent samples and ADS's respondent-recruitment experience. The public page does not provide a technical account of the training data, sampling design or model construction.
Results reported by the company, with details about how they were tested.
The supplier's page describes the pretest workflow, examples of materials it can assess and a case-study link. It does not state a public numerical accuracy result, hold-out comparison or full validation design for Alice's synthetic responses.
The public material describes an early-stage pretest service. It does not report response accuracy or decision performance.
ADS Alice explicitly says the service does not replace real respondents and that important final decisions should still be validated with real-audience research.
The provider frames Alice's outputs as risks, strengths and questions for validation before recruitment. That is a narrower role than a final fieldwork result or a claim about a population.
Uses groups of AI personas for simulated focus groups and structured discussions, drawing on Microsoft's TinyTroupe framework.
Runs synthetic user studies for product questions, concept feedback and message testing, returning a research report from the interviews.
A managed research service covering synthetic surveys, depth interviews and focus groups, with human analysis and a tailored findings report.
