--- title: "SINUS Milieu-Minds pass reality test against SINUS… | Minds" canonical_url: "https://getminds.ai/press/sinus-milieu-minds-reality-test" last_updated: 2026-10-22 meta: description: "In a comparison with a SINUS survey of 2,005 people, synthetic SINUS Milieu-Minds reproduced answer distributions per milieu with 91.7% closeness. SINUS and..." "og:description": "In a comparison with a SINUS survey of 2,005 people, synthetic SINUS Milieu-Minds reproduced answer distributions per milieu with 91.7% closeness. SINUS and..." "og:title": "SINUS Milieu-Minds pass reality test against SINUS… | Minds" "twitter:description": "In a comparison with a SINUS survey of 2,005 people, synthetic SINUS Milieu-Minds reproduced answer distributions per milieu with 91.7% closeness. SINUS and..." "twitter:title": "SINUS Milieu-Minds pass reality test against SINUS… | Minds" --- Minds October 22, 2026·Press release # **SINUS Milieu-Minds pass reality test against SINUS survey** In a comparison with a SINUS survey of 2,005 people, synthetic SINUS Milieu-Minds reproduced answer distributions per milieu with 91.7% closeness. SINUS and Minds position them as a fast pre-check before fieldwork.![Minds and sinus logos on the Minds yellow-green gradient background](https://getminds.ai/images/newsroom/minds-sinus-cover.webp)_Synthetic respondents from all ten Sinus-Milieus come close to real survey results and are intended as a pre-check before fieldwork, not as its replacement._ HEIDELBERG / BERLIN, October 22, 2026. The SINUS Milieu-Minds, the AI personas for the Sinus-Milieus developed by SINUS-Institut and the Berlin start-up Minds, have been compared with a real SINUS survey. 200 Milieu-Minds, 20 from each of the ten Sinus-Milieus, answered 24 questions on personal concerns, from very great concern to no concern at all. SINUS had asked the same questions to 2,005 people in June 2026. The Minds did not see the survey results. Per milieu, the answer distributions of the Minds reached a closeness of 91.7 percent to the survey. On average, they differed by 8.3 percentage points per answer level. As a reference point, two real SINUS survey waves with the same questionnaire, from March and June 2026, reached 96.3 percent against each other. Across all milieus, with the Minds weighted by the milieu composition of the survey, closeness was 95.2 percent. All 4,800 answers were valid and were available after about 26 minutes. ## Clear gap to general AI models A separate method control asked general AI models the same questions with a simple instruction to answer as a member of a named milieu. Gemini 3.7 Flash and Claude Sonnet 5 both reached about 72 percent, around 20 points less than the Milieu-Minds. Asked the same question 20 times, the models usually gave the same answer 20 times: Gemini in 194 and Claude in 211 of 240 combinations of question and milieu. The Milieu-Minds never answered unanimously. Adding the official milieu portraits from the SINUS website did not improve Gemini's result either. A follow-up analysis of the same data looked at the differences between milieus. Where two milieus differ clearly in the survey, the Minds ranked them correctly in 86.5 percent of cases, the general AI model with only the milieu name in 67.4 percent, and the earlier real survey wave in 97.1 percent. On exact shares per milieu, this lead does not hold: a fixed answer mix from the March survey was slightly more accurate. The Minds are strongest at ranking topics and milieus. ## A pre-check before fieldwork SINUS and Minds position the Milieu-Minds as a fast pre-check, not a replacement for primary research. The results give indications of topics and answer distributions within a milieu; individual shares can deviate clearly. For a realistic estimate of actual shares, a survey is still needed. The results are not nationally representative and are not projected to the population. Typical uses are pre-selecting campaigns and messages for each milieu before a pretest, sharpening hypotheses and questionnaires before fieldwork starts, and understanding reasons, because the Minds explain their answers. Whether real respondents give the same reasons is shown by fieldwork. The reality test covered attitude questions; for campaign decisions, SINUS and Minds recommend using the Minds for the pre-selection and a field study for the decision. The SINUS Milieu-Minds are available to SINUS clients through SINUS and the Minds Marketplace. ## About the comparison 200 SINUS Milieu-Minds with 4,800 answers to 24 concern questions, run on August 25, 2026, compared with SINUS surveys from June 2026 (n = 2,005, primary comparison) and March 2026 (n = 1,047). Closeness is 100 minus the mean absolute difference in percentage points between synthetic and real answer shares. Values per milieu are simple means across the ten milieus. Tested AI models: Gemini 3.7 Flash and Claude Sonnet 5, runs in August and September 2026. Details on method and limitations are available on request. ## About Minds Minds (getminds.ai) is the end-to-end platform for commercial synthetic research, founded in Berlin. Teams use Minds to build synthetic audiences and expert panels and to run qualitative and quantitative studies, from interviews and questionnaires to methods such as MaxDiff, before committing time and budget to fieldwork or production. Minds sees its technology as a complement to research with real people: it makes assumptions visible and improves the quality of the research that follows. More information about Minds at [www.getminds.ai](https://www.getminds.ai/). ## Media Contact Alexander Doudkin Co-Founder & CEO, Minds [alexander@getminds.ai](https://getminds.ai/mailto:alexander@getminds.ai) Jan Hecht Director Research & Consulting, SINUS-Institut [jan.hecht@sinus-institut.de](https://getminds.ai/mailto:jan.hecht@sinus-institut.de) [Minds](https://getminds.ai/)© 2026 Minds. Your target audience. AI-driven and grounded in transparent evidence. Build within minutes. [Minds on X (Twitter)](https://x.com/mindsai_co) [Minds on LinkedIn](https://www.linkedin.com/company/mindsaicompany/) [Minds on Instagram](https://www.instagram.com/getminds.ai/)Minds is part of [![ESOMAR](https://getminds.ai/images/newsroom/logos/esomar-logo.svg)ESOMAR](https://esomar.org/) [![GreenBook](https://getminds.ai/images/newsroom/logos/greenbook.svg)GreenBook Directory](https://greenbook.org/company/Minds) [![Insight Platforms](https://getminds.ai/images/newsroom/logos/insight-platforms.png)Insight Platforms](https://www.insightplatforms.com/platforms/minds/) [![Capterra](https://getminds.ai/images/newsroom/logos/capterra.svg)Capterra](https://www.capterra.com/p/10046203/Minds/) [![G2](https://getminds.ai/images/newsroom/logos/g2.svg)G2](https://www.g2.com/products/minds/reviews) [![CSSDA Best UX Design Award](https://getminds.ai/images/newsroom/logos/cssda-best-ux-award.png)CSSDA Best UX Design Award](https://www.cssdesignawards.com/) [![CSSDA Best Innovation Award](https://getminds.ai/images/newsroom/logos/cssda-best-innovation-award.png)CSSDA Best Innovation Award](https://www.cssdesignawards.com/) [![CSSDA Best UI Design Award](https://getminds.ai/images/newsroom/logos/cssda-best-ui-award.png)CSSDA Best UI Design Award](https://www.cssdesignawards.com/)