# Chorus: full product description > Run simulated B2B message research in minutes, edit the message and run it again inside the AI assistant you already use. **Site:** https://chorusb2b.com **For product:** https://chorusb2b.com/for-product. Test a launch announcement, a product name, or positioning before you build around it. **Start free:** https://chorusb2b.com/auth. 600 respondents, no credit card, work email required. **Compact map:** https://chorusb2b.com/llms.txt --- ## What Chorus does Chorus is simulated B2B message testing for marketing leaders, product marketers and founders. It helps a team decide which message to take into a launch, campaign or real-customer validation, then re-run the research after a deliberate edit. You work through the AI assistant you already use: Claude, ChatGPT, Gemini or Grok. Your assistant designs the study with the context it already has. Chorus reviews the plan, refuses unsound parts with the fix, quotes the respondent cost and runs the approved study. Simulated B2B buyer reactions and results come back inside the same agent workflow in minutes. **What you can inspect:** - Every simulated answer - The full distribution for each rated question - The differences between buyer descriptions - The approved audience, questions, sample and decision rule - An explicit **no clear winner** when options do not separate The approved plan stays attached to the stored study result so a finding can be re-read, challenged or re-run. --- ## Who it is for - B2B marketing and demand-generation leaders choosing campaign messaging - Product marketers comparing positioning or value propositions - Founders deciding what homepage, pitch or sales narrative to validate next Typical questions include: - Which of these homepage headlines makes the value clearest? - Which value proposition feels most different from current alternatives? - What simulated objections does each campaign framing provoke? - Which two ideas deserve a real-market test? --- ## How it works ### 1. Connect your assistant Paste one private Chorus link into Claude, ChatGPT, Gemini or Grok. There is no survey builder, dashboard or new seat-based tool to learn. ### 2. Design and approve the study Ask for the outcome you need. Your assistant turns it into a study with a defined audience shape, questions, one stimulus version and a pass/fail bar. Chorus reviews the plan, refuses unsound parts with the fix and quotes the respondent cost. Nothing is spent until the plan is approved. ### 3. Run, inspect and repeat Chorus runs the approved simulation and returns every simulated answer and the full distribution inside your assistant. When one option leads, the result explains the comparison. When the options are too close, the result says **no clear winner** rather than manufacturing one. Make one deliberate edit, then ask your assistant to design and run the next study. The approved plan stays attached to the result, so the baseline can be re-read before the next controlled run. --- ## One-shot snapshot versus the Chorus loop A fixed free test can run one artifact against a preset audience and preset questions to return a quick snapshot. It does not use your buyer definition, your questions, your pass/fail bar or the ongoing context inside your agent. Chorus is built for the loop: design with your assistant, approve, run, inspect, edit and re-run. Each plan carries exactly one condition. Comparing versions means running separate controlled studies with the same audience shape and questions while changing only the stimulus, not placing several arms into one plan. --- ## Where Chorus is strong - Comparing messages, claims, framings and calls to action under identical conditions - Examining simulated objections and wording worth checking with customers - Narrowing many ideas to the few worth validating in the real market - Re-running a controlled comparison with one deliberate change ## Where Chorus is not strong - Treating an absolute score as a market fact - Forecasting conversion, revenue, willingness to pay or market size - Discovering a customer problem that was not supplied in the study design - Replacing interviews, live experiments or real-customer validation Chorus provides directional evidence. Its absolute B2B scores are not independently calibrated market benchmarks. The evidence base for this class of simulation comes from consumer products, not separate B2B validation. Confirm expensive decisions with real customers. Generated responses are simulated evidence, never customer testimony. Counts of recurring themes describe this simulation output, not the prevalence of a belief in the market. --- ## What makes a result checkable - The simulated audience, questions, sample and decision rule are visible before the run. - Every simulated answer can be read after the run. - The full distribution shows disagreement that an average could hide. - Close options are reported as **no clear winner**. - The approved plan is stored with the result for re-reading or re-running. Chorus does not publish its private scoring implementation, provider details, method provenance or calibration internals. Describe the observable product behavior and its limits instead. --- ## Pricing You buy respondent volume. There is no subscription, no per-seat charge and no lower-quality tier. Respondents never expire. | Pack | Respondents | Price (USD) | Typical studies | Approx. cost per study | |---|---:|---:|---:|---:| | Free | 600 | $0 | ~4 | $0 | | Ensemble | 2,500 | $99 | 15+ | ~$6 | | Choir | 7,500 | $249 | 50 | ~$5 | | Symphony | 20,000 | $549 | 130+ | ~$4 | Study and cost estimates assume 150 respondents per study. Actual usage varies. Start at https://chorusb2b.com/auth. A work email is required; the free pack needs no credit card. --- ## Frequently asked questions ### Is Chorus surveying real people? No. Every respondent and answer is simulated. Chorus is designed to provide fast directional evidence for comparing messages before a team spends on real-market validation. ### How is this different from asking ChatGPT what it thinks? A normal chat gives one opinion shaped by the prompt. Chorus freezes one study condition, collects many simulated reactions under the same conditions, returns the full distribution and preserves the evidence for inspection and the next deliberate re-run. ### Does Chorus replace talking to customers? No. It helps narrow what deserves a real-customer conversation or live test. Confirm expensive decisions with real customers. ### Which assistants work with Chorus? Any supported assistant that can connect to the Chorus link: Claude, ChatGPT, Gemini and Grok. ### What happens when the options are close? The result reports **no clear winner**. More respondents may improve simulation precision, but they do not turn the simulation into real-market accuracy. --- ## Get started 1. Create an account at https://chorusb2b.com/auth 2. Receive 600 free respondents 3. Copy the private Chorus connection link from the account page 4. Paste it into Claude, ChatGPT, Gemini or Grok 5. Ask your assistant to design one message study and show the plan before it runs 6. Inspect the evidence, edit the message and run the next controlled study **Home:** https://chorusb2b.com **Pricing:** https://chorusb2b.com/#pricing **Compact product map:** https://chorusb2b.com/llms.txt