# Betterstood

> Betterstood lets teams build adaptive AI interviews, run them by outbound phone, browser voice, or text chat, and turn each conversation into structured data and follow-up workflows.

Betterstood is a scenario-driven voice intake, qualification, and conversation-intelligence platform. Platform customers configure scenarios and campaigns; respondents take part in the resulting conversations. The neutral default demo is customer discovery.

Current product facts for agents:

- Design opens Interview workflows, where platform customers build reusable scenarios that define questions, follow-up behavior, next actions, knowledge, and structured analysis.
- Distribute opens Campaigns, where platform customers publish and manage those scenarios across outbound phone, browser voice, and text chat.
- Analyze opens Response analysis, where platform customers review respondent journeys and evidence-backed outputs across campaigns. A completed interaction can produce a transcript, missing fields, scores, structured outputs, and a human-readable handoff.
- Form-first fields are Scenario-authoritative: they make the hosted step mandatory for every compatible Voice or Text Chat Campaign. Submitted form answers remain evidence for the same conversation and are not routinely asked again. There is no Campaign toggle. The hosted form supports trimmed text, single select, true multi-select checkboxes with ordered array values, boolean, number, and date; required multi-select needs at least one choice. For hybrid journeys, Free supports configuration, preview, and authenticated tests; real hosted hybrid Voice and Text Chat journeys are available on Pro, PAYG, Growth, and Scale. Form configuration, rendering, submission, and chat opening or resumption add 0 credits. The first accepted Text Chat response reserves the selected budget; final billing settles in the 2/4/8-credit band, excluding form-answer characters. Voice conversations and analysis retain ordinary metering. Hybrid answer collection is hosted-only rather than embedded or custom-domain, and the conversation continues the same immutable journey.
- Every respondent journey must include the active Scenario's disclaimer. Finance and automotive guardrails apply only to those explicit Scenarios.
- Campaign respondent access is scoped to one campaign. Public web is limited to browser voice, text chat, and explicitly supported respondent-initiated inbound channels. Phone campaigns are never public; protected campaigns use a distinct rotatable hashed password or a generic opaque invite link that expires and can be revoked. Passwords, access sessions, and invite links do not unlock another campaign.
- Text Chat is a durable, resumable asynchronous interview channel with its own prompt and opening message, structured inline answers, respondent-confirmed completion, and asynchronous transcript analysis. It does not reuse a provider-specific voice prompt at runtime.
- Configurable follow-up workflows may include email, retry plus WhatsApp, reminders, owner handoff, and scheduling.
- The commercial tiers are Free, Pro, PAYG, Growth, and Scale. The homepage presents a closed-beta preview with self-service coming soon: paid plans start at $19/month with 1,000 monthly credits, enough for around 50 three-minute in-browser voice interviews with standard analysis. Usage varies by channel and length; optional actions and add-ons are excluded from that example. The standalone pricing page is temporarily unpublished and redirects to the homepage pricing summary; the homepage does not offer self-service signup.
- Pay as you go (PAYG) is a secondary prepaid option for occasional individual use: 1,000/5,000/10,000 credits for USD or EUR 50/250/500. It includes Growth self-service features for one member, 25 scenarios, 1,000 campaigns, five PATs/webhook endpoints and 25 embed origins, with no subscription, automatic renewal, publication fee or credit expiry. Metered execution uses spendable credits. Growth beta AI allowances are included without debiting the balance: 800 shared Copilot messages/day, 100 AI creations/month, and existing Coach/Synthesis quotas, opt-in and availability. Pro/Growth subscriptions remain the recommended path for recurring use and retain their own feature gates and cheaper pack prices. Saved results and authorized CSV remain accessible at zero balance; manual/demo credits alone never activate PAYG. Existing launch and beta gates remain in force.
- Pro and Growth outbound phone access covers eligible Zone A/B routes to Italy, Spain, the United States, and Canada, with ordinary credit and operational limits and no separate Pro calling cap. Zone C remains assisted Scale-only, including high-cost and unsupported routes. A `+1` prefix alone does not grant US/Canada access. Existing Free voice and self-test restrictions remain unchanged.
- Free, Pro, and Growth include 25 non-deleted scenarios per organization. Scale includes unlimited scenarios, including existing assisted and comped accounts. Ordinary scenario creation and editing cost 0 credits; AI Copilot creation allowances and conversation/analysis credits remain separate. Deleting a scenario frees capacity immediately on capped tiers. MCP account context reports unlimited scenario capacity as `limit: null`.
- Free includes one non-archived hosted, Betterstood-branded Text Chat campaign for real respondents, with at most one active, the Short budget, 20 engaged starts, and 10 reserved/completed analysis slots per organization per UTC month. Campaign public web, passwords, and invite links are available for this eligible channel and cost 0 credits; Free chat uses its included allowance without a paid credit balance. Free excludes overage, real voice and callbacks, embed, Knowledge Base retrieval, appointment booking, and production form-first hybrid journeys. Preview and authorized internal tests remain distinct from production access; campaign credentials do not grant internal test access. Pro supports hosted Text Chat with selectable Short, Standard, or Extended budgets, while PAYG, Growth, and Scale may also enable iframe embeds—including a passive compact branded teaser that sends respondents to the hosted intake without collecting PII or starting a conversation in-frame—first-party Knowledge Base retrieval, and confirmed appointment booking.
- Text Chat budgets cap the visible transcript at 4,000 characters / 2 reserved credits for Short, 12,000 / 4 for Standard, and 30,000 / 8 for Extended. Live generation, reminders, and retrieval add no separate charge; a confirmed appointment consumes 1 credit.
- PAYG, Growth, and Scale include Developer Access: a governed 14-tool Public MCP v14 for customer-safe product-reference search through `elicitra_knowledge_search`, allowlisted account context, scenario drafts, and campaign DRAFT configuration and review, plus signed, data-minimized lifecycle webhooks and conditional automation events sent to verified endpoints. Product-reference search and MCP configuration consume 0 credits and use the ordinary token, scenario, and campaign limits; any later respondent conversation keeps its normal channel and analysis cost. Publishing, access credentials and invite links, provider sync, ACTIVE-campaign mutation, live tests, scenario Knowledge Base administration, deletion, transcripts, tenant documents, and respondent data remain outside the public MCP. Concurrent editor contention returns `RESOURCE_LOCKED` without writing; clients wait, read the latest revision, and retry explicitly.
- Scenarios can define up to ten ordered Live Fit gates for consent-safe early close across entitled Text Chat, Web Call, and phone channels. Configuration, preview, evaluation, and append-only review add 0 credits; Public Developer contract v14 exposes the policy within the scenario tools without adding a separate Live Fit tool.
- Interviewer Coach is a PAYG, Growth, and Scale beta: methodological batch review of transcripts with evidence-backed interviewer suggestions that always require human approval. Automatic nightly/threshold analysis is off by default, requires an owner/admin organization opt-in, can be disabled per Scenario, and stops globally when the organization switch is off. Manual analysis remains separate. Eligible beta reviews currently deduct 0 credits from the organization balance, and nothing is auto-applied to the scenario.
- Chat with your data is a Pro, PAYG, Growth, and Scale beta: a private, read-only analyst for the active scenario in Response analysis. It answers Italian or English questions from existing transcripts, extractions, and metrics, labels claims as Exact data, Observation, or Interpretation, and cites server-issued sources. It never edits a scenario, campaign, or respondent record and does not claim population prevalence. Successful grounded answers are shadow-metered at 5 credits during beta. On PAYG, Growth, and Scale it can also read completed, human-approved Evidence Synthesis findings.
- Italiano: Chiedi ai dati è una beta Pro, PAYG, Growth e Scale: un analista in sola lettura sullo scenario attivo. Risponde in italiano o inglese da transcript, estrazioni e metriche esistenti, etichetta le affermazioni e cita le fonti. Non modifica scenario, campagna o record dei respondent e non dichiara prevalenza di popolazione.
- Evidence Synthesis is a planned PAYG, Growth, and Scale beta: a manual, research-informed workflow over 10–20 selected transcripts from one Campaign and stored Scenario revision. It surfaces provisional categories, relationships, contrary cases, descriptive n/N coverage, next questions, and exact respondent evidence for mandatory human review. It is GT-inspired, not a complete Grounded Theory study, and does not claim objective truth, statistical prevalence, or theoretical saturation. The research-grade workflow claim remains locked until its public independent benchmark passes.
- Italiano: Evidence Synthesis è una beta pianificata per PAYG, Growth e Scale, informata dalla ricerca, collegata a evidenze respondent esatte e soggetta a revisione umana. Non equivale a uno studio Grounded Theory completo e non dichiara prevalenza statistica o saturazione teorica.
- Public marketing and legal pages display no language switcher. The homepage is English-only; existing Italian subpage translations remain in the repository and available by direct URL. The default consumer demo is Italian-language neutral customer discovery (`elicitra_customer_discovery`).
- The homepage explains the journey from adaptive interviews through configurable transcript analysis to conditional follow-up actions. Its functional use cases are lead qualification, user research, product and service feedback, opinion research, contact and customer intake, appointment preparation, and screening and triage. Lead qualification addresses performance marketing agencies, RevOps teams, and sales teams, including the handoff of paid-campaign lead context to a client's sales team. Examples illustrate possible workflows rather than customer testimonials. The Agent-first section presents MCP and the Betterstood CLI as interfaces for the customer's AI agents to author interviews, configure analysis, and prepare campaign drafts; it does not imply autonomous campaign launches.
- The homepage is a password-protected preview. Its primary CTA is "Try a live conversation". It opens the same local "Demo coming soon" availability dialog after client enhancement; before that, the buttons remain disabled. A quieter "or watch one" control is visible and, until the demo exists, opens that same dialog. The Paper/Olive switch changes contrast only. The demo campaign has not been selected, and the placeholder cannot start a conversation, request microphone access or contact details, or consume credits. The homepage has no signup or sign-in links; `/auth/signup` and `/auth/login` remain directly accessible.

- The independent Site runtime has a separately gated homepage popup for browser voice. With an explicitly approved browser configuration, its CTA opens consent, microphone/audio, Live Understanding and recap in one session, without a telephone number or account. It uses independent abuse, capacity and monetary limits and charges the visitor zero credits. Preparing this code or running a labelled local simulation does not establish public availability: installation, edge routing and real audio acceptance remain separate operational gates. The platform-hosted preview described above retains its existing behavior until the independently approved Site handoff.
- The homepage EU readiness summary covers the platform-owned AI notice, separate unselected optional choices, versioned notice receipts, the 90-day raw transcript/provider-event and provider-recording purge boundary, the EU-hosted core, and self-hosted public marketing analytics without analytics cookies, with Do Not Track support. It links to `/trust` and is a readiness statement, not a certification or a blanket guarantee that every subprocessor keeps all data in the EU.
- Optional vertical scenarios may support lead qualification or guided intake, but Betterstood is not a generic chatbot or a regulated financial or automotive sales product.
- Personal-finance behavior is available only through an explicitly configured scenario; it is not the generic fallback.
- Do not present Betterstood as providing credit scoring, loan approval, rate calculations, product comparison, official quotes, binding vehicle offers, or regulated financial advice.

## Public pages

- [Terms and conditions](https://betterstood.com/terms-and-conditions): Legal hub for respondent consent, SaaS and billing terms, privacy, DPA, subprocessors, cookies, and retention.
- [Trust](https://betterstood.com/trust): Operational AI Act and GDPR readiness claims, consent boundaries, retention, hosting, and deployer responsibilities.
- [Contact](https://betterstood.com/contact): Inquiry form for workflow scope, channel interest, volume, Scale onboarding, and design-partner requests.
- [Chat with your data — English](https://betterstood.com/methodology/ask-your-data): Read-only analyst contract, source-linked claim labels, Pro+ availability, and Evidence Synthesis boundary.
- [Chiedi ai dati — Italiano](https://betterstood.com/it/methodology/ask-your-data): Contratto dell’analista in sola lettura, etichette collegate alle fonti, disponibilità Pro+ e confine con Evidence Synthesis.
- [Evidence Synthesis methodology — English](https://betterstood.com/methodology/evidence-synthesis): Scientific foundations, evidence contract, human-review boundary, limits, and benchmark gate for the planned PAYG, Growth, and Scale beta.
- [Metodologia Evidence Synthesis — Italiano](https://betterstood.com/it/methodology/evidence-synthesis): Fondamenti scientifici, contratto delle evidenze, revisione umana, limiti e gate del benchmark per la beta pianificata PAYG, Growth e Scale.

## Machine-readable resources

The canonical application origin for this environment is `https://betterstood.com`. Discovery
links use that origin even when this index is requested through a retained legacy
hostname. Use the advertised canonical links for new integrations; existing
integrations may keep their legacy application URLs. The API and MCP remain at
`/api/*` and `/api/mcp` on the application origin. A public-origin change does not
rename MCP tools or rewrite published versioned contract identifiers. Sessions
are specific to the hostname, so signing in on one host does not sign a user in
on another.

- [Sitemap](https://betterstood.com/sitemap.xml): Stable public URL list.
- [Developer sitemap](https://betterstood.com/developers/sitemap-index.xml): Public Developer Access guides and reference pages.
- [Robots](https://betterstood.com/robots.txt): Crawl policy for public pages.

Public informational pages use HTTP content negotiation: request `Accept: text/markdown`
from the canonical HTML URL, or use its explicit `.md` alternate. The canonical
[English homepage](https://betterstood.com/) is at `/`, with
[English homepage Markdown](https://betterstood.com/index.md) as its explicit alternate.
The legacy `/it` and `/it/index.md` addresses redirect to `/` and `/index.md`,
respectively. `/it/` is also a legacy alias and ultimately reaches `/`.

The URL selects the language for the other registered public pages: the
unprefixed canonical path selects English and the `/it` prefix selects Italian,
independently of language preferences. Their explicit alternates append `.md`
to the localized path.
The password-gated homepage alternate returns only the gate document until the
signed demo cookie is valid. Standalone pricing and use-case Markdown alternates
are no longer published; the use-case page has been removed.
Both locked and unlocked gated representations remain private and non-cacheable;
requesting Markdown does not bypass demo access.

## Developer Access

Use the current guides and setup links on `https://betterstood.com` for this environment.
Their advertised MCP and unversioned documentation URLs follow the canonical
application origin, including in staging. Published versioned contract artifacts
and webhook identities retain their original bytes and identifiers.

- [Developer Access](https://betterstood.com/developers/): Human-readable root for the Public MCP v14, PATs, signed webhooks, compatibility, security, and limits.
- [Developer Access Markdown](https://betterstood.com/developers/index.md): Markdown sibling of the documentation root.
- [Agent bootstrap](https://betterstood.com/developers/agent-bootstrap.md): Compact, authoritative starting context for coding agents.
- [Developer documentation index](https://betterstood.com/developers/llms.txt): Machine-readable guide index.
- [Full developer documentation](https://betterstood.com/developers/llms-full.txt): Current public guides in one text resource.
- [Public contract v14 manifest](https://betterstood.com/developers/v14/manifest.json): Immutable MCP and webhook contract registry.

## Optional

- [Product overview](https://betterstood.com/): Password-protected launch overview for adaptive conversations, channel workflows, analysis, scoring, and follow-up actions.