Answers to
what you're thinking.

Mirror is a decision engine for sales conversations. Five available modules — Situation Solver, Brief, Live Assist, Practice, and Pipeline — each built for one specific moment in a sales conversation. Pitch Trainer, Analyse, and Coaching are launching soon. Mirror is different because the modules are designed to build on each other: Brief gives you the DISC profile, Practice lets you rehearse it, Live Assist helps you during the call.

B2B sales reps in the Nordics and Europe running complex deals — typically €5,000+ with decision-maker meetings and DISC-sensitive dynamics. Mid-market AEs, Heads of Revenue, CROs. Mirror is not built for SDR blasting or transactional inside sales.

No. Mirror is the decision layer on top of your pipeline — what helps you win the deal, not manage it. CRM integration (HubSpot, Pipedrive) launches H2 2026.

The loop. Gong only analyses backwards. Crystal only before the meeting. SalesEcho only during training. Mirror is designed to cover every moment: Brief (prepare) → Practice (rehearse) → Live Assist (during the call) → Analyse + Coaching (learn afterwards). In V1 the first three are available — Analyse and Coaching launch soon and close the loop.

Brief gives you the DISC profile based on CVR and LinkedIn. Practice lets you rehearse the opener and objections. Live Assist helps you during the meeting. Situation Solver resolves the acute decisions. In V1 these four modules plus Pipeline are available. Analyse and Coaching launch soon and will tie them together into one learning loop.

Memory is an upcoming layer that will store your own sales patterns over time — what works with C-types for you, objections you have rehearsed, signals from meetings. In V1, DISC profiling is based on CVR, LinkedIn, and Situation context. Memory launches as part of the Coaching module.

AI is used for pattern-matching and suggestions — not to write on your behalf. Brief verifies facts from CVR/LinkedIn, it doesn't generate them. Live Assist suggests tactical responses based on Mirror's tactical playbook, not ChatGPT with a sales prompt. We intentionally under-claim how AI-heavy the modules are.

You lead the conversation. Mirror suggests, you decide. Live Assist gives you the words, but you say them. Practice is one session at a time, not an automated routine. Mirror is built to make you sharper — not to replace you.

The CVR registry (public Danish company data), LinkedIn (public profiles only), and your own Mirror Memory (previous meetings, your notes). Three sources — no scraping, no closed LinkedIn groups, no email pulls from your inbox.

CVR is the official source, updated daily. Mirror pulls directly from the cvr.dk API — the same data the Danish Tax Agency and banks use. Data is typically less than 24 hours old.

No. Mirror has no microphone access and never listens to audio. You type what the prospect says (a few words is enough), Mirror suggests a tactical response. This is a deliberate choice — we do not want Mirror to be a surveillance tool.

Typically 1–2 seconds from when you type what the prospect says to when the suggestion appears. Frankfurt hosting + optimised inference gateway. Under high load it may rise to 2–3s.

Yes — Live Assist runs in your browser alongside your meeting. No plugin, no integration required. You just have an extra tab open.

Pitch Trainer is not available in V1 — launching soon. It is a structured coaching session with an AI prospect calibrated per DISC type and live hints. Practice (drill format) is available now.

Analyse is not available in V1 — launching soon. Planned: Audio (.mp3, .wav, .m4a, .ogg) up to 60 min, video (.mp4, .mov) up to 60 min — only the audio track is used — or plain text notes.

Coaching is not available in V1 — launching soon. Planned: typically 8–12 meetings + 5–10 Practice sessions. Patterns are only shown when there is a statistical signal — not based on single sessions.

Follow-up is not available in V1. When it launches: Mirror suggests the angle + draft email — you edit and send it yourself from your own email system. No automation. No mass-send.

Free = permanent freemium. 1 sales rep, 8 briefs/month, 3 Situation Solver analyses/day, 30-day Memory retention. Pro removes the limits: unlimited briefs and analyses, infinite Memory retention, unlimited Practice reps, priority support. Pro costs €67/mo on annual billing, €107/mo on monthly.

Team adds Memory sharing across the team (winning patterns spread), SSO (SAML/OIDC), DPA + GDPR addendum, audit log export, dedicated CSM. Price on request — depends on seat count from 5+. No lock-in.

Yes. Switch happens in settings, no interruption, no migration. Memory retention stretches automatically to infinite when you upgrade.

Yes. No lock-in — you cancel in settings with one click. Monthly plan: access until the end of the billing period. Annual plan: you keep access until the end of the year. No 'why are you leaving us' email.

With 10+ seats on Team tier you get a discount. Write to sales@mirrorsub.com with the number of reps — we'll send a quote within 1 business day.

Frankfurt (eu-central-1) as the primary region. Backups in Stockholm (eu-north-1). Certain calls — AI inference, payments, auth — go to US-based sub-processors (Anthropic, Stripe, Clerk) under EU Standard Contractual Clauses (SCC). Full sub-processor list: /juridisk/underleverandoerer.

By default: no. Your meetings and messages are not used for model training unless you explicitly choose to share them. It is opt-in — not opt-out. Anthropic discards inference logs after 24h and does not train on API calls from customers either. See terms: /juridisk/underleverandoerer.

Standard DPA on all Team and Enterprise plans, signed PDF available on request via security@mirrorsub.com. Custom clauses possible on Enterprise.

Mirror employees cannot view customer data without a ticket + 4-eyes approval + MFA. Production access is logged. Audit logs can be exported on Enterprise tier to SIEM (Splunk, Datadog).

30-day export window. You receive a database dump in an open format (JSON + Parquet). No lock-in at the data layer. This is hard-coded in our company articles.

Yes. Cancellation triggers a 30-day grace period + automatic deletion from all systems (production + backup). Audit trail of the deletion available on request.

External penetration testing is planned before we open up to larger contracts. Internally we run continuous dependency scanning and SAST. We will update this page when external pentest is completed.

SOC 2 Type II audit Q4 2026. ISO 27001 audit Q3 2026. NIS2 compliant from day 1. GDPR from day 1.

Privacy@mirrorsub.com — we respond within 30 days as GDPR requires. Self-service export available from settings on Pro+.

CRM integration is planned for H2 2026. HubSpot and Pipedrive launch first, Salesforce on the roadmap. Mirror already works without a CRM — all modules run standalone.

Calendar and email integration is on the roadmap (planned H2 2026). No integration is required to use Mirror — all modules work standalone.

Planned on the roadmap. Mirror currently does not send notifications to Slack or Teams.

Danish + English fully. Norwegian + Swedish work (same Nordic language model). German and Dutch are on the roadmap.

Yes. Mirror's tactical playbook is trained on Danish B2B data — the patterns and objections that apply in the Nordics are not identical to American sales. Dutch and German will get dedicated calibration.

Under 60 seconds. You can have your first Brief ready quickly after a name is entered — typically within a few seconds. Team onboarding: ~30 min for core features, one week to integrate into daily workflow.

Founders (Morten + Martin) on Pro. Dedicated CSM on Team+. Ticket response within 4 hours during business hours. No support bot.

Pro: best-effort, 4h response. Team: 99.5% uptime, 2h response. Enterprise: 99.9%, 1h critical.

Our codified methodology for B2B sales conversations. Built on established negotiation techniques (calibrated questions, mirror, label, tactical empathy) but calibrated specifically for B2B SaaS, mid-market complex deals, and DISC patterns in the Nordics/Europe. It is not a black-box AI — every suggestion can be traced back to a named technique.

Illustrative examples and cases on /kunder are marked as hypothetical or verified. We never use 'up to X%' phrasing or fabricated statistics. All figures cited are either from public data sources or clearly marked as estimates.

Each case study has a methodology note: data period, n, baseline, measurement period. Hypothetical examples are clearly marked. Request the full methodology report at research@mirrorsub.com.