Real onboarding. Honest plans.

Recon works on Sales Hub Professional and above, plus Marketing Hub Professional and above. Free and Starter portals do not expose enough deal and engagement data for useful clustering.

Every plan connects to all of them: HubSpot, Apollo, GA4, Google Ads, LinkedIn Ads, Meta Ads, n8n webhooks, and CSV upload. Starter and Team get the same source breadth. Scale adds multi-portal HubSpot and custom data integrations.

No per-seat fees on any plan. Every plan includes unlimited team members. Pricing scales with data sources, Scout experiments, governance features, and support level, not with how many people log in. Bring your whole RevOps, marketing, and exec team.

Recon AI is the always-on GTM analyst built into the product. Think of it as the RevOps director you cannot afford to hire full-time. It ingests every system you connect (CRM, ads, analytics, billing), classifies every contact and account, recommends where to spend, where to cut, and which segments to push next, then tracks whether the calls you took actually moved revenue. Answers come back with citations to the underlying deals and accounts. Built on Anthropic Claude Sonnet 4.6 (the agent) and Haiku 4.5 (classification). Strip the models out and the product would not work.

The closed loop. Breeze and Einstein ship recommendations into a void and never check whether they worked. GPT wrappers are read-only chat boxes layered on CRM data. Recon is a pipeline (ingest, classify, derive, recommend, brief, track) and the recommendation_outcomes ledger turns every call into a 30/60/90-day verdict that sharpens the next one. Three properties stacked (CRM-native, AI-native, closed-loop) are what's never been combined in production until Recon. Native recon_* writeback to HubSpot is the last edge of the loop and ships next.

Connect HubSpot, hit go, and the initial backfill completes within a few hours for most portals. You will see a usable Brief dashboard the same day. Full ICP convergence takes 7 to 14 days as Recon learns from your win patterns.

All customer data is stored in Supabase EU (Frankfurt). Application processing runs on Vercel's EU region. AI inference (Anthropic Claude) is routed through EU regions where the model provider supports them, otherwise US, per the published sub-processor list. Recon is GDPR compliant by default. We sign DPAs with every paying customer.

Recon AI is built on Anthropic Claude. Your data is never used for model training. Anthropic processes prompts in EU regions where available and does not retain content for training. We can show you the audit log of every Recon AI call against your workspace.

Yes, opt-in. After you connect HubSpot you can enable writeback from /app/connect. Recon idempotently creates a small set of custom properties on your contact, company, and deal schemas (recon_tier, recon_fit, recon_segment, recon_last_synced) and PATCHes them after every scoring cycle. Switch it off any time and the existing values stay on your portal (we never delete on disable). The field-mapping disclosure on /app/connect lists every scope and matches the actual OAuth code line-for-line.

Yes. Monthly plans are cancel-anytime. While Recon is in early access there is no annual lock-in. Annual contracts will be available at general availability with a discount.

All workspace data is hard-deleted within 30 days of cancellation. You can request earlier deletion via support. Backups are purged on the same schedule.

Every beta customer starts with a single upfront 90-day Beta Access fee — paid once, less than three months at full price. No auto-renew during the trial. After 90 days you pick monthly or annual at the rates on the pricing page. Get on the early access list and we'll reach out within 48 hours to scope your trial.

The calculator on the homepage uses your inputs for marketing spend, sales headcount, and tooling, multiplied by the share of pipeline that is off-ICP. We blend a loaded cost of around £130,000 per UK sales rep, an SDR and AE midpoint with employer NIC, benefits, OTE at plan, and ramp factored in. UK rep cost benchmarks come from Bluebird and Live Digital UK SaaS salary surveys 2025 and the Pavilion 2024 B2B SaaS Performance Metrics report. The 21% wasted marketing budget figure is from Forrester research on bad data and targeting. The 40 to 60% pipeline lost to no-decision figure is from Matthew Dixon's JOLT Effect research, 2022. The 67% quota-attainment uplift figure is from TOPO and Gartner sales analytics research. Adjust the slider to match your reality.