Modjo

What it is

Modjo is a Paris-built conversation intelligence platform: it records sales calls and video meetings, transcribes them in 100+ languages, scores them, and writes the result back into Salesforce or HubSpot. Founded in 2019, it has raised more than €14 million (including a €7M Series A in 2021 backed by Daphni and Aglaé Ventures) and claims 600+ B2B revenue teams as customers, most of them in Europe.

Two things separate it from the Gong template. First, the data residency: calls are stored in France, on AWS in the Paris region, under SOC 2 Type II and GDPR. For a European sales org whose DPO has already said no to a US-hosted recorder, that is not a footnote — it is the shortlist. Second, the licensing unit. Modjo bills recording seats only; managers, enablement, leadership and anyone who listens without recording get unlimited free listener licenses. On a team where five managers coach thirty reps, you pay for thirty.

On top of the recorder sit two layers that matter in 2026. AI CRM Filling summarizes each call and writes to any native or custom field on any object, which is how teams get MEDDIC or BANT fields populated without asking reps to type. And agents: prebuilt ones — Prep’ My Meeting, Challenge My Deal, Build ROI Case, My Sales Coach, Email Follow Up, Cross-Sell Detection, Churn Signals — plus a builder for your own, running against the call and deal record.

Pricing reality

Quote-only. The pricing page checked on 10 October 2026 publishes no figures and no tier names; it routes you to a demo form segmented by license count and states the product is “optimal from 3 sales reps.” There is no free plan and no self-serve trial.

The numbers below are triangulated from procurement and review data, not list price. Negotiate against them; don’t quote them back to the vendor.

• Per recording seat: about €99/user/month on annual billing (roughly $108), the most commonly reported figure. Spendbase’s dataset shows a band — around €79-89 at a 10-seat commitment, €99 at 15, and €119-139 on 25+ seat deals.
• Seat floor: a 15-seat minimum is the figure most often reported, despite the “from 3 reps” line on the site. Treat smaller deals as possible but not standard.
• Implementation: a one-time onboarding fee of roughly €2,000-5,000 is reported on larger deployments. A named account manager runs a 30-day onboarding.
• Listener seats: included, unlimited, no charge.

Worked for the case that put Modjo in this queue — a 35-rep team asking for a Gong-like analyzer under $100/seat: 35 × €99 × 12 is about €41,600/year (≈$45,000), with managers free. The same 35 seats at Gong’s triangulated $1,400-2,000/seat/year plus a $5,000-50,000 platform fee lands at roughly $54,000-120,000. Modjo comes in at about 40-85% of Gong depending on where your Gong quote falls — cheaper, not a different price class. Against Avoma fully configured ($82/seat/month annual for recording, coaching and deal intelligence, published), Modjo is about 30% more per seat; you are paying that premium for EU hosting and the agent layer.

Best for

A RevOps or sales leader at a European B2B company with 15-80 quota-carrying reps on Salesforce or HubSpot, where data residency is a procurement requirement and the goal is CRM hygiene plus coaching on one contract. The scoped case where Modjo wins outright: a French or DACH sales team with a manager-heavy coaching model, where Gong’s per-user pricing would bill every listening manager and Modjo’s would not.

Not for

Teams under roughly 15 reps — the reported floor makes the per-rep cost worst exactly where budgets are smallest, and Fathom or Fireflies at $0-30/seat/month plus Claude-side analysis covers the decision value. It is also not the pick for 150+ rep US-centric orgs that need cross-market deal inspection and forecasting at the depth Gong and Clari sell; Modjo’s deal layer is younger and its US customer base thin.

Versus the alternatives

• Gong — the share leader (ARR past $500M, 5,000+ customers). Pick Gong when you have 40+ reps, US data residency is acceptable, and coaching is a funded program with a named owner. Pick Modjo when EU hosting is mandatory or when free listener seats change the math on a manager-heavy team.
• Chorus (ZoomInfo) — the other incumbent by share. Pick it when ZoomInfo is already your data layer and Chorus is an add-on line on that contract. Pick Modjo when you don’t buy ZoomInfo, or when European language coverage on calls matters more than US contact data.
• Attention — the fastest-growing entrant ($30M Series B in June 2026, ARR up 4x). Its MCP server writes to the CRM; Modjo’s is read-only. Pick Attention when agents taking the action is the deliverable. Pick Modjo when EU residency and a read-only connector are what security will sign.
• Avoma — published pricing and no seat floor. Pick Avoma under 15 reps or when you need booking and inbound routing in the same product. Pick Modjo above that size when residency in France matters.

If none of these fit, the tell is usually that the team is too small for conversation intelligence at all: buy a notetaker, export transcripts, and run the analysis in Claude.

The MCP server

Modjo ships an MCP server at https://api.mcp.modjo.ai/v1/mcp, listed in Claude’s connector directory and documented for ChatGPT, Cursor, Dust, n8n, Make, Zapier, Codex CLI and Gemini CLI. It exposes 12 tools, all read or analysis: get_calls, get_deals, get_accounts, get_contacts, get_emails, get_transcript, semantic_search_activities, three ask_anything_on_* tools for call, deal and account, plus get_users and get_agents. Nothing writes. A rep can ask Claude why last quarter’s deals were lost and get reasons clustered from the closing calls; the assistant cannot touch a CRM field.

Watch-outs

• The API-key path bypasses role permissions. OAuth connections inherit the user’s Modjo access; a workspace bearer token is organization-scoped and reads every call, deal and account regardless of role. Guard: connect every human through OAuth, keep the workspace key with one admin for server-side automations only, and rotate it when that person changes role.
• CRM Filling writes to production fields. “Any object or field” includes stage and amount fields a forecast depends on. Guard: map fields in a Salesforce sandbox or a HubSpot test property first, start with empty-only custom fields (MEDDIC, next step), and audit two weeks of writes before enabling standard fields.
• No trial, and the floor is a contract term. You find out about accuracy after you have committed to roughly 15 seats for a year. Guard: before signing, send 10 recorded calls in your actual languages and accents for a test run, and get a 60-day exit or seat-reduction clause written into the order form.
• Dialer imports filter calls. Aircall’s integration page states only two-participant calls longer than two minutes are imported, so short connects and conference calls vanish from coaching data. Guard: confirm the import rules for your dialer in writing, and track connect-rate in the dialer itself, not in Modjo.

Related: alternatives to Gong covers the wider replacement set, and the conversation intelligence stack prices the layers around it.