Attention

What is Attention?

Attention is an AI-native revenue and conversation-intelligence platform that captures sales conversations (calls, emails, messages) and deploys AI agents to automate post-call administrative work. It records and transcribes meetings in 100+ languages, automatically extracts deal data and fills CRM fields, generates personalized follow-up emails, delivers AI coaching scorecards against sales frameworks (MEDDIC, BANT, SPIN), and identifies deal risk patterns—reducing manual CRM entry and enabling sales teams to focus on selling.

Attention news

2026-06-24
Attention Raises $30M Series B Led by RTP Global
Announced Series B funding of $30 million led by RTP Global, with participation from returning investors Aglaé Ventures, Eniac, Alven, and new investor Linea. At the time, the platform was running 20+ million agent actions per month across 500+ customers including Abridge, Scale AI, Lovable, Preply, Engine, and BambooHR. ARR was growing at 4x year-over-year.
Source: Attention (company announcement) ↗
2026-10-29
Attention Dropped Live Coaching, Doubled Down on CRM Auto-Fill
CEO Anis Bennaceur revealed that Attention deprecated real-time live coaching features in favor of post-call follow-up automation (CRM field filling, follow-up emails). The shift was driven by customer validation: sales teams will pay to have CRM filled automatically long before they pay for AI coaching. Daily pain (empty CRM) outweighs nice-to-have features (coaching).
Source: Built This Week (CEO interview) ↗
2026-11-10
Attention Platform Now Runs 30+ Post-Call AI Agents; MCP Server Expands to 82 Tools
Attention expanded its agentic surface from knowledge base, follow-up email, deal risk, and CRM auto-update agents to 30+ forkable agents (lead qualification, competitive intelligence, sales forecast, etc.). The MCP server now exposes 82 tools across 15 groups (Calls, AI Analysis, Scorecards, Forecasting, CRM, Reports, Users & Roles), enabling deep Claude integration and workspace-level automation.
Source: Attention MCP Documentation ↗

Key features

Call Recording & Transcription

Records sales meetings on Zoom, Google Meet, Microsoft Teams, and other platforms; transcribes in real-time across 100+ languages with speaker identification and sentiment analysis.

Automatic CRM Updates

AI automatically fills Salesforce or HubSpot CRM fields after each call—extracting deal stage, next steps, pain points, objections, timeline, decision-makers, and competitor mentions without rep manual entry.

Follow-Up Email Generation

AI drafts personalized follow-up emails reflecting specific conversation points, next steps, and prospect concerns, ready for rep review and send within seconds.

Real-Time Coaching

During live calls, surfaces relevant talk tracks, objection handling suggestions, competitive responses, and battlecards via browser overlay without interrupting the call.

AI Coaching Scorecards

Generates automated scorecards assessing each call against configured frameworks; identifies rep strengths, gaps, and patterns from top performers across the team.

Deal & Account Health Scoring

Analyzes conversations to score deal probability, identify churn risk, and surface pipeline forecasting insights grounded in actual call content rather than rep self-reporting.

AI Sales Agents

Customizable automated agents that handle knowledge base updates, lead qualification, competitive intelligence tracking, deal risk monitoring, upsell/cross-sell identification, and meeting recaps with action items.

MCP Server / Claude Integration

Exposes 82 tools across 15 groups via MCP protocol, enabling Claude and other LLMs to search calls, analyze deals, configure scorecards, and manage workspace—differentiating from read-only connectors.

200+ Integration Catalog

Connects to CRMs (Salesforce, HubSpot), video conferencing, email, Slack, sales engagement platforms, and custom tools; supports API, webhooks, and MCP for workflow automation.

Attention pricing

Attention does not publish pricing publicly; pricing page (attention.com/pricing) returns 404. All pricing is quote-based and gated behind a scheduled sales demo. No free trial or free plan available. Industry benchmarks and aggregators estimate Team tiers at ~$800–$1,200/user/year and Enterprise at ~$1,200–$1,800/user/year, plus potential platform fees ($5K–$20K) and onboarding costs ($7.5K–$28.5K). Customers typically pay based on seat count, number of agent actions, CRM integrations, and annual contract commitment; no seat minimum publicly disclosed. Enterprise ACVs reportedly average $120K.

Pros & cons

Pros

+ Eliminates manual CRM data entry: 90% of post-call admin work automated; reps save 10+ hours per week per user.
+ Near-human accuracy on CRM field extraction: AI fills dozens of fields from unstructured conversation (e.g., renewal date, tech stack, spending habits) that reps rarely log manually.
+ Deep agentic surface: 82 MCP tools enable autonomous agents beyond call analysis—knowledge base building, lead qualification, competitive tracking, deal risk monitoring—and integrate directly with Claude and other LLMs.
+ Transparent deal and forecast intelligence: Every coaching score, recommendation, and CRM field tied to verbatim transcript quotes; no 'vibes' or unexplainable outputs.
+ Multi-language support: Transcription and analysis across 100+ languages supports global sales teams.
+ Proven win-rate lift: Customers report up to 40% improvement in win rates and 5x coaching efficiency gains (e.g., Abridge).
+ Integration depth: 200+ integrations and native CRM write-back (Salesforce, HubSpot) reduces tool sprawl.
+ Customizable scorecards and frameworks: YAML-based rubrics are editable and forkable; frameworks (MEDDIC, BANT, SPIN) ship as templates.

Cons

− No public pricing: Enterprise sales cycle required; opaque pricing makes budget planning difficult and prevents self-serve evaluation.
− No free trial: All evaluation requires demo with sales team; no low-friction onboarding for smaller teams or exploratory use.
− Quote-only model creates information asymmetry: Third-party sites publish conflicting per-seat estimates; no official per-seat or per-action price guidance.
− Real-time coaching is post-call, not live: Attention deprecated live coaching in favor of post-call follow-up; competitors like Quotadot offer live mid-call guidance.
− CRM-dependent: Platform requires integration with Salesforce, HubSpot, or another CRM to deliver full value; standalone call analysis without CRM connection is limited.
− Pricing scales with actions, not just seats: High-volume call centers may hit action ceilings; cost per agent action can escalate with usage.
− Minimum contract terms apply: No month-to-month pricing published; sales deals are typically annual with minimum seat or ARR commitments.

Who is Attention best for?

• Enterprise sales teams (25–500 reps) with high CRM adoption and strong desire to eliminate post-call admin and boost forecast accuracy.
• Revenue-focused organizations (Sales, RevOps, Sales Enablement) that standardize coaching methodologies (MEDDIC, BANT, etc.) and measure win-rate drivers.
• Companies with existing Salesforce or HubSpot deployments seeking native CRM auto-fill and field mapping without custom integration.
• Global sales teams: 100+ language support and multi-timezone call analysis reduce friction.
• Sales leaders and RevOps teams that want Claude-native agentic workflows (via MCP) to automate deal risk, forecasting, and competitive intelligence.
• Organizations comfortable with enterprise sales cycles and annual contracts; not suitable for teams seeking self-serve pricing or free trials.

Frequently asked questions

Can Attention work with CRMs other than Salesforce and HubSpot?

Attention integrates natively with Salesforce and HubSpot; generic API and webhook support enables custom CRM connections, but mapping and automation may require custom development.

Does Attention record calls automatically or require rep action?

Automatic: Attention connects to the rep's calendar and auto-joins Zoom, Google Meet, Microsoft Teams, and other platforms. No manual record-start required.

Can Attention be used for customer success or post-sale conversations?

Yes. Attention offers a customer-success solution that monitors call-based health and churn signals; it is not limited to pre-sale or sales conversations.

What is the difference between Attention and Gong or Chorus?

Attention focuses on agentic automation (auto-fill CRM, draft follow-ups, trigger workflows) and uses Claude for reasoning; Gong and Chorus emphasize call intelligence and reporting. Attention's MCP integration is deeper than competitors' read-only connectors. Pricing: Gong is $1,200–$1,800/rep/year with a 25–50 seat minimum; Attention's quote-based model claims lower seat floors.

How does Attention handle accuracy and hallucinations in CRM field extraction?

Attention spent 3+ years handling edge cases and errors. All coaching scores and CRM recommendations tie to verbatim transcript quotes. Fields are AI-extracted but reviewed on the CRM tab before manual sync or automatic push, depending on configuration.

What is Attention's relationship to Claude and Anthropic?

Attention uses Claude as its core reasoning engine; Anthropic features in case studies and marketing. Attention publishes an MCP server and Claude Connector allowing Claude users to query and act on Attention call data natively.

Does Attention offer a free or freemium plan?

No. There is no free tier, free trial, or freemium option. All evaluation is gated behind a sales demo.