Account Research

Using Octave in Claude

Workflow

Account
Research

Surface

Claude Chat (MCP)

Cowork

Claude Code

Difficulty

Easy

Medium

Hard

Audience

Sales Reps, SDRs, and AEs doing outbound

What it does

Build a skill that connects the Octave MCP to your internal data sources and synthesizes everything into a unified account brief

Scenario

You've got a target account. Before you reach out, you want the full picture -what they do, what they're using, what conversations your team has already had, what their product usage looks like. Right now that means opening four or five different tools and stitching it all together yourself. By the time you've assembled everything, you've lost half your morning and the picture is already stale.

How It Works

You create a skill that does the orchestration. The skill connects to the Octave MCP for GTM intelligence -playbook match, segment fit, what messaging works for accounts like this -and hooks up to your internal data sources via their MCP connections for the facts: CRM history, product usage, call recordings. One prompt triggers the skill; it handles every call and synthesizes the output into a single account brief.

Pick Account

Name or domain

Octave MCP

Playbook, segment, personas

Internal Data

CRM, product data, calls -via MCP

Account Brief

Unified, ready to act

The Old Way

❌ Tab Surfing

Jumping between CRM, product dashboards, call recorder, and web research to piece together the picture

❌ No GTM Context

You see raw data but nobody's telling you what it means -which playbook fits, which persona you're talking to, what's worked before

❌ Manual Synthesis

You're connecting the dots yourself across every source, every time, for every account

❌ Stale on Arrival

By the time you've assembled everything, the picture is already incomplete -someone else had a call yesterday you don't know about

⏱️ 30–45 minutes per account

The New Way

✓ One Prompt

Run your skill with an account name -it calls the Octave MCP and your internal data MCPs automatically

✓ GTM-Enriched

Octave matches the account to playbooks, segments, and personas so you know the "so what," not just the "what"

✓ Auto-Synthesized

Your skill pulls CRM data, product usage, and call recordings via MCP and combines them with Octave intelligence into one brief

✓ Always Current

Pulls live data from every source at the moment you need it -no stale spreadsheets, no missing context

⏱️ 2–3 minutes per account

What You Get

✓ Account Brief

Everything in one place -company profile, product usage, CRM history, call highlights, and web research synthesized into a single document

✓ Playbook Match

Octave tells you which playbook fits this account and why, based on your library's segments, personas, and use cases

✓ Talking Points

Key angles grounded in what's worked with similar accounts in similar segments -not generic templates

✓ Ready to Act

You know who to talk to, what to say, and why it matters -before you ever pick up the phone

Step by Step

1

Connect your MCPs

Your skill needs access to the Octave MCP plus your internal data MCPs. Each connection gives the skill a specific set of tools it can call.

MCPs your skill needs: 1. Octave MCP → enrich_company, qualify_company 2. CRM MCP (e.g. Salesforce) → query accounts, deals, activities 3. Call Recording MCP (e.g. Gong) → search calls by account 4. Product Data MCP (e.g. BigQuery) → query usage by domain

You set up MCPs once. The skill references them every time it runs. Start with Octave + CRM, add more sources as you go.

2

Build the skill: define your inputs

Your skill takes an account domain as input. Everything else is derived from the MCP calls. Keep the interface simple so anyone on the team can use it.

Skill inputs: - companyDomain (required): "acme.com" - depth (optional): "quick" or "full" quick = Octave only, full = Octave + all internal sources

3

Build the skill: Octave MCP calls

The skill calls enrich_company to get company profile, firmographics, and tech stack, then qualify_company to match against your library's segments and playbooks. This is the GTM intelligence layer.

Octave MCP sequence: 1. enrich_company({ companyDomain }) → Octave MCP Returns: company profile, firmographics, tech stack 2. qualify_company({ companyDomain }) → Octave MCP Returns: segment match, playbook fit, persona map, competitive landscape, suggested talking points

4

Build the skill: internal data enrichment

The skill calls your internal MCPs to get the current state of the account. This is the "what's actually happening" layer that pairs with Octave's "what it means."

For the account: 1. CRM MCP → query account by domain Returns: deal stage, owner, recent activities, open tasks 2. Call Recording MCP → search calls by account name Returns: recent call summaries, key topics, objections raised 3. Product Data MCP → query usage by domain Returns: active users, feature adoption, usage trends

This is where Medium difficulty comes from. Each MCP connection adds a data layer. Start with CRM if you only have one internal source.

5

Build the skill: output format

The skill synthesizes all sources into a single account brief. Define the sections you want so the output is consistent every time, for every rep.

Account brief sections: - Company Overview: profile, firmographics, tech stack - GTM Match: segment, playbook fit, competitive landscape - CRM Status: deal stage, owner, recent activities - Call History: recent conversations, topics, objections - Product Usage: active users, feature adoption, trends - Talking Points: angles grounded in playbook + context - Recommended Next Step: based on all of the above

6

Run the skill

Give it a domain. The skill handles all the MCP calls, enrichment, qualification, and synthesis. You get back a complete account brief.

Research acme.com for me. Pull everything we know: Octave enrichment, CRM history, product usage, and any recent calls. Give me a full account brief.

7

Iterate, then chain into Find + Qualify People

Review the brief. Ask follow-ups: "what's the competitive situation?" or "any red flags?" Once the brief is solid, chain directly into the Find + Qualify People workflow. The account context carries forward.

Now find the VP of Engineering and VP of Platform at this account. Qualify them against our personas and pull internal context for each.