Skip to content
Book a call
MenuClose
GTM AI Bootcamp4 weeks · $5,000 for the first five teams, then $10,000Packages and pricing
SolutionsMarketingSales developmentSalesCustomer successRevOpsGTM engineeringProductExecutive
How we workOur processWhat you own: your GTM brainGTM data warehouseAgentic tool implementationAI reporting and analysisGTM strategyCustom signals and intentSkillsAgentsAutomated workflowsAI workspaceCustom apps and MCPsAI telemetryAI self-improvementGTM project management
PackagesImplementationAI operating partnershipManaged Agents
TechnologyGTM technology advisoryClaude CodeClaude CoworkClayHubSpot, Salesforce and AttioEverything else we work with
Why B2B CatalystWhy B2B CatalystAbout
Book a call
How we work · What we build · Step 03

Your best rep's process, as a skill every rep can run

Repeatable AI procedures like “prep this call” or “draft this follow-up”, written from how your top performers work and loaded with your context.

Book a call30 min · no prep needed
Step 01ConnectStep 02PrioritizeStep 03ShipStep 04Improve

What you're after

01

The whole team working like your best rep

The steps your top performer takes at your costliest buyer decision, run the same way by your newest hire.

02

Buyers who hear your best argument

Call prep and first drafts start from the argument that wins at that moment, whoever the buyer talks to.

03

Prompts that don't walk out the door

One shared library in your repo. When we improve a skill, everyone gets the new version at once.

Open one skill file

A skill is a short file your AI workspace loads when the job comes up. Illustrative: a pre-call planning skill. Companies, people and numbers are fictional.

PartWhat it holdsIn Northwind's pre-call planning skill
TriggerWhen the skill runs, in the words people use“Prep my call with Hartwell Freight”, or a first meeting on tomorrow's calendar
InputsWhat it needs and where it gets itThe account and attendees, through one connector to your whole business
Context it readsThe files that make it yoursNorthwind's ICP, the VP Operations persona file, past calls with this account
Playbook stepsYour best rep's approach, in orderList what changed since the last touch. Tie it to peak-season readiness: missed deliveries, overtime. Draft three questions and one proof point.
ExamplesWhat good looks like, and what doesn'tTwo briefs the top AE wrote, plus one weak brief that leads with route-optimization features, and why it's weak
ChecksWhat it confirms before handing anything backEvery claim cites a source. Names match the CRM. One page, no more.
Version and ownerWho changes it, and what changedv7 in your repo, with a change log. Owner: enablement.

Where the content of a skill comes from

The people who win at that moment

We sit with the reps who handle your costliest decision best and write down what they do, including the steps they skip.

Your won and lost calls

The arguments that moved buyers, pulled from the raw transcripts.

GTM strategy

The revenue play it serves

Each skill is written for the play we're shipping at that decision. We don't hand over a generic prompt library.

Why scattered prompts don't add up

Prompts in chat history
  • Every rep writes their own
  • Nobody knows which ones work
  • The best ones leave with the best rep
Skills in your repo
  • One skill per job, built from your top performer
  • Checked against real examples and a written definition of good
  • Shared and versioned, with a named owner

Skills across the four steps

  1. Step 011

    Connect your data

    Skills read your warehouse and strategy files through one connector, so they start from what your best people know.

  2. Step 022

    Prioritize by revenue impact

    We write skills for the buyer decision that costs the most revenue, not the easiest demo.

  3. Step 033

    Ship into the tools your team uses

    Skills ship into your team's AI workspace with training on real accounts. Working versions in days.

  4. Step 044

    Keep improving

    When the read on a play changes, the owner edits the skill and the version number moves.

Who runs them

Enablement

Enablement is plugin development now. Your enablement team owns the skill files and the examples of good, and we train them to change both.

Sales

Call prep, follow-ups and executive briefs.

Read more

Sales development

First touches and account research.

Read more

Your agents

Always-on agents run the same skills, so people and agents follow one playbook.

Where to start

GTM AI Bootcamp

The bootcamp ships the first skills for the play we build with your users. More come through Implementation or the AI operating partnership.

Book a callSee GTM AI Bootcamp

Before you book

Can't our reps just write prompts?

They can. A skill is a prompt made reliable: tied to your context, checked against real examples, and shared so everyone gets the improvement.

Our builders already write skills in Claude. What's left for you?

Choosing which job gets a skill. We find the buyer decision where you lose the most before anyone writes one. What we add for builders.

Do we need formal evaluations?

At most teams' volume, no. Each skill gets a written definition of good and a few real examples to check against. Heavier testing comes when volume justifies it.

Skill or agent?

A skill does one job when a person asks. An agent starts on its own, runs skills, and waits for a named person to approve. Most teams need skills first.

Where's your line stuck?

Find the first skill worth writing

30 minutes. Tell us the buyer decision your team handles worst. We'll say which skills come first.

Book a call