Codify the knowledge of your top performers into organizational skills.

Memory is what your organisation knows: ICPs, personas, messaging, sales process, win patterns, the rules and judgement behind how you sell. Skills are what your organisation does with it: codified GTM actions like Prospecting, Meeting Prep, Pipeline Review, Discovery, Objection Handling and QBR Prep. Every skill connects to memories - a Meeting Prep skill reads your personas, your competitive frames, your qualification criteria - so when Memory versions up, every skill built on it sharpens with it. Knowledge without execution sits on a wiki. Execution without knowledge is automation guessing. You need both, and they're separate layers on purpose.

Playbooks get read once, then forgotten. A playbook is a document that describes the play; a skill is the play, executable. The Objection Handling skill doesn't sit in a folder waiting to be searched mid-call - it preps the rep before the call with the specific objection this account is likely to raise and how your top performers answered it on closed-won deals. Keep your document store. Codify the plays that should run.

No. A sales leader put it to us directly: "I don't want to make people who aren't Bill, Bill." Skills raise the floor, not clone the personality. What gets codified is the pattern - which stakeholder your best rep multi-threads at stage two, which proof point they lead with against this competitor - not their phrasing or their charm. The rep runs the play in their own voice. The floor rises; the ceiling stays theirs.

You do. The system surfaces patterns from closed outcomes; nothing becomes canon without approval. Every proposed skill and every proposed change carries the evidence that justified it - the deals, the calls, the numbers - so you can see why before you approve. RevOps owns the approval, the versioning, and the rollback: if v12 underperforms, v11 is back across the team instantly. The system never silently changes how your team sells.

From what's already happened. The system reads your closed-won and closed-lost deals - the calls, the emails, the stage movement, the stakeholder pattern - and surfaces what your top performers actually do. We then sit with you to confirm: this is what your best reps do; do you want it to become the team standard? Most teams find their codified skills are sharper than the playbook they tried, and failed, to write themselves.

A starter library on day one: Prospecting, Meeting Prep, Pipeline Review, Discovery, Objection Handling, QBR Prep, account research, follow-up drafting, deal scoring, coaching briefings. We tune the foundation to your business in week one. Then you build your own: if a person on your team does it regularly with judgement, you can codify it as a skill. Describe it in natural language; the system codifies it; you version it, govern it, ship it across the team.

For low-risk skills, yes - research, enrichment, scoring, draft generation. For higher-stakes skills - sending the email, advancing the stage, contacting a customer - there's a human approval step by default. You set the line, per agent, per skill, per team. Every run is traced, so you can move a skill from "needs approval" to "auto-run" once you've watched it perform.

Skills are built on Memory, so they inherit the change. When your ICP definition versions up, every skill that reads ICP - Prospecting, deal scoring, territory prioritisation - switches over the same day. No re-training, no re-writing six playbooks that reference the old ICP. And Learning works the other direction too: when closed outcomes show a skill drifting from what wins, it proposes the upgrade and you approve it.

Yes. Skills segment by anything you need - product line, region, ICP tier, sales motion - and version independently. The Discovery skill your EMEA mid-market team runs can differ from the North American enterprise one. Company-wide standards inherit down; local skills override where it matters.