gtm-motion

GTM motion

The strategy layer for selling: who to pursue, when, and in what order, written down so it can be run every week. It sits above the other skills in this plugin - signal-prospecting finds accounts with a reason to talk, account-play plans one account in depth, and outbound-sequences gets the messages delivered.

Works for any role and product: a founder doing their own sales, an enterprise account executive with a named territory, an account manager growing existing customers, a consultant selling services.

What it produces

A single file, gtm-motion.md, in the user's working folder (or pasted back if there is no folder), containing:

1. The motion type and why.
2. The ideal customer profile (ICP), disqualifiers and the fit-scoring rubric.
3. A tiered account list, or the search brief to build one.
4. The signal library with weights tuned to the product.
5. The buying personas and what each cares about.
6. The play shape per tier and the weekly routine.
7. The guardrails: what data may be used, channels allowed, and that nothing is sent without the seller's approval.

Ground rules

1. Ask about employer data first. If the seller works for an employer, ask whether they may use outside AI tools with customer or pipeline data. If the answer is no or unsure, work from public information only and do not ask for CRM exports. See references/seller-guardrails.md.
2. Every fact about an account carries a source and date. No source, no fact.
3. No invented customers, results or product capabilities.
4. Nothing is sent. This skill plans and drafts; the seller approves and sends.

Step 1: Intake, one question at a time

Skip anything the user has already said, and ask only for what blocks the next step. If the seller works for an employer and has not said what data they may use, ask the employer-data question (ground rule 1) first, before anything else. Then get, in this order:

1. Role and who they sell for - their own company or an employer.
2. What they sell - the product, the problem it solves, and two or three things it is genuinely good at. A product page link helps.
3. Who buys - three to five best customers or recent wins, or the buyer they believe fits if there are none yet.
4. Motion - do they choose their accounts, were they given a named list, or are they growing existing customers?
5. Channels and tools - email, social, phone, events; which CRM or tracker.

Do not block on the account list. If the seller has a territory or a book but has not pasted the accounts yet, write the motion now - motion type, ICP or fit rubric, signals, tier rules, personas and the weekly routine - and leave the tiered list as the first task of week one, saying exactly what to paste in to fill it.

Step 2: Choose the motion

Use references/motion-types.md to map the answers to one of new-logo outbound, named territory or expansion. Recommend one. A second motion can come after four weeks of the first.

Step 3: Write the ICP and score fit

Follow references/icp-and-tiering.md: write the ICP from the best customers, add disqualifiers, and set the fit rubric. If there were no customers to learn from, mark the ICP as a hypothesis.

Step 4: Tune the signals

Start from references/signal-scoring.md. Raise the weights of the signals that link most directly to the problem the product solves, and drop any that never apply. Keep the freshness bands.

Step 5: Tier the accounts

Whatever the account count, keep tier 1 to ten accounts or fewer. If accounts were given, research each briefly (website, recent news - use web search if available), score fit and timing, and place each in a tier. Cap tier 1 at ten. If no accounts were given and the motion is outbound, write the search brief and hand to signal-prospecting.

Step 6: Personas, plays and routine

• Name the buying roles for this product and the angle each cares about (the role table in references/icp-and-tiering.md).
• Set the play shape per tier: bespoke for tier 1 (use account-play), light for tier 2.
• Set the weekly routine from references/operating-rhythm.md.

Step 7: Hand back

Show a one-screen summary: the motion and why, the ICP, the top accounts with one line each, and the first week's plan. Write the full motion to gtm-motion.md. Offer the next step: run a signal sweep now, or build a play for the top account.

Example

An enterprise account executive at a product analytics company has 30 named accounts.

• Motion: named territory - the employer chose the accounts, effort per account is hours, and the constraint is getting to the right people.
• ICP: consumer or B2B software companies with a product team of 20 or more, a data warehouse in place and a growth or experimentation function. Disqualifier: no in-house product team.
• Signals weighted up: a new head of product or growth (3), job posts for product analysts or experimentation roles (3), a warehouse or data-stack change (2), a public push into self-serve or product-led growth (2).
• Employer data: not allowed in outside tools, so the motion runs on public information and the account names the AE typed in.
• Week one: sweep all 30, tier 1 is six accounts, plays built for the top two.

For an expansion motion the shape is the same but the signals come from the customer: usage changes, a new team adopting, a limit hit, a renewal coming up, a new leader or a champion leaving, and questions about an add-on. Cold prospecting sequences are never sent to existing customers.

Configuration

• Search access: web search or fetch makes Step 5 and the weekly sweep far stronger. Without it, the skill works from what the user pastes.
• Working folder: where gtm-motion.md is saved so later runs start from it.