Skip to content
Helix GTM tools
Claude plugin · 3 skills

Which GTM motion should lead, how should you position, and why are your prompts getting generic answers?

GTM Skills is a free Claude plugin with three skills built on Shashwat Ghosh's own frameworks. EPIC picks the go-to-market motion to lead with, IMPACT finds the market you can own and the message to go with it, and CRAFT rewrites a prompt so Claude stops giving generic answers.

What does each skill do, and who is it for?

SkillWhat it doesWho it is for
EPIC Motion Diagnostic
epic-motion-diagnostic
Scores the four EPIC motions (Ecosystem and ABM, Product-Led, Inbound and Outbound, Community) for your stage, deal cycle, NRR and region, then names the motion to lead with and what to do first.Founders, CMOs and GTM leaders choosing a motion before choosing channels.
IMPACT Quick Positioning
impact-quick-positioning
Finds the market you can own (sub-category, category or new category) and builds a three-level message hierarchy in the buyer's words, using IMPACT's Anchor and Craft steps.Founders and product marketers whose product sounds like everyone else's.
CRAFT Context Engineering
craft-context-engineering
Checks a prompt against the CRAFT framework, names the missing context and rewrites it.Anyone whose AI output keeps coming back vague.

Who is it not ideal for?

How do I install it?

Claude Code

/plugin marketplace add shashwatgtm/gtm-claude-skills
/plugin install gtm-skills@gtm-claude-skills

Or from a terminal: claude plugin marketplace add shashwatgtm/gtm-claude-skills then claude plugin install gtm-skills@gtm-claude-skills.

Cowork

  1. Open Customize in the sidebar, then Plugins.
  2. Select Add marketplace and enter shashwatgtm/gtm-claude-skills.
  3. Find the plugin in the list and select Install. Its skills apply when relevant, or type / to run one.

What does it produce? Real outputs from a test run

On 25 September 2026 each skill was run once on the prompt shown, in a fresh Claude Code session with the plugin installed from GitHub. The outputs below are shown as produced, with these edits only: dashes are written out in words to match this site's style, paths to files saved on the test computer are replaced, where an output quotes this site's old byline, the author's title is updated to his current one (Co-Founder and Fractional CMO), one link a web search found to a private person's LinkedIn profile is removed (the text stays), and a placeholder attendee name that the search matched to a real person is shown as "Dana Example". Company and product names are placeholders.

EPIC Motion Diagnostic: real output

Prompt used in the test run (Claude Code 2.1.281, fresh session, plugin installed from GitHub; skills Claude used: gtm-skills:epic-motion-diagnostic):

We are a Series B logistics SaaS company selling route-planning software to enterprise supply chain teams in India and the Middle East. ARR is $2M, NRR is 120%, the sales cycle is about 90 days, ACV is around $45,000, and 80% of deals come from outbound. There are roughly 3,000 target accounts. Which GTM motion should lead, and what should we do first?

Short answer: Outbound should lead, run as account-based outbound. Your inputs put Outbound only one point ahead of Ecosystem/ABM, and two of your numbers sit exactly on scoring thresholds, so check those before you commit budget. In the Middle East, a partner-led (Ecosystem) motion should lead.

EPIC scores (India + Middle East combined)

StepEPIC
Series B starting scores7566
ACV $45K (in the $5K, $50K band, no change)7566
Sales cycle about 90 days (on the threshold, no change)7566
NRR 120% (on the threshold, no change)7566
3,000 target accounts (in the 500 to 10K band, no change)7566
Logistics cap: Product-Led can't score above 47466
80% of deals from outbound: Outbound +27486
Two regions: no geography change (see the split below)7486

E (Ecosystem/ABM): 7 | P (Product-Led): 4 | I (Inbound/Outbound): 8 | C (Community): 6

Scored per region, the result changes:

  • India (E +1, C +1): E 8, I 8, C 7. Ecosystem and Outbound tie.
  • Middle East (E +2): E 9, I 8. Ecosystem leads.

Primary motion: Outbound. It's where 80% of your deals come from, and 3,000 accounts is enough for structured outbound. But at $45K ACV with 90-day cycles, cold volume won't work. It has to be multi-threaded and account-based.

Secondary motion: Ecosystem/ABM. It's close behind overall and it leads in the Middle East. There, relationships and partners (system integrators, ERP/TMS resellers, 3PL alliances) will likely open more doors than cold sequences.

Critical warning: the lead depends on two numbers.

  • If your median sales cycle is over 90 days, Ecosystem goes to 9 and Outbound drops to 7, so Ecosystem leads outright.
  • If your NRR is over 120%, Community rises to 7.

"About 90 days" and "120%" are the two values where the answer flips, so get the exact figures first.

No other warnings apply. NRR is above 100%, so there's no retention problem to fix before adding pipeline. Product-Led is capped for logistics, so a product-led land plus ABM expand (hybrid) doesn't apply.

Three actions for Monday morning

  1. Get the two numbers that decide the lead. From your CRM, take the last four quarters of closed-won deals. Pull the median days from first touch to close and the trailing NRR, both split by India and the Middle East. If the cycle is over 90 days, move budget toward ABM before scaling SDR headcount.
  2. Tier the 3,000 accounts instead of working them all the same way. Pick roughly the top 10% by fit and intent signals. For those accounts, map each buying committee (Head of Supply Chain, Logistics/Transport lead, IT, Procurement) and reach several people per account at once. Run lighter sequences on the rest. Outbound spread evenly across 3,000 accounts with 90-day cycles uses up your list and your reputation fast.
  3. Use the 120% NRR as sales proof. Find which customers expanded and why. Get 5 to 10 of them to agree to be references, and put their stories into outbound sequences and late-stage deals. In India and the Middle East, buyers trust peers, and this is your cheapest way into Community.

What to stop doing: Running the same SDR playbook in the Middle East as in India. Sign one or two partners there and give them joint pipeline targets. Also avoid any free trial or self-serve effort. Logistics enterprise buyers rarely buy that way.

Assumptions (correct any of these and I'll rescore):

  • Assumption: "About 90 days" is 90 days or less. If the median is higher, Ecosystem leads.
  • Assumption: NRR is exactly 120%. If it's higher, Community goes up by 1.
  • Assumption: I treated India + Middle East as two regions, so the overall score has no geography adjustment. If one region brings most of your revenue, use that region's scores.
  • Assumption: Outbound is the only channel you run at scale today. You didn't give team size or current channels, so "what to stop doing" may need adjusting. [User to add: GTM team size and current channel mix]
  • Assumption: There's no self-serve version of the product.

These scores will change as the company grows, so rerun this in six months or after the CRM numbers come in. I can also turn this into a one-page summary for your leadership team.

IMPACT Quick Positioning: real output

Prompt used in the test run (Claude Code 2.1.281, fresh session, plugin installed from GitHub; skills Claude used: gtm-skills:impact-quick-positioning):

We sell procurement automation to mid-market manufacturers in India (200 to 2,000 employees). We keep losing to a larger procurement suite in competitive deals, and buyers say we sound like everyone else. Our strengths: setup in 3 weeks instead of 6 months, and it works with the ERP they already have. Help me position.

Why you sound like everyone else

"Fast implementation" and "integrates with your ERP" are on every procurement vendor's homepage, the large suite's included. Buyers treat both as claims no one can check, so they fall back on the safe choice: the big brand.

There's a second problem. "3 weeks vs 6 months" can make your product sound small. To a plant head or CFO, a 6-month project sounds thorough and a 3-week one sounds like a light tool. Speed only helps you if you say what it buys them.

So the fix isn't new adjectives. Stop saying you're a faster, lighter version of the suite. Say that the suite is a large project, and you're a procurement layer on the ERP they already run. That way you aren't claiming to be better than the suite. You're offering a different kind of purchase.

Step 1: Anchor

  • Option A, own "procurement software / P2P suite" head-on: Not viable. That's where the suite is strongest (brand, breadth, analyst coverage). You'd be comparing feature lists against a bigger list.
  • Option C, invent a category (e.g. "Procurement Intelligence"): No one searches for it, and it would take 18 to 24 months of educating buyers. Rejected.
  • Option B, sub-category (recommended):

Recommended category: Procurement automation for mid-market manufacturers that runs on the ERP they already have
Positioning type: Sub-category
Why this anchor: Buyers already search for "procurement automation", so the category is findable. The qualifier ("on your existing ERP, for manufacturers") turns your two strengths into what defines the category, where right now they're just claims. It also changes what the deal is judged on. The buyer stops asking "which suite has more modules?" and starts asking "do we want a 6-month transformation project, or control of purchasing this quarter?" At 200 to 2,000 employees, many buyers have no procurement transformation team to run the first option.
Risk: If a buyer actually needs sourcing, contracts and supplier management all in one place (the full suite scope), this position tells them you're not that, and you'll lose those deals openly. That's acceptable if those deals are a minority. It's a problem if they're most of your pipeline.
Test question for 5 prospects: "Would you rather have a full procurement suite in 6+ months, or purchase control running on your current ERP within 3 weeks? What would make you pick each?"

Step 2: Message hierarchy

LEVEL 1: Core message (9 words)
Control plant spend this quarter, without replacing your ERP.

Alternate: Procurement control on your existing ERP, live in three weeks.

LEVEL 2: Elevator pitch (30 seconds)
[Company] is procurement automation for Indian manufacturers with 200 to 2,000 employees, built to run on the ERP you already have. Most procurement suites start with a 6-month project. You change processes, retrain people, and keep buying on email and Excel until go-live. We connect to [User to add: ERPs you support, e.g. SAP B1 / Tally / Dynamics] and have your team raising and approving POs through us in three weeks. [User to add: named customer + result, e.g. "X went live in 19 days and now routes all indirect spend through approvals"].

LEVEL 3: Full narrative (2 minutes, ~250 words)

Problem: In a mid-sized manufacturing company, purchasing runs on the ERP plus email, WhatsApp and spreadsheets. Indents come from the shop floor, approvals are chased by phone, and the CFO sees what was spent only after the invoice arrives. Everyone knows it's leaky. The usual fix, a full procurement suite, is a transformation project that most 500-person manufacturers don't have the team to run.

Why now: [User to add: a trigger you can back up. Options: raw material price volatility making mid-quarter visibility urgent; promoter or PE pressure on working capital; ERP upgrade cycles creating a window. Pick only what your customers actually say.]

Solution: We don't replace your ERP or ask you to redesign procurement around new software. We sit on top of the ERP you run today. Your masters, vendors and GL codes stay where they are. We add the control layer: indent-to-PO workflows, approval rules, and spend visibility. Because nothing underneath changes, go-live takes three weeks, not two quarters.

Proof: [User to add: number of manufacturing customers, 1 to 2 named logos, actual go-live times, one spend or cycle-time outcome with the customer's permission.]

CTA: Give us one plant and three weeks. If POs aren't flowing through the system by then, you've lost nothing. [User to decide whether you can contractually commit to this.]

Do's and don'ts

  • Do make the speed claim checkable. Say "first PO in week 3" and name the ERPs you support. "Seamless integration" is what everyone says. "Runs on SAP Business One, no customisation" is a claim a buyer can test.
  • Do sell to the CFO or promoter on risk: no replacement project, no gap while you wait for go-live, no consultant fees. Suites win mid-market deals on perceived safety, so show the risk is on their side.
  • Avoid comparing features with the suite. Every time you match their module list, you're competing on their terms. When a buyer asks "do you do X?", answer "Is X something you need in the first year, or something the suite is selling you?"

Assumptions (correct any of these)

  • Assumption: Your customers mostly run mid-market ERPs (SAP B1, Tally, Dynamics, Oracle NetSuite or home-grown) rather than SAP S/4HANA. If they're mostly on S/4, "works with your ERP" is a weaker point of difference and the anchor needs work.
  • Assumption: You lose on perceived breadth and brand safety, not on a specific missing capability. If losses trace back to a real gap (e.g. supplier portal, e-auctions, GST/e-invoicing), the fix is partly product, not messaging.
  • Assumption: The "3 weeks" figure is typical across your customers, not a best case. Put only your median go-live time in front of buyers.
  • I haven't named or researched the suite you lose to, so none of this depends on claims about them.

What would sharpen this

  1. Why you lost your last 3 competitive deals, in the buyer's words if you have them.
  2. Which ERPs your current customers run, and roughly how many customers are manufacturers.
  3. One or two customers who would let you name them, with their actual go-live time.

With those, I can tighten the anchor and fill in the proof. If you name the suite you lose to, I can also build a battle card for your reps: the questions to raise early in a deal and how to answer the "they're the safe choice" objection. I can also turn this into a shareable page for your sales team.

CRAFT Context Engineering: real output

Prompt used in the test run (Claude Code 2.1.281, fresh session, plugin installed from GitHub; skills Claude used: gtm-skills:craft-context-engineering):

I asked Claude to build a competitive analysis but the output was too generic. My prompt was: 'analyze our competitors.' We sell HR software to 50 to 500 person companies in the UK. What am I missing, and can you rewrite the prompt?

Your prompt had three words and no context, so Claude had to guess everything: which competitors, what questions to answer, who would read the result, and what format to use. When Claude has to guess, it gives you the average answer for any company, and that's the generic output you got. Better wording won't fix it. More context will.

What's missing

ElementIn your prompt?What Claude assumed instead
Character (who Claude should think as)MissingA general business analyst with no HR tech or UK experience.
Result (what "done" looks like)Missing"Analyze" had no defined goal. Claude couldn't tell whether you wanted help with pricing, sales, a roadmap, or positioning, so it covered a bit of all four.
Artifact (the output format)MissingProbably headings and bullet points, which nobody can use directly in a deal or a planning meeting.
Frame (your market, rivals and limits)Mostly missingYou gave me the segment (UK, 50 to 500 employees), but that wasn't in the prompt. Claude also didn't know your competitors, your ACV, who you lose to, or which UK requirements buyers care about.
Timeline (how you'll review it)MissingA one-shot answer, with no chance to fix gaps or approve the approach first.

The biggest gap is Result. A competitive analysis for a pricing decision looks very different from one for your sales team, and Claude can't guess which you need.

Rewritten prompt

Fill in the [User to add] fields before you use it. Those fields are what make the output specific to you.

CHARACTER
You are a product marketing lead who has spent 10+ years in UK HR tech,
selling HRIS and people software to mid-market companies (50-500 employees).
You have built competitive programs that sales teams actually use. You judge
competitors on why deals are won and lost, not on feature checklists. You are
skeptical of vendor marketing claims and separate what a competitor says from
what customers report.

RESULT
Produce a competitive analysis that helps us [User to add: the decision this
supports, e.g. "win more head-to-head deals", "set 2027 pricing", "decide
roadmap priorities for Q1"]. It must include:
1. A landscape map of the 4-6 competitors we meet most in UK mid-market deals,
   grouped by type (UK-native HR platforms, global HRIS moving down-market,
   payroll-led vendors adding HR, all-in-one SMB tools moving up-market).
2. For each competitor: target segment, pricing model (per employee per
   month, modules, minimums, as far as public info allows), core strengths,
   known weaknesses, and who at the buyer they usually sell to.
3. How each one handles the UK requirements our buyers weigh: UK payroll and
   HMRC RTI integration, pension auto-enrolment, statutory leave and pay
   (SSP, SMP), right-to-work checks, UK GDPR and data residency.
4. Where we win, where we lose, and why, against each one.
5. The 3 biggest threats to our position in the next 12 months, and 3 gaps
   we could exploit.
Out of scope: enterprise vendors that only sell to 1,000+ employee companies,
unless they are actively moving into our segment.

ARTIFACT
- A comparison table first, with columns: Competitor | Type | Target segment |
  Pricing model | Main strength | Main weakness | Threat level (High/Med/Low).
- Then one section per competitor, max 250 words each, in short paragraphs.
- End with a one-page summary for [User to add: audience, e.g. "CEO and
  Head of Sales"].
- Cite a source for every factual claim about a competitor (pricing, features,
  funding, acquisitions). Use official sites, G2/Capterra, or reputable press
  only. Mark anything you cannot verify as [Unverified].

FRAME
About us:
- Product: [User to add: what your HR software does, e.g. core HRIS +
  absence + performance; do you do payroll?]
- ICP: UK companies with 50-500 employees. Buyer: [User to add: e.g. HR
  Manager / Head of People, with FD or MD signing off]
- ACV / pricing: [User to add]
- Our real differentiators: [User to add: 2-3, honestly stated]
- Our known weaknesses: [User to add: where we genuinely fall short]

Competitive context:
- Competitors we meet most often: [User to add: names]
- Who we lose to most, and the reasons given: [User to add]
- Who we beat most, and why: [User to add]
- Win/loss notes or CRM data: [User to add: paste if available]

Constraints:
- UK market only. Ignore US-only features and pricing.
- Tone: direct, evidence-based, no marketing language.
- Do not invent pricing, customer counts, or market-share figures. If public
  data is not available, say so.

TIMELINE
1. Before starting, list any gaps in this brief that would change your
   analysis, and ask me about them.
2. Propose the list of competitors to cover and your outline. Wait for my
   approval.
3. Draft the full analysis. I'll mark each section KEEP, REVISE, or REMOVE,
   and you revise once.

Before and after

Your original prompt could have been written by any HR software company anywhere, so it got a textbook overview. The rewrite ties the analysis to one decision, one UK mid-market buyer, the rivals you actually meet, and the UK-specific requirements that decide deals in your segment. It also asks for sources and makes Claude flag anything it can't verify. That matters because made-up competitor pricing is a common problem in AI competitive analysis.

If you're short on time: at minimum, fill in the decision the analysis supports, your competitor names, and who you lose to and why. Those three fields do most of the work.

Assumption: You sell core HR software, not payroll-only or recruitment-only. If payroll is your main product, the landscape groupings and the UK requirements section should change. Tell me and I'll adjust them.

Assumption: You want a strategic analysis. If what you really need is something reps can use against one named competitor in a live deal, ask for a battle card instead. That's a narrower prompt, and I can write it.

Licence, source and privacy

All connectors and plugins