Growth Unhinged is proudly supported by Cleverbridge
The shift to digital buying is happening whether you're ready or not. 67% of software buyers want to purchase without talking to a sales rep. I suspect 100% of AI agent buyers prefer a self-serve path, too 😉
PLG and AI-native companies have been ahead of the curve. Now the shift is going upmarket. Enterprise software companies like PTC ($2B+ ARR) are moving renewals and add-ons into digital buying, alongside higher-touch sales where it adds the most value.
Cleverbridge just launched a new content series on this shift. It explores how to improve margins, increase efficiency, and build a foundation for agentic commerce.
McKinsey-caliber research is the first Claude Cowork use case that made me a believer.
Creating my GTM research skill, or a curated set of instructions and resources for Claude, became the foundation for every more advanced GTM workflow that I stacked on top. Great research led to better content insights, more relevant campaigns, more accurate competitor analysis, and fresher pricing insights.
But I didn’t spend weeks painstakingly making the Claude skill. I simply pointed Claude Cowork to this GTM deep research resource and the skill effectively built itself. Then I was free to run the skill whenever I wanted.
It suddenly hit me: what if five years of newsletter editions was just a set of Claude skills for GTM and pricing analysis?
Instead of reading a 2,500 word newsletter, you could plug the frameworks into any Claude product and immediately apply them. These skills could become living assets, too, getting smarter over time.
I’ve built more than a dozen Claude skills for GTM and pricing analysis. Now I’m sharing some of my favorites with you, and will teach you how to make your own. These skills are a v1 — if you try them, please let me know what you think.
How to build and use GTM skills in Claude
I asked Claude to create a [NAME].md skill using the skill-creator tool, which is built-in for Claude Cowork and Chat, and pointed it to the best source materials. Claude did all the heavy lifting from there. Here’s a prompt you can use to do it yourself:
Create a Claude skill called /deep-research-gtm. It should be a SKILL.md file with a description that tells Claude to reference this skill whenever I ask for research that involves synthesizing multiple sources into a decision-ready output. This skill is for reports worth spending 10-30 minutes on, not one-off searches. Use the frameworks and methodology mentioned in [insert source file].Running the skill files is pretty straightforward, too. I believe you can even use them on the free version of Claude. Here are step-by-step instructions:
Step 1: Download the skill files below. They’re free for readers.
Step 2: Upload them to Claude.
Install it in Claude Chat or the desktop app — Go to Customize → Skills → Create skill → Upload a skill → Attach the files above.
Step 3: Start using them.
You can directly reference the skill in the prompt bar by tagging the file name or Claude will suggest the skill if it’s relevant for your prompt.
Step 4: Make the skills yours.
The magic of skills is customizing them to your business context and personal preferences.
When you give Claude feedback on the outputs, you can also ask it to edit the skill file accordingly.
/deep-gtm-research skill for GTM research
Use case
When you need analyst quality research that helps you make more informed GTM decisions and cites its sources.
Why it matters
AI can be fantastic at research, but makes frustrating intern-level mistakes. It’ll sometimes over-index on one blog post, treat a Reddit thread as data, or waste all your tokens on research that doesn’t answer your core questions.
What it does
Runs a structured workflow: gathers context first, proposes a research plan for your sign-off, prioritizes primary sources, and delivers a pyramid-structured report with in-text citations and a source table. It essentially follows the same workflow of a strategy consultant or BizOps manager, as outlined in Torsten Wallbaum’s fantastic guide.
Pro-tip
You don’t need to tag /deep-research-gtm to use it, although you certainly can if you want. When you ask for things like a deep dive, competitive analysis, landing page teardown, or in-house playbook, the skill automatically gets suggested. It’ll ask a few follow-up questions if the context is thin.
I’ll reference the deep research skill when I’m doing research for Growth Unhinged or onboarding with a new advising client. The potential uses are endless: you could analyze which ads competitors are running, audit a homepage, compare competitors’ product features, assess markets for international expansion, or estimate the market size for a new product.
As a test case, I used the skill to analyze the pricing models of all 50 companies in the new Forbes 2026 AI 50 list. Here’s the prompt I used to run the AI 50 pricing research:
Please go through the new 2026 AI 50 list and use the
/deep-research-gtm skill to analyze how these companies approach pricing and packaging including whether they offer free plans, whether they have public pricing, and what is their main pricing model.It did a pretty solid job — you can check out the report it created. Claude found that 52% offer a free plan, 56% publish pricing (the ones that didn’t were in legal, healthcare, robotics, and enterprise infra), and 10% don’t actually have a commercial product yet. This type of research used to take me a couple of hours to do myself. Claude does it in a few minutes (and probably better than me!).
📄 Get the skill: Deep GTM research skill
📚 Sample output: Forbes 2026 AI 50 pricing report
🔎 Reference: How to use Deep Research for GTM
/pricing-teardown skill for pricing analysis
Use case
When you want outside feedback on how to optimize your pricing page for both human buyers and (increasingly) AI agents.
Why it matters
Pricing pages were already important for human buyers; I believe they’re the second most important page on your website behind the homepage. Many pricing pages were already overdue for a refresh, and AI makes this more urgent. LLMs are increasingly recommending tools to buyers, and opaque pricing means you could be invisible or misrepresented.
What it does
Evaluates your pricing page across 10 dimensions — 7 for the human buyer experience (value prop, plan clarity, cognitive load, trust, behavioral psychology, transparency) and 3 for AI agent readiness (machine-readable pricing, FAQ coverage, per-tier depth). Each dimension gets a 1-4 score with a specific finding from your actual page, an overall letter grade, and a prioritized set of quick wins vs. strategic improvements.
Pro-tip
For the most reliable results, run the skill via Claude Chat. This has better browsing scraping capabilities than Cowork, especially if you have a PC and can’t use the Chrome Connector 😢
I’ll reference the pricing teardown skill ahead of a call with a potential pricing consulting client. It helps me get a sense for the company’s pricing maturity and usually gives me a couple of quick wins I can sprinkle into the conversation.
I tested and refined the skill using Linear as an example. Linear got a D – Claude might’ve woken up on the wrong side of the GPU – with the following recommendations.
Its verdict: Exceptionally clean and low-friction for human buyers, but dangerously thin on value framing and AI agent readability — the page looks great and says almost nothing.
In v1 of the /pricing-teardown skill, Claude deducted points for lack of social proof. This was a miss – Linear highlights fantastic customer logos and a great quote from OpenAI. I suspect the miss was because LLMs like Claude famously struggle with images so they get dropped in favor of text. This bug should be fixed in the latest version. The magic of Claude skills is in incrementally improving them as you go.
📄 Get the skill: Pricing teardown skill
📚 Sample output: Linear pricing teardown report