What this episode is about
Wes Bush (founder of ProductLed) and Aakash Gupta (the AI PM, former PM at Apollo, Affirm, and Epic Games, followed by close to 300,000 tech people) spent an hour mapping out exactly how AI is changing product-led growth.
Wes has worked with 400+ SaaS companies to generate over $1B in self-serve revenue. His book "The Product-Led Playbook" is the complete system for accelerating growth with a lean team.
In this episode, Wes and Aakash took that foundation and mapped out how AI transforms each layer of PLG.
Here's what’s covered:
Your AI Stack for PLG
Core LLMs:
- Claude Projects (your trained copilot)
- NotebookLM (sole context, no hallucinations)
- Dictation tools that 3x your prompting speed
Agent Platforms:
- Lindy (S-tier for most people)
- n8n (most powerful, steeper learning curve)
- Relay (middle ground, great for non-technical GTM operators)
Use cases: competitive analysis agents, AI EAs that replace $1,200/month assistants, 40-person marketing teams automated
Prototyping:
- Bolt (my pick - built my entire latest site on it)
- Lovable (hit $120M ARR in under a year)
- Base44 (one guy, $80M exit, all vibe coded)
How AI Changes Every PLG Layer
Activation: Your time-to-value needs to be 1/10th of what it is now.
Example: Apollo could analyze your domain, find 10 perfect contacts, and draft outreach emails before you even search. That's AI-assisted activation.
You now need a ladder: AI-assisted activation → user-driven activation → habit.
Pricing: AI has COGS. Your pricing needs to adapt:
- Best: Outcome-based (Fin, Harvey AI, 11x)
- Good: Usage-based
- Workable: Seats with overages + multiple plans
Cursor gives you a free IDE with unlimited usage if you bring your own Claude API key. Genius.
Model: Reverse trials are the new normal. Give premium AI features upfront, then convert when credits run out.
Free trial with qualification (business email) is becoming standard to manage COGS.
Expansion: AI identifies which users at an account aren't fully activated, finds relevant use cases for each person, and crafts personalized outreach.
Example: Target senior PMs who haven't adopted key workflows, send them a customer insight that matches their role.
GTM: You can do 10x more marketing now. Teams are:
- Using Veo3 and Nano Banana to create viral ads
- Running Outrank for daily SEO articles
- Automating employee LinkedIn at scale
- Creating AI avatars for outbound (yes, they work)
Data & Team:
- Prompt-based experimentation (prototype → experiment → live in minutes)
- Bandit algorithms that shift traffic to winners in real-time
- Teams are getting 50% smaller, 10x more productive
The Real Differentiator
Features are now commoditized. Anyone can vibe code your core product (or close to it) in Bolt.
Taste is the new moat.
ChatGPT shows you a blank screen with a text box. That's taste.
Linear has zero tolerance for bugs. They solve your bug the same day even if you're the only person experiencing it. That's taste.
Taste = only showing people what they need to see, when they need to see it. Plus obsessive attention to eliminating every rough edge.
Where To Find Aakash
📧 Subscribe to his Product Growth, the ultimate AI PM newsletter
Resources from ProductLed
🚀 Follow Wes on LinkedIn
📘 Get the ProductLed Playbook for Free
🎯 Find what’s broken in your ProductLed motion. Book a PLG Diagnostic.
🧩 Subscribe to the ProductLed Newsletter – Join 43k+ founders, PMs, and operators learning how to scale smarter with product-led growth.
One Action for You This Wee
Pick one tool from this list you haven't tried yet:
- Week 1: Superwhisper (dictation)
- Week 2: Lindy (agents)
- Week 3: Claude Projects (copilot)
- Week 4: Bolt (prototyping)
- Week 5: NotebookLM (sole context)
Don't try to do everything. Just add one tool per week to your stack.
The arbitrage opportunity in AI + PLG is massive right now. But it won't last forever.
Listen to the full episode 👇We go way deeper on each tool, show live demos, and cover specifics I couldn't fit here.
