Ronnie Fieg on Why Kith Deliberately Underprices Its Products
Ronnie Fieg started working in a Queens stockroom at age 13, learning product construction and retail floor dynamics from the ground up before founding Kith.
Theme
Pricing: what the top podcasts are saying.
How operators set, test, and raise prices. 15 write-ups from 10 shows so far, the newest from September 2026.
15 write-ups10 shows
The short version
Founders test extreme premium pricing across consumer software and enterprise sales to escape commodity rates. In the AI market, aggressive discounting threatens raw token costs, and buyers increasingly evaluate models on the total expense per completed workflow.
Top talking points
High prices filter for trust and reveal demand
Founders position software as a premium service through high subscription fees. Anish Acharya notes that testing tiers up to $10,000 reveals true consumer demand. Mentava charges a $500 monthly fee to signal immediate quality to parents evaluating educational tools.
- Why Mentava Charges $500 a Month to Teach Toddlers to Read The Tim Ferriss Show · Sep 13
- Price Consumer Software Like a Birkin Bag: Anish Acharya on Ambition Lenny's Podcast · Sep 6
AI pricing shifts from tokens to completed tasks
Raw token charts mislead buyers evaluating modern language models. Steven Heidel observes that reasoning systems use fewer prompts to finish workflows, making GPT-6 Astra cheaper per finished task. Startups like Lovable route requests dynamically to optimize computing expenses.
- Why Cost Per Task Replaces Per-Token Pricing in AI TBPN · Sep 6
- GPT-6 Astra: $50 Output Tokens, Computer Use, and New Benchmarks How I AI · Sep 6
- Lovable Hit $500M: Their AI Mistake-Learning Loop & Token Pricing All-In Podcast · Jul 19
Price wars and hardware limits define frontier AI
XAI challenges competitors with disruptive rates, and Bill Maris states Google could cut token prices by 80%. Rising component costs constrain these drops. Memory bottlenecks force providers to charge five times more for output tokens than input data.
- Grock 4.6's Disruptive Pricing Puts OpenAI on Notice All-In Podcast · Aug 16
- Google's 80% Token Cut: The AI Market Killer Threat All-In Podcast · Jun 14
- LLM API Costs: Context Length Reveals Memory Bottlenecks Dwarkesh Podcast · Apr 26
Custom manufacturing and win rates signal pricing power
Pat LaFrieda Jr. and Ronnie Fieg maintain premium margins by creating custom products and rejecting wholesale channels. Enterprise software teams monitor win rates closely. Closing more than 35% of deals serves as direct evidence of underpricing.
- Sell a $28 Burger? Pat LaFrieda's Method for Premium Pricing My First Million · Jun 7
- Ronnie Fieg on Why Kith Deliberately Underprices Its Products Founders Podcast · Sep 27
- Why High Win Rates Mean Your Enterprise Price Is Too Low Lenny's Podcast · Aug 23
Most interesting insights
High software pricing acts as an immediate trust filter for skeptical parents who assume cheap educational apps are low-quality scams.
From Why Mentava Charges $500 a Month to Teach Toddlers to Read, The Tim Ferriss Show · Sep 13
Raw token pricing charts are misleading: a model like 3.8 Flash looks 13 times cheaper per token than Astra, but Astra costs less per finished task.
From Why Cost Per Task Replaces Per-Token Pricing in AI, TBPN · Sep 6
Healthy enterprise sales win rates sit between 30% and 35%; anything higher means you are leaving money on the table.
From Why High Win Rates Mean Your Enterprise Price Is Too Low, Lenny's Podcast · Aug 23
Latest write-ups
Why Mentava Charges $500 a Month to Teach Toddlers to Read
Mentava charges $500 per month for early literacy software, replacing $100-an-hour private tutors for toddlers and preschoolers.
Price Consumer Software Like a Birkin Bag: Anish Acharya on Ambition
Venture capital has inverted its core filter: three years ago Andreessen Horowitz passed on pitches that looked too complex, but today general partner Anish Acharya turns down ideas that are too small.
GPT-6 Astra: $50 Output Tokens, Computer Use, and New Benchmarks
OpenAI priced GPT-6 Astra at $10 per million input tokens and $50 per million output tokens, targeting high-value enterprise execution over cheap conversational chat.
Why Cost Per Task Replaces Per-Token Pricing in AI
Academic benchmarks saturate at 99% to 100% accuracy within weeks of release, forcing builders back to subjective vibe checks.
Moderna's $500K Cancer Shot and Regulatory Capture
Moderna's market valuation tripled from $20 billion to $60 billion after positive trial readouts for its personalized melanoma mRNA therapy.
Why High Win Rates Mean Your Enterprise Price Is Too Low
Healthy enterprise sales win rates sit between 30% and 35%; anything higher means you are leaving money on the table.
Alec's Edge: Pricing Uncertainty, Not Just Risk
Investor Alec distinguishes between 'risk,' where probabilities are known, and 'uncertainty,' where no existing model applies, arguing that true market edge lies in resolving the latter.
H1 2026 PE: Why 12.5x EBITDA for 'A-Assets' Only
US private equity buyout multiples held firm at 12.5x EBITDA in H1 2026, matching 2021 peaks, despite a drop in deal volume and tighter debt markets.
Grock 4.6's Disruptive Pricing Puts OpenAI on Notice
Elon Musk's XAI is challenging the OpenAI and Anthropic duopoly in frontier AI models with Grock 4.6, offering "disruptive pricing" while maintaining high quality.
Lovable Hit $500M: Their AI Mistake-Learning Loop & Token Pricing
Lovable, led by Oika, scaled to $500 million in just 20 months, demonstrating that hypergrowth is still achievable with a precise AI strategy.
Apple Hikes Prices: Memory Shortages Shift Billions
Apple has significantly increased prices on numerous products, including MacBooks, iPads, and Apple TV.