Issue No. 40Week ending Sunday, October 4, 2026485 episodes · 2075 articles
The Throughline ↓
The Podcast Summary.

10+ hours of podcasts, in 5 minutes.

Theme

Unit economics and margins: what the top podcasts are saying.

Where the money is made or lost inside a business. 11 write-ups from 8 shows so far, the newest from September 2026.

11 write-ups8 shows

The short version

Heavy token consumption routinely breaks traditional software gross margins, pushing AI startups to train in-house models and route prompts to restore profitability. Outside of software, operators focus on tracking true customer acquisition costs to prevent automated marketing spend from eroding capital efficiency.

Top talking points

  1. Heavy AI token usage collapses software margins

    Reasoning agents create direct expenses from user actions. Harvey watched gross margins drop to negative 50% after token volumes jumped 20-fold. Aaron Katz asserts AI companies with fast top-line growth evade traditional software margin expectations entirely.

  2. In-house models and prompt routing restore profitability

    Companies repair unit economics by owning their infrastructure. Canva cut inference costs 90% by training in-house models, and Harvey restored positive margins using model routing. The 20VC podcast counters that routing tokens burns engineering hours better spent on distribution.

  3. Unmeasured marketing spend obscures true acquisition costs

    Marketing platforms double-count sales, typically reporting customer acquisition costs 20% below reality. Many middle-market executives spend $500,000 monthly without tracking returns. Meanwhile, IM8 automated its Meta targeting to scale spend to $33 million while dropping acquisition costs to $239.

Most interesting insights

At an estimated $400 million ARR run rate ($33 million monthly revenue), a negative 50% gross margin created a burn of roughly $16 million per month purely on inference costs.

From Why Agentic AI Broke Harvey's Gross Margins, TBPN · Sep 27

Canva slashed operational inference costs by 90% by acquiring Leonardo and training proprietary image generation models in-house.

From Canva Cut AI Inference Costs 90% by Ditching API Wrappers, Cheeky Pint · Aug 30

Scaling spend from $17 million to $33 million in one quarter lowered IM8's customer acquisition cost from $305 to $301, and down to $239 the following month.

From How IM8 Scaled Meta Ads to $33M a Quarter While Lowering CAC, My First Million · Sep 13

Latest write-ups

Shows covering this

The Sunday Email

Get next Sunday's issue in your inbox.

10+ hours of podcasts, distilled into one 5-minute read. Free, every Sunday morning.

Newsletters

One email a week. Unsubscribe with one click.