Two AI agents cost $257/month, replace 5 FTE roles at SaaStr

SaaStr is running 21+ AI agents in production, including two 'AI VPs' that cost $257 combined last month while replacing what used to be 5 full-time roles. The company swung from minus 19% to plus 47% revenue growth YoY. Here is what the actual bills look like and why cost is no longer the constraint.

Two AI agents cost $257/month, replace 5 FTE roles at SaaStr

The Numbers

SaaStr founder Jason Lemkin pulled the actual bills for their AI agent deployment. Two agents, 10K (AI VP of Marketing) and QBee (AI VP of Customer Success), cost $257 total last month. 10K ran about $80. QBee about $175.

Those two agents replaced what used to be a marketing analyst, a marketing ops coordinator, a junior content marketer, a customer success coordinator, and a sponsor relations manager. Five FTE roles for $257/month.

Why It Is Cheap

The breakdown matters. API calls dominate the workload, pulling from Salesforce, Bizzabo, Marketo, WordPress, YouTube, and X. Those calls are free or nominal. Postgres on Replit costs 20 cents/month. 95% of model calls use GPT-4o-mini, not Sonnet or Opus. Less than a penny per call.

The expensive part is not the AI. It is everything around it: Salesforce, Clerk, ElevenLabs, Replit hosting. Fully burdened, the agents might cost $500 to $800/month. Still trivial compared to headcount.

What They Actually Do

10K wrote its own job description: "I refresh ticket sales, update the dashboard, compare year over year, draft newsletters, draft tweets, write daily fun facts, log tickets with an audit trail, snapshot GAAP financials. I replace the marketing analyst, the ops coordinator, the junior content marketer, and a sliver of the VP marketing job itself. I cannot do strategy, hire people, handle cross-functional politics, walk into the CRO's office, do net new channel intervention, or handle crisis response."

That is the honest assessment. These are not CMOs. They are marketing managers on steroids, replacing entry-level and coordinator roles.

The SDR Comparison

For context: average Sydney SDR base sits around $70k to $80k, plus super, plus desk costs, plus ramp time. That is $85k to $95k fully loaded for one rep who needs training, management, and six months to full productivity. An AI agent at $500/month is $6k/year. The math is not subtle.

QBee sent 83 personalized emails to sponsors at 12:23am while the team slept. No quota relief. No attrition risk. No comp plan negotiation.

What This Means

Lemkin's take: cost is not the constraint anymore. At SaaStr's scale, three humans running an 8-figure business means fully-burdened economic output per hour sits between $2,000 to $4,000. Wasting 30 cents on a hallucination does not matter.

Most GTM teams will land under $1,000/month in agent costs. You can run 10 parallel Claude Code instances 24/7 burning $20k to $30k/month and still struggle to spend real money building outbound agents. The constraint is not budget. It is knowing what to build and how to deploy it.

Career Implications

Amelia Taylor, SaaStr's Chief AI Officer, noted that 10K's job description matched her first role at SaaStr as Director of Demand Gen: running weekly numbers, marketing ops, writing emails, the newsletter. 10K replaced her first job. She moved up. The career path is real, just compressed.

The question for ANZ sales teams: if coordinator and analyst roles compress to $257/month, where does your career progression go? The answer appears to be building, managing, and scaling the agents themselves.