How SaaStr Built an AI VP of Marketing That Actually Plans Campaigns
SaaStr runs 20 AI agents across its business. When the company looked for an AI tool to handle marketing strategy and planning, every option it found did the same thing: write content.
Blog posts. Social captions. Email copy. That wasn't the problem. SaaStr has 5,000 pieces of content spanning 13 years - 20 million words. The bottleneck was something else entirely: knowing what to do, when to do it, and executing a coherent plan across every channel every single day.
No AI marketing agent on the market could do that. So SaaStr built its own.
What the AI VP Actually Does
The tool, called 10K, runs daily. It analyzes five years of historical data from SaaStr's systems, pulls real-time inputs from vendor APIs, and connects directly to Salesforce to track pipeline and predict sponsor outreach timing.
Each day, 10K plans weekly and quarterly marketing activities, builds campaigns with specific offers and messaging, identifies which team members should do what, and auto-generates a calendar to execute against. It updates in real time.
The system uses Claude Opus to analyze the data and feeds it into a Replit app that has designed every campaign through mid-2026. When 10K can automate actions, it does. The rest it assigns to humans.
The setup works because 10K has two clear goals: reach 10,000 attendees for SaaStr AI 2026 and hit $10 million in revenue this year. It optimizes for those targets alone.
Where It Beats Human Judgment
10K has no agenda. No politics. No sunk cost fallacy. When a campaign fails, it flags the problem on Day 2, not Day 60. When data shows one channel outperforming another by 4x, it doesn't protect the underperforming channel because someone on the team built it.
It processes data better than most humans because most humans have bias. That honesty might be worth the build cost alone.
Is 10K better than a great human VP of Marketing? Not exactly. But it's already better than an average marketer. The real advantage: run both in parallel, compare outputs, and debate the recommendations.
What It Can't Do Yet
10K can analyze outputs from other AI agents but cannot manage them directly. It cannot tell Artisan to adjust an outbound sequence or tell Qualified to change qualification criteria.
The problem is technical. Each AI agent has different guardrails, input processes, and training requirements. The APIs aren't standardized. Training interfaces differ. Tuning one agent might break another.
10K coordinates strategy across all 20+ agents but lacks the hands-on coaching capability to train them from a single place. That's the goal. It's not there yet.
The Broader Shift at SaaStr
In 2020, SaaStr had 20+ full-time employees: three designers, five in sales, three in content, plus operations, marketing, and event logistics. Today it has three full-time humans and 20+ full-time AI agents at the same revenue scale.
The three remaining humans work harder than they did in 2020. They handle strategic decisions, relationship management with speakers and sponsors, high-stakes negotiations, and creative direction. Orchestrating 20+ AI agents takes 30% of the Chief AI Officer's time daily.
But the output is 10 times better with roughly zero drama.
What This Means for B2B Companies
A truly great VP of Marketing costs $300,000 to $400,000 fully loaded. Half the time, the hire doesn't work out anyway.
A B2B company under $20 million in annual revenue can build an AI VP of Marketing today. It won't be perfect. But it can handle 60 to 70 percent of strategic planning work, keep you honest on data, and make human marketers dramatically more effective.
The combination of AI plus human judgment beats either one alone.
SaaStr followed one rule: buy 90 percent of what you need, build only the 10 percent where no solution exists. It tried to buy first. No product existed that did real strategic orchestration. So it built 10K.
How to Build One
Start with data. Connect your CRM, analytics, and historical campaign data. The system is only as good as the data you feed it.
Set clear goals. Revenue targets. Pipeline targets. Quarterly OKRs. Don't say "do marketing." Tell the AI what winning looks like.
Run it daily. Not weekly. Not monthly. The compounding effect of daily AI-driven planning is enormous.
Augment, don't replace. Use AI for strategy and analysis. Use humans for judgment calls and creative execution.
Budget for maintenance. This is software you own. Plan for 30 minutes daily, minimum.
SaaStr built 10+ production apps on Replit that have been used over 860,000 times. A valuation calculator hit 334,000 uses in its first month. A speaker review app replaced a $300,000-per-year agency overnight.
10K is the latest and arguably most strategically important because it orchestrates the go-to-market motion that used to require multiple humans to run.
Your membership also unlocks:
