Podcast Summary: SaaStr 821 – Swapping Stories: Building AI-First Companies with Zapier CEO Wade Foster
Episode Date: September 17, 2025
Podcast: The Official SaaStr Podcast
Host: Amelia Larue (Chief AI Officer) & Guillaume Ghuban
Guest: Wade Foster, CEO and Co-Founder of Zapier
Overview
In this episode of the new SaaStr series “Swapping Stories,” hosts Amelia Larue and Guillaume Ghuban sit down with Wade Foster, CEO and Co-Founder of Zapier, to discuss what it truly means to build an "AI-first" company. The conversation touches on the evolution of Zapier’s product, the rise and realities of agentic workflows, AI fluency as a hiring requirement, and practical lessons from both Zapier and its customer base on AI adoption in the SaaS world.
Key Discussion Points & Insights
1. Zapier’s Evolution: From Workflow Automation to AI Orchestration
- Original Zapier: Started as a simple integration tool (“If this then that” for businesses).
- Phase Two: Embraced complex workflows – a trigger could launch multiple actions across apps.
- Today: Zapier is positioned as the “most connected AI orchestration platform,” embedding AI end-to-end in workflows, including AI agents that can perform complex autonomous tasks.
[04:15] Wade Foster: “Now with AI, you're able to embed AI as part of those workflows end to end… Zapier has become this full on automation suite that uses AI to power any of the workflows that you might have inside of an organization.”
2. Deterministic vs. Agentic Workflows
- Definition & Importance:
- Deterministic workflows = repeatable, reliable, predictable actions (e.g., putting a new lead into the CRM).
- Agentic workflows = processes requiring autonomy, judgment, or “human-like” decisions (e.g., summarizing a customer account for a renewal).
- Advanced users combine the two, letting AI handle ambiguity and humans handle the final step for accuracy.
[06:19] Wade Foster:
“What is most misunderstood here?... The people actually deploying automation at scale are finding ways to mix and match deterministic workflows with agentic workflows and do those together.”
3. Real-World AI Use Case: The Renewals Agent at Zapier
[08:36] Wade Foster:
- Automated Renewal Process: For renewing enterprise contracts, a Zapier agent:
- Gathers account data (usage trends, support tickets, sales call transcripts).
- Analyzes for upsell, renewal or contraction signals using prompts and historical examples.
- Inserts recommendations and drafts emails directly in HubSpot for AEs.
- Takes 90% of the work off the AE; the human adds the final touch and executes.
[09:22] Wade Foster:
“The agent is basically doing 90 plus percent of it and then leaves the last mile to the account rep…”
4. The Role of Humans in the Loop
- Why Human Oversight Still Matters: AI delivers 90%+ accuracy, but nuance and error handling still require humans—especially when stakes are high (e.g., enterprise sales).
- Example of AI Limitation: AI hallucinated Saster London speakers who hadn’t been announced, pulling data from previous events.
[16:24] Amelia Larue:
“That final mile typically is better served by a human… Sometimes, one, it's wrong or it hallucinates. Two, …the AI doesn't really quite understand those complexities in the same natural way that a human being would.”
[13:26] Wade Foster:
“It takes a lot of work to get that last mile good enough… there are a lot of risks today if you don't do that last mile work.”
5. Cultural Buy-In & AI Adoption
- Zapier’s Internal AI Culture:
- Small, automation-friendly sales team; existing culture of “don’t be a robot, build a robot.”
- High receptivity to automation—AI agents allowed the team to cover more accounts, not take jobs.
[17:35] Wade Foster:
“We've always been an automation company and the value is don't be a robot, build a robot.”
- Customer Base Adoption:
- Existing Zapier users (familiar with deterministic workflows) are “quick on the uptick” with AI.
- New users are eager but less skilled, starting simple and building up gradually.
[19:19] Wade Foster:
“Even if you aren't great at it yet, AI is good enough to start to translate some of what you were trying to think and get you started.”
6. “Vibe Coding” and the Expanding No-Code Ecosystem
- Trend: More people are “vibe coding” (experimenting, building with AI/no-code tools) and using platforms like Zapier to glue their stacks together.
- Impact: The barrier to entry is lower; even non-technical people can get started and scale up ideas.
[21:59] Wade Foster:
“The learning curve to getting into this stuff is just so much less than it used to be... All these tools that make it easier for people to bring their own ideas to life are benefiting.”
7. MCPs (Multi-Agent Communication Protocols) vs. APIs
- MCPs: A new, agent-to-agent communications protocol; enables tools/AI agents to talk and act across platforms (e.g., Claude can now operate within HubSpot via MCP).
- Comparison: MCPs are flexible but less deterministic than APIs; both will have their place. [23:23] Wade Foster: "APIs are deterministic and cheap... where MCP is a lot more flexible, the agent is getting to choose this stuff... Both are still quite valuable."
8. Zapier’s Approach to AI Fluency in Hiring
- Hackathons & Knowledge Sharing: Regular AI-focused hackathons; weekly “show and tell” at all-hands meetings to spread best practices and inspire.
- Hiring for AI Fluency: Defined expectations by function, with skills tests and behavioral questions in interviews (not just “Do you use ChatGPT?”).
- Advice: Don’t stop at the CEO memo—define what AI fluency means for each team and assess accordingly.
[30:54] Wade Foster:
“It follows hiring best practices… ask what they're experimenting with, do skills tests, screen shares. Go function by function and set a baseline for your company.”
Notable Quotes & Memorable Moments
-
Deterministic vs. Agentic:
“The people actually deploying automation at scale are finding ways to mix and match deterministic workflows with agentic workflows.” — Wade Foster [06:19] -
On AI’s Last Mile:
“It takes a lot of work to get that last mile good enough where you would say, hey, we’re actually going to feel confident actually having the AI act as the sales rep or act as the CSM in these equations…” — Wade Foster [13:26] -
On Company Values:
“Don’t be a robot, build a robot.” — Wade Foster [17:35] -
Vibe Coding’s Democratization:
“All these sort of like tools that make it easier for people to bring their own ideas to life, I think are benefiting from this…” — Wade Foster [21:59] -
On AI Fluency:
“I talked to a lot of people who… do the CEO memo and then it’s just like chaos inside the organization being like I don’t know what it actually means… just go function by function and start to define some of these things.” — Wade Foster [30:54]
Highlighted Timestamps for Key Segments
- AI Orchestration Platform and Workflow Evolution: [04:15]
- Mixing Deterministic & Agentic Workflows: [06:19]
- Renewals Agent Use Case (Deep Dive): [08:36]
- Human in the Loop and AI Limitations: [13:26], [16:24]
- Cultural Buy-In at Zapier: [17:35]
- AI Adoption by Users: [19:19]
- Vibe Coding and No-Code Trends: [21:59]
- MCPs vs APIs: [23:23]
- AI Fluency Hiring Tactics: [28:54], [30:54]
- Rapid Fire: Favorite AI Tools (Granola, Monologue): [33:19], [34:27]
- Upcoming Zap Connect Event Sneak Peek: [35:19]
Takeaways for SaaS Founders and Operators
- Start experimenting with small, hybrid AI workflows—don’t try to automate everything at once.
- Define “AI fluency” for every team; set clear expectations during hiring and ongoing training.
- Leverage the lower barriers of entry: encourage “vibe coding” and empower non-engineers to prototype with AI tools.
- Place humans in control of the ‘last mile’ until technology matures and error rates are fully acceptable for your business risks.
- Stay abreast of new protocols like MCPs that are enabling richer agent-to-agent integrations, but don’t discard classic APIs.
- Continually share learnings internally as AI adoption evolves rapidly.
For more on use cases and concrete workflows:
Check out Zapier’s upcoming Zap Connect event, featuring real customer demos and actionable examples.
This summary aims to condense and clarify the episode’s rich discussion, giving SaaS and AI practitioners concrete insights and a peek into AI-first culture at a market leader. For the full experience, listen to the episode for voice and context.
