Mar 14, 2025 · 49m · 20product
20Product: How to Design and Build Products in a World of Agents | Why AI Will Kill Many SaaS Products | What Products Will Thrive and Die in a World of 100M Developers with Matt Biilmann, Co-Founder and CEO @ Netlify
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of 20 Product, Netlify co-founder and CEO Matt Biilmann joins Harry Stebbings to discuss the emergence of Agent Experience (AX), the replacement of generic SaaS with custom AI-built applications, and how 100 million new developers will shape the future of software infrastructure.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Harry holds 30.7% of the talking time here. How this is scored →
Harry as informed peer 4.6 Guest teaching 4.6 Guest disagreement 1.6 Harry pushing back 4.2
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 13:05 Matt rejects 15-year enterprise adoption timeline
Matt directly rejects Harry's conservative 15-year timeline estimate for enterprise adoption of AI builders, cutting his own estimate down to 5 years based on rapid tool evolution.
Hardest push from Harry ▶ 11:07 Harry disputes replacing generic SaaS with custom AI appsHarry forcefully pushes back against Matt's claim that internal custom apps will displace generic SaaS, calling out Klarna's claims as unrealistic.
Biggest teaching moment ▶ 14:21 Matt outlines zero-ops infrastructure requirementsMatt educates Harry on why custom app proliferation will crash under maintenance costs unless built on top of zero-operations serverless building blocks like Jamstack.
Harry holds his own ▶ 29:25 Harry defends closed ecosystem dominance with hardware and distribution factsHarry demonstrates deep industry knowledge by citing Google's proprietary chips, compute scale, and consumer touchpoints alongside Microsoft's enterprise distribution to challenge Matt's open-source thesis.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Defining Agent Experience and Building for Agents | 2 | 5 | 1 | 4 | Harry interrupts to ask bluntly how building for agents differs from normal building. Matt explains Netlify's claim-flow system that allows agents to deploy sites without upfront user signup. | |
| Design, User Experience, and Strategy for Agentic Products | 3 | 4 | 1 | 4 | Harry uses aggressive phrasing to ask how founders should direct product strategy. Matt patiently outlines how user experience shifts to accommodating specific agent ecosystems. | |
| Debating Generic SaaS vs. Custom Internal Applications | 7 | 5 | 3 | 8 | Harry directly challenges Matt's claim that custom internal apps will replace generic SaaS, citing Klarna's claims as unrealistic. Matt holds his ground, arguing that trade-offs in adapting to generic SaaS will drive companies to build tailored internal tools. | |
| Enterprise Adoption Timeline and the Rise of AI Application Builders | 5 | 6 | 3 | 6 | Harry rejects Matt's timeline, arguing enterprises change slowly over a 15-year horizon and asking how maintenance is managed. Matt counters by shortening his timeline estimate to 5 years and detailing the necessity of zero-ops serverless infrastructure. | |
| The Future of Design Tools and Prompt-Based Prototyping | 4 | 5 | 2 | 4 | Harry questions whether conversational chat is the ideal interface for AI by making an analogy to retail shopping. Matt explains that while chat works for simple queries, brand experience and non-deterministic UI require guided digital interactions. | |
| AI Revenue Sustainability and the 100 Million Developer Era | 5 | 5 | 2 | 4 | Harry compares AI builders to no-code platforms like Webflow, noting non-technical users still struggle. Matt acknowledges skill barriers but predicts a new wave of 100 million developers operating at a higher abstraction layer. | |
| Software Abstraction Layers and Historical Tech Shifts | 3 | 6 | 1 | 2 | Harry asks how developers will be split across different tool categories. Matt provides a historical overview of software abstraction layers from Ruby to React, demonstrating how new tech layers add developers without killing underlying roles. | |
| Open vs. Closed Models and Ecosystem Dynamics | 7 | 4 | 2 | 6 | Harry argues strongly for closed walled-garden ecosystems, citing Google's silicon, compute, and consumer touchpoints. Matt maintains that open ecosystems win because developers prefer flexibility across multiple LLMs. | |
| LLM Specialization and AI Investment Pick: Anthropic | 5 | 4 | 1 | 3 | Harry sets up a valuation trade game comparing Anthropic, OpenAI, and Grok. Matt picks Anthropic, explaining that their early safety research inadvertently helped them build superior practical steering for real-world code generation. | |
| Value Accrual, Outcome Pricing, and Infrastructure Winners | 6 | 3 | 1 | 2 | Harry offers a detailed thesis on Stripe expanding internet GDP and capturing enormous value in an agentic economy. Matt agrees, reinforcing that zero-ops building blocks will thrive. | |
| Reflection on AI Safety Panic and Early Over-Regulation | 4 | 4 | 1 | 3 | Harry pokes fun at existential AI doom panics, noting models currently struggle with basic tasks. The segment shifts to a quickfire round covering Netlify's fundraising history and valuation peaks. |
Statements from this episode (27)
Biilmann: Next 100M developers will emerge around agentic tools
“I think we'll see the next hundred million software developers emerge around tools that are Gentic, but I still think it will be a profession to use those tools to do stuff, to build stuff for other people.”
Biilmann: Google has failed to productize generative AI compellingly
“If I look at Google's products, like I now have all these Genesis buttons all over and I never use them. They just haven't managed to productize it in a way that feels super compelling.”
Software companies defer user signup to support AI agent onboarding
“We've seen a lot of companies now start adopting similar flows, right? Like in the authentication space, in the database space and so on to make the tools available to agents without an explicit user's sign up and set up step first and delaying that step to af…”
Biilmann: Great DX will require great agent experience within five years
“I just don't think anyone can have a great developer experience five years from now, if they don't have a great agent experience.”
Biilmann: AI agents will likely reduce Salesforce to a system of record
“But they definitely feel it very strongly as like the biggest threat to them to be reduced to that system of record and not the actual like workflow platform. And I think that's pretty likely what would happen, right?”
Biilmann: Cheaper development will trigger radical surge in custom software demand
“Like I think what we're seeing now is definitely the cost of building custom workflows, custom UIs, custom applications is going to dive down dramatically. And typically that will mean that demand for doing that will Go up radically and people will start build…”
Biilmann: Build vs buy equation will shift toward custom internal apps
“I think it's a balance that will really start tipping, right? Like where today, if you're a company and you have a specific business need, hiring a development team to build, deploy, operate an internal application just for you to solve that business need, the…”
Stebbings: Klarna is not replacing all of its SaaS with AI
“And we have companies like Klarna who are saying, oh, we're replacing all of our SaaS with AI. They're not, they're absolutely not. And they won't.”
Stebbings: AI replacement of enterprise SaaS will take 15 years
“Respectfully, I think it is a very small customer segment that will be doing that in the next 15 years.”
Biilmann: AI custom apps will challenge enterprise SaaS within five years
“And now my timeline is more like five years and 15 years before it's going to be like really relevant.”
Biilmann: Custom AI apps fail on Kubernetes due to overwhelming maintenance costs
“Like otherwise, if it's just fully custom built Kubernetes cluster stuff, all of it, the viability of that whole approach goes away, right? Like the maintenance cost just overtakes everything.”
Biilmann: AI agents are 15-20 years away from maintaining complex software infrastructure
“And then you could speculate, can the agents maintain it? But I think. I think that, that then we are definitely in a 15, 20 year horizon before, before they could be relied on for anything like that.”
Biilmann: Figma faces existential revenue threat without end-to-end AI building tools
“But if Figma doesn't manage to connect that to the full end to end chain of prototyping and building and so on, then this could be a pretty existential threat. To companies like that over time, or at least like a threat to their current revenue model”
Biilmann: Starting new software projects is shifting from design templates to AI prompting
“There's no doubt that the way people start projects in general are really already starting to shift a lot, both from traditional templates and bootstraps and front end frameworks on the one end, on the other end from like design prototypes and visual design on…”
Biilmann: Chat Prompting Will Never Deliver Differentiated E-Commerce Experiences
“If I follow a link and I land on an e-commerce from like a large brand that's trying to really differentiate themselves and be something different, right? Like I definitely just, I wouldn't want just to have some box saying like, you come up with what you want…”
Biilmann: AI Web Builders Will Force Brands to Adopt Novel Interfaces
“When the cost of building an e-commerce that looks like any of the current brands today goes down to like, just Ask Bolt to do it for you. Ask Lovable to do it for you. And you have that. Those brands are going to have to come up with something much more engag…”
Biilmann: AI builder tool revenue is durable, not a sugar high
“When I look at tools like bold and lovable, the category itself has something where it seems obvious that there will be a lot of money spent there. Just because like, when you see some of the examples of what people can do with those tools that they just could…”
Biilmann: Higher software abstraction layers did not shrink traditional developer populations
“But the interesting thing is that the layer below of like Ruby and Rails developers and Java developers and .NET developers. It's probably still at least as large as this was before that. The new layer just got added on top.”
Biilmann: Open AI ecosystems will defeat closed verticalized models
“And it'll actually be an advantage for like the agent builders to not be tied to a specific model and be able to like, just jump around between whichever model is best. And I think once that start is happening, we're going to have like that, that's going to me…”
Stebbings: Meta's productization of Llama AI models has been very poor
“I would say that Facebook is the same like Lama and the productization around Lama has been bluntly very, very poor at the moment.”
Biilmann: Bullish on Anthropic due to coding and programming superiority
“Anthropic definitely has really stood out in, in their ability to solve programming and coding style tasks in a way that I think is like I'm very bullish on Anthropic for that reason, because I think that's such a groundbreaking part of what makes LLMs multipl…”
Biilmann: Would buy Anthropic over OpenAI and Grok at current valuations
“I would probably go for Anthropic. It's a really interesting thing from the outside where a lot of their motivation started around these like safety concerns and like, how do we allow the models to not take over the world and so on. And I think after sort of a…”
Biilmann: Would sell Grok over OpenAI and Anthropic
“I would probably sell Grok. I still think like it sits a little more in a middle space”
Stebbings: Stripe's value and internet GDP will soar via AI entrepreneurs
“I think Stripe will be massively more valuable. In a world that we now enter, because I think there will be so many people who are enabled to create stores, entrepreneurial endeavors, and will inevitably use the default of Stripe as their payments mechanism th…”
Biilmann: Microsoft is the safest 10-year stock buy under Satya Nadella
“It's a hard one, but I would probably say Microsoft. I think again, if you can only buy one, you probably want it to be a pretty safe one. You want it to be one that benefits from all of the disruption AI is going to bring to the field rather than like falls f…”
Biilmann: WordPress has reached the end of its innovation cycle
“I mean, I think WordPress is, is in a challenging position where they are kind of like in the end of their innovation cycle. And I don't think they're in a great position to respond to all of this disruption. So I think like it'll be seen more and more as a bi…”