In my previous post I wrote about what I call the age of the “incremental SaaS”. In this article, I dig a bit more into what the term “radical SaaS” covers and how I distinguish radical from incremental SaaS.
Radical SaaS: Creating a “before and after” effect
I use the term “radical” in the literal sense. As a user, “radical SaaS” products enable you to do things that you couldn’t do before. A key aspect is that there is a real “before and after” effect once you adopt such a solution. There’s a real sense of novelty and the impression that it opens you the doors to new possibilities.
But the best is to illustrate what I mean with a couple of examples that I myself experienced.
Intercom.I discovered Intercom in 2012 and I used it in production for the first time in 2013. I still remember being blown away by the concept. It was the first tool that enabled me to chat with users and collect in-product user data at the same time. Nowadays it’s normal to see a chat button on the bottom right of every SaaS product. But ten years ago it was not. And when we discovered it with my friends, we could directly feel that there would be a “before and after” Intercom.
Stripe. Stripe definitely had the same “things will never be the same again” feeling when I discovered it. Back in 2012/2012 it was a real pain to manage payments for a startup. And when we saw Stripe coming with a customizable developer friendly payment API, it was a dream come true. A fun fact is that unlike Intercom, Stripe was not available in Europe from day one. So we saw the solution available in the US but we couldn’t use it and we were begging them to come to Europe asap. The OGs will remember Paymill which was the German copycat that many of us used before Stripe was available in Europe. Again, we could feel that there would be a “before and after” Stripe.
Zendesk. I discovered Zendesk thanks to Maxime Prades back in 2009. The product was so new and so great that Maxime was actually kind of an informal “reseller” of the solution. And when we saw how businesses reacted when Maxime was demoing the product, we directly understood that they were onto something. A SaaS ticketing system seems normal today, but it was a game changer at that time.
Slack, GitHub, Algolia, Mailjet, Dropbox (and more). I could also speak of Slack and how it changed the way we communicated at eFounders (we were using Skype as our chat system before). Or how I discovered GitHub thanks to Greg Tappero. Or about the first wave of API first products such as Algolia or Mailjet that really changed how developers built SaaS products. How Dropbox emerged as the winner of the myriad of startups that changed how we shared large files online. But to sum it up, the common denominator of all these SaaS is that there was a “before and after” effect with them.
Misc thoughts
All of the radical SaaS winners were products of their time. They did not emerge from “nothing”. When you read the examples I just gave, you might think that all these startups were revolutionary companies that invented all these concepts. But of course these companies were products of their time. You could find other startups with similar concepts back then. Together with Slack you had Hipchat. Same with GitHub and Bitbucket or Intercom and a couple of startups that started with a similar approach back then. The important point is not who was the first but rather that when these companies had their breakthrough, the majority of businesses were not equipped with such software. So they radically changed the software landscape they were in. Market timing is key as radical SaaS winners are not necessarily the “first to do it”, but rather the “first to do it at the right time”.
Not all radical SaaS ended up becoming huge companies. It’s not because you are a radical SaaS that you necessarily become a Salesforce (indeed there’s only one Salesforce). Today Stripe is worth tens of billions of dollars. I don’t know their valuation number, but Intercom is probably worth less than Stripe. And Sendgrid was acquired by Twilio for $2B. So you can see a wide spectrum of outcomes. Radical SaaS do not escape the constraints of the markets they are in (market size. price dynamics, commoditization etc…).
Many of today’s incremental SaaS have better products than their radical counterparts. But it’s not enough. I often read in SaaS startups’ deck that they will displace incumbents because they have a much better product. Which can be true (that they have a better product, disrupting incumbents is another story). The issue I see is that they are most of the time incrementally better. Even a product with a 10X better UI/UX will not create the same “before/after” effect as its radical incumbent initially did. This is why I think the timing aspect is much more important than the quality of the product once a radical SaaS has captured a market. It’s hard to create a “before/after” effect if your product is a better version of something that already exists.
A good example I like to think of are video games consoles. A Playstation 5 is 1000X better than the first Playstation. But it didn’t have the same “before/after” effect as many of us had when we first discovered 3D games on the PS1. The PS1 was radical, the PS5 is incremental.
Incremental SaaS can be big businesses. A message that I do not want to convey is that incremental SaaS cannot become big businesses. They can. It’s just that only a tiny fraction of them will achieve that and the vast majority should better run a bootstrap playbook until they reach the $1M ARR threshold and then decide if it makes sense to raise money with VCs to aim for the moonshot. Incremental SaaS are not VC compatible at pre-seed and seed because they are not compatible with VCs’ time horizon (need to raise every 18 to 24 months).
As a side note, I also notice that most of the incremental SaaS that escape velocity are in big markets and target mostly mid-market/enterprise customers through a well oiled sales model. Maybe a topic for another blog post.
How to distinguish a radical SaaS from an incremental SaaS
This part is very much a work in progress. So this “mini” framework will probably evolve.
I thought for a long time about how to assess whether a SaaS startup is an incremental SaaS or if it has the potential to be radical. So far I ask myself a couple of questions around two dimensions: product and budget. I’m super happy to hear your comments about other dimensions and the questions I potentially miss.
1- Product: Before/after effect. Does the product change the way the user works/currently does things? And if yes, is it a radical/fundamental change or an incremental change?
I think it’s not necessarily obvious, especially early on, whether a product fundamentally changes the way someone does its job. Maybe because waves of truly innovative products are rare by essence. However it’s easier to detect products that simply improve how people work but do not bring a paradigm shift (sorry I hate this term, but I couldn’t find a better term). This is especially true for products that promise to do X and Y better than existing incumbents.
2- Budget: Competing for customers’ budget. Does the product compete for the budget of similar other apps that the customer already pays for?
When you sell a SaaS, if your customer needs to pay for your app by moving part of a budget already allocated to a similar tool, then imo it is very hard to replicate this process at scale. I noticed that many radical products do not compete for the budget of existing products, but rather create a “new budget need” on the customer’s side.
Applying this radical/incremental SaaS framework to GenAI startups
GenAI SaaS are all the rage at the moment and looking at them through the scope of this radical/incremental mini-framework I believe that the truly radical products in this GenAI wave are the infrastructure ones (LLMs provider and potentially some other developer tools) while the vast majority of the end-users facing GenAI SaaS (like the ones that help you write better blog posts) are incremental SaaS.
Let’s start with Infrastructure GenAI software such as LLMs providers (OpenAI) or developer tools such as LangChain and HuggingFace.
1- Product: Before/after effect. Does the product change the way the user works/currently does things? And if yes, is it a radical/fundamental change or an incremental change?
In my opinion they do. Without OpenAI’s API (and other LLMs providers) and tools such as LangChain or HuggingFace, 99% of developers simply couldn’t integrate AI powered features in their product. They really enable something completely new for many product teams and developers. So there is a real “before/after” effect here.
2- Budget: Competing for customers’ budget. Does the product compete for the budget of similar other apps that the customer already pays for?
In most cases no. Many companies that leverage these AI infrastructure providers did not have an “AI infrastructure” budget before. They do now.
And because of #1 and #2 I believe that there are radical SaaS in the GenAI infra space. I don’t know who will win ultimately, but these are radical SaaS.
Let’s move to the application layer (a.ka. end-user facing apps) such as SaaS that generate blog posts for you or help you do something you already do (Copilot for X).
1- Product: Before/after effect. Does the product change the way the user works/currently does things? And if yes, is it a radical/fundamental change or an incremental change?
I will speak about my own experience with GenAI text products and GenAI research products that I now use on a weekly/daily basis. At the beginning I clearly had a “wow effect” and I thought that there was a real before/after effect. I have to admit that my enthusiasm has faded a bit. Why? Because more than changing the way I work, it mostly accelerates the way I work. Most of these tools are fantastic accelerators rather than game changers. And where do I want to see these “accelerators”? In the SaaS products I already use. So the day when Google Docs integrates real GenAI features, I will probably stop using the dedicated tools I’m currently using.
And imo this is the issue that many GenAI SaaS face: They do improve/change the way users work, but this change is not radical. And when change is not radical, the incumbents can more easily integrate them in their product. It’s radical change that threatens incumbents. Not incremental change.
2- Budget: Competing for customers’ budget. Does the product compete for the budget of similar other apps that the customer already pays for?
I think one could argue with me that I’m harsh with the answer to the first question and that these GenAI SaaS bring more radical change than what I think. And I can understand it.
Where I think it’s harder to argue is on the budget side. For me there is little doubt that most GenAI SaaS do compete with the budget of existing tools in the customer’s stack. If a business already pays for marketing, sales or productivity tools and that these tools start to offer GenAI features, the similar standalone GenAI SaaS will have to compete with them on the budget side. And competing for an existing budget is hard when you are a newcomer.
This is because of these two reasons that I believe that most GenAI SaaS are more incremental SaaS than radical SaaS.