MicroConf Tactics: The ‘Old Rules’ of SaaS Still Win

**Rob Walling** (0:07)
Welcome back to the MicroConf Podcast. This is a Tactics episode where I look at the old rules of SaaS and how they still win in this age of AI.
This is audio taken from one of our most popular YouTube videos. Speaking of that, if you haven't subscribed to the MicroConf YouTube channel, we have published all of the talks from MicroConf Portland, which happened just over, what, about six, eight weeks ago.
youtube.com/atmicroconf, it is a treasure trove of incredible videos from 15-ish years of MicroConf. And with that, let's dive into the old rules of SaaS and why they still win in the age of AI.
I see a lot of keyboard warriors out there shouting that AI is changing everything about SaaS, that the old rules are dead, that if you're not building AI native, you're already obsolete. They're f***ing wrong.
After building multiple SaaS companies and investing in over 230 more, I can tell you, there are fundamentals that haven't changed. If you ignore these, no amount of AI is going to save you. And one of them, AI is actually making it worse, not better. Now, I want to be clear, I'm not saying AI doesn't matter. It's actually the biggest change to SaaS I've seen in my 20 years of doing this. You can build faster, test cheaper, ship more often. But those are tools. The fundamentals, the stuff that actually determines whether your business survives, those haven't changed. The first is you still have to solve a problem people care about. AI makes it easy to build solutions, but solutions without painful problems are just toys. Problem, solution, fit hasn't gone anywhere. And I'd argue it matters more now than ever. If you solve a problem that isn't painful enough for your customers, faster coding doesn't save you. In our last batch of companies that we funded through my SaaS accelerator, Tiny Seed, we reviewed hundreds and hundreds of AI-powered applications. And we funded...
wait for it...
three.
The difference wasn't the AI, it was whether they'd found a real problem. The question hasn't changed. Would someone pay for this if it didn't have AI in the H1, or in the description, or anywhere in the code? Number two, you still have to talk to customers.
AI makes it easy to feel like you've done customer research. You can scrape Reddit, you can summarize the complaints, generate a list of pain points, but that's not research. That's homework avoidance. Real product sense comes from talking to real people, and there's no shortcut for that. AI can analyze feedback at scale, it's true, but someone still has to understand the humans. You still have to talk to customers, gather feedback, and figure out what to build. This is the role of product, and AI will not do this for you. AI can summarize it, but someone still has to understand the humans, what they're saying, and what they're not saying. One of our Tiny Seed founders, Jason Buckingham, was a development manager at Microsoft. He wanted to build a SaaS, but he had no idea what to build, so he made 70 cold calls. Not to validate an idea, but to find one. And out of those conversations, he discovered this whole world of senior living placement agents, basically realtors for elderly care, who are running their businesses on spreadsheets and sticky notes. With his co-founder, he built Senior Place. And because he actually talked to those people, he learned things no AI summary would ever surface, like the fact that his average customer is a woman in her 50s or 60s who is, in her own words, not even acquainted with computers. That shaped everything about the product. You don't get that from a reddit scrape. You get that from a phone call. Number three, you still have to market. AI can write your emails, your ads, your blog posts. It still can't make people care. Distribution is still the hard part. Marketing and sales are still the hard part, and no tool is going to do this for you. AI can help you produce that content. It can send outreach. I'm receiving dozens of these junky AI-generated emails every day. It's not going to do distribution for you. You still need to think through and find a channel that works. You still have to earn attention. The founders who win aren't the ones with the best AI tools. They're the ones who figured out where their customers hang out and how to reach them. If you're thinking about this very topic about how to find a marketing channel that works, check out our recent video about the big five marketing strategies. We'll link it up in the description. Number four. You still have to understand churn, and AI makes it worse. AI-native companies have the worst retention I've ever seen in SaaS apps. I'm seeing 60, 70, 80% annual churn. That's not a revolution. That's a leaky bucket. So this is one where AI isn't just neutral. It's actively making things harder. It's making them worse for SaaS founders. The data is brutal. AI-native products have median gross retention, around 40%, compared to 90-plus percent for traditional B2B SaaS. A lot of what's being counted as ARR is really experimental budget, companies trying things out and not committing. People aren't leaving because of payment failures. They're leaving because the product didn't stick. It didn't work. It didn't live up to the promise that was made. Founders still have to diagnose why customers leave and fix root causes and not mask churn. AI doesn't solve churn and in most cases, it accelerates it. These fundamentals aren't complicated, but they can be hard. And they're a lot harder if you're trying to figure them out on your own. One of the best things I ever did for my entrepreneurial journey was to join a mastermind. It's a small group of founders who meet regularly to share wins, talk through challenges, and hold each other accountable. And finding the right people is the hard part. That's why we started Mastermind Matching at MicroConf. We've successfully matched almost 1,800 founders across 64 countries, and applications for our next group of masterminds are open now. You can go to microconfmasterminds.com to apply. Alright, so I have one more fundamental, and this one is less to do with technology, and more to do with your mindset. It's execution over ideas. AI has made ideas cheap, execution is still rare. This might be the most fundamental lesson in this video.