No, SaaS Isn't Dead... Yet.

I'm on day 5 of 75 Hard. On day 2 (day two!) I got bored of my own tracking and built myself a website to track it instead. Two workouts, diet, reading, the whole checklist. I even gave myself little gamified badges because apparently I need a gold star to drink a gallon of water.

It took me a Sunday afternoon. An afternoon! To build a whole tracking site!

Naturally, every product and engineering person I told this to had the same reaction: “Well, that's it then. SaaS is dead. We're all just going to build our own tiny custom app for our own tiny custom problem and never pay for software again.”

But, I don't think that's quite right. Here's my case.

The building is not the only hard part

If I do something twice, I'm already annoyed enough to go build the thing that saves me from a third time. But here's what I've realized building my little 75 Hard tracker: the building was never actually the bottleneck, even before AI made it ridiculously fast. The bottleneck was the noticing… catching yourself mid-task and going “wait, this is dumb, I've done this five times, there should be a button for this.”

Most people spend their day doing their work, not analyzing it. They're busy being great at their actual job — closing deals, treating patients, teaching kids, running events. Noticing the friction and turning it into “here's a tool for that” is its own specific skill, and it's a genuinely useful one to have around you, whether that's you personally or someone on your team who thinks that way. It's not that non-product people are missing something. It's just not what they're optimizing for, and that's a totally reasonable way to spend your energy.

That instinct isn't unique to product people, but it does get trained into us. That noticing muscle is worth building, for everyone, not just the product and engineering crowd. Not “learn to code.” Just learn to catch the moment.

So here's where I actually land on SaaS

These are my observations right now and I fully expect some of them to look outdated in three months — maybe three weeks, at the rate things are moving. But here's where I'm at:

1. Could an engineering team rebuild all of your SaaS products? Build a better Salesforce or HubSpot, more custom and cheaper? Yes. Should they? Not yet. Your team could absolutely build something more tailored to how you actually work. But unless your company's core value prop is building CRMs, your engineers' time is hopefully worth more solving the problems that make your product valuable and your customers happy. Rebuilding Salesforce is a distraction dressed up as a cost save.

2. Not everyone is going to build their own tools — and that's genuinely fine. The “everyone will just build their own software” take assumes everyone spends their day scanning for problems to solve, the way product people are trained to. Most people don't, and honestly, why would they? Most people feel overworked in their day to day job — spending time analyzing their job is extra work, and it takes real time and effort.

3. Being good at sales doesn't make you good at building product. Being good at product doesn't make you good at selling. These are still, today, largely different skill sets living in different people. The best product people I know aren't the ones who think they can do everyone else's job — they're the ones who know exactly what they don't know and go learn enough about sales, or ops, or support, to build the right thing collaboratively. Speed of building doesn't collapse that need for cross-functional fluency. If anything it raises the bar for it.

4. We still need technical people. Today. Yes, most code will probably be written by AI by 2027. But architecture, scale, security, the decisions that hold up when you go from 10 users to 10,000… we are not there yet. Tools like Lovable let you go from idea to working product shockingly fast (I've built an entire LMS in hours that would've taken a team the better part of a year, and a USTA rules chat app for my tennis team in an afternoon. I am a huge Lovable fan-girl). But fast-to-build isn't the same as built-to-last. The gap shows up the moment you try to scale.

The takeaway

What's actually happening isn't the death of SaaS. It's a massive drop in the cost of testing an idea, which is genuinely exciting. So play with it. Build your own tracker, your own scrappy little tool, whatever's bugging you — and actually compare it to what you're paying for. You might realize you don't need that subscription. You might realize you really do.

But spend your company's resources, and your own, on what you're actually best at and what creates real value — for you and for your business. For the big, complex, organization-wide platforms — the Salesforces, the HubSpots, the systems half your company touches every day — leave those to the people who do that for a living. For now, anyway.