In 2023, one HR tech startup lost 800 monthly subscribers in just three months — even as it kept adding new features. The reason: a competitor began offering the same capabilities for free, packaged as a GPT-based plugin. The company's CEO later summed it up in one blunt line: "We were selling software. What our customers wanted was faster hiring."
The weight of this story isn't simply that "competition got tougher." It cuts at a more fundamental question: as software features commoditize at breakneck speed, what exactly should founders be selling their customers? SaaS — the subscription software model — has been the default playbook of the startup world for the past decade: charge monthly, keep shipping updates, keep churn low. Investors know the formula by heart, and founders have learned it just as well. But as AI and LLMs start to undercut the assumptions the formula rests on, a growing chorus is arguing it's time to rewrite the pitch deck.
Features Are Turning Into Commodities Faster Than Ever
Ivan Nikkhoo, a venture investor and co-founder of Navigate, laid out the shift SaaS founders now face in a recent op-ed for Crunchbase News. His point is simple: AI has accelerated the commoditization of software features, making it far harder for feature sets alone to sustain a competitive edge.
Several trends have converged in SaaS over the past two to three years. Revenue multiples for many SaaS companies have fallen to less than half of their 2021 peak. GPT-4-era startups have repeatedly shown they can replicate, with a two-person team, what an incumbent needed ten engineers to build. And customers themselves have started asking a sharper, more frequent question: "What does using this tool actually change for our business?"
What worries Nikkhoo is the response founders default to. When software-level competition heats up, the temptation is to bolt on consulting or managed services. That brings in revenue in the short term, but it kills scalability and eats into margins. The moment a software company starts operating like an agency, investors reach for a different yardstick — they get valued on service multiples, not software multiples. Worse, this approach still doesn't answer the core question: why should the customer keep using this at all?
Before Accepting That SaaS Is Broken
Not everyone agrees with this diagnosis. Enterprise SaaS players like Salesforce, ServiceNow, and Workday have kept growing since the arrival of AI — if anything, they've accelerated. By folding AI in as a feature enhancement rather than treating it as an external threat, they've expanded their customer base and kept ARR climbing steadily. There's real evidence that declarations of "SaaS is dead" are overblown.
There's also plenty of skepticism about outcome-based pricing models. How do you measure customer outcomes? How do you separate a bad result caused by the software from one caused by how the customer used it? And when revenue forecasting gets murkier, what do you show investors? Outcome-based pricing is logically compelling, but putting it into practice demands contract redesign, measurement infrastructure, and organizational readiness all at once. Some startups that rushed the transition ended up with tangled cash flow. It's worth acknowledging this isn't the right fix for every founder.
Even so, Nikkhoo's argument carries weight. His point isn't that SaaS is disappearing — it's that the strategy of surviving on features alone has run its course. Blur that distinction, and you end up prescribing the wrong medicine.
Embed Yourself in the Workflow, or Watch the Cancellations Come
At the center of Nikkhoo's survival strategy is what he calls "workflow ownership." The idea isn't just to provide a tool — it's to have your software burrow deep into the work customers have to do every single day, no matter what.
This matters in practice because replicating a single feature and rewiring a team's entire daily workflow are decisions on completely different scales. A competitor can copy the former quickly; customers don't decide to do the latter lightly. Software that's deeply embedded in a workflow tends to churn less, and it can be structured so that revenue grows in step with usage — which is exactly where usage-based or outcome-linked pricing starts to work.
This is where thinking like a finance person helps. What a customer actually calculates when buying software isn't how many features it has — it's "what number does paying $400 a month for this tool produce for our business?" More and more customers now run that calculation the way they'd evaluate an investment: what cash goes out, and what profitability improvement comes back. A founder who opens the conversation in that language, and one who opens by listing features, are having two entirely different conversations — even if they're selling the exact same product.
How you talk about AI splits along the same line. "Our product now has AI" and "customers using our product cut their processing time by an average of 40%" aren't just different in length. The first is one more checkbox in a features arms race. The second signals that you can measure and prove what actually happened to your customer's work. I'd argue that gap is the single most important fork in SaaS pitches right now.
Own the Workflow — Don't Just Sell the Software
For solo founders in Korea building SaaS, or running software-driven businesses on their own, this shift lands from a slightly different angle.
If you've actually listened to why customers cancel, that's where to start. "It's too expensive" is, more often than not, code for "the results aren't clear enough." What you're really looking for behind that sentence is which piece of the customer's workflow never actually got solved — that's your product direction.
You also need to be honest about where your software sits in the customer's day. If it occupies a nice-to-have slot, the moment a similar tool shows up, the cancellation decision comes fast. To become the tool someone has to open every single morning, you need to walk through, together with the customer, what actually breaks when the tool isn't there. The more painful that gap is, the stronger your defense.
Apply the same test before adding any AI feature: can you put a number on exactly which task it speeds up, and by how much? If you can, that's a reason customers keep paying. If you can't, the AI feature might grab attention for a moment, but it won't be the reason anyone stays.
The "deep moat" Nikkhoo emphasizes is, if anything, a more realistic path for a solo founder. Compete head-on with the big players on their terms, and you lose on scale. Going deep into one industry, one workflow, one specific customer segment is what actually builds a defense. Go broad, and you need scale. Go narrow and deep, and depth itself becomes the asset.
SaaS isn't going away. But surviving by listing out features has stopped working. The founders who move to the next stage are the ones embedded deep in their customers' workflows, showing the results in numbers. The moment the first page of the pitch deck is filled with proof of outcomes, the conversation with customers changes too.
