Software isn’t dying, but it is becoming more honest
Have you heard? "SaaS is dead." A lot of these essays come from the investors who, to be fair, benefited most from the SaaS explosion over the last 15 years. But while I agree that the subscription-based billing side of SaaS is reaching its expiration date, I’ve started to believe that’s actually a good thing. Software isn’t going anywhere, but the business model underneath it is finally evolving.
We often conflate “Saas” with “fixed subscriptions,” but they aren’t the same thing. At its core, SaaS was a shift in how software was accounted for—moving it from a fixed capital expense (the old “license and maintenance” era where you bought hardware and paid a massive upfront fee) to a recurring service.
But not all software businesses are SaaS businesses, and not all SaaS businesses have to be tied to a fixed subscription. The panic we’re seeing right now is coming from the realization that “recurring” doesn’t have to mean “predictable seat-based billing,” even if investors want it to 🙂
The seat-based status quo
For a long time, seat-based enterprise software was a fantastic deal for vendors and a mediocre one for customers. It created "shelfware"—licenses bought "just in case" that sit idle, never get activated, yet keep the recurring-revenue ticker moving. We’ve all seen it: a company "trips the core" and adds a few dozen seats that no one actually uses.
Outcome-based billing changes that dynamic. Instead of paying for potential capacity, you pay for what you actually accomplish.
I was an early investor in Metronome (acquired by Stripe), which pioneered metered, or usage-based, billing. They took off because AI companies like OpenAI and Anthropic needed to align their own infrastructure costs with their consumption. It’s hard to charge customers per token if you’re stuck paying a fixed, flat subscription for the underlying pipes. Usage-based billing is a big step in the right direction. You're paying for consumption (effort), but you're still not necessarily paying for results.
What we’ve accepted
Look at the account-based marketing (ABM) space as an example. Platforms like 6sense, Demandbase, and Terminus charge $100K+ per year based on the number of accounts you want to target, which platform tier you choose, and how many seats you need.
But here's what’s crazy: these platforms promise to drive qualified pipeline (that's literally the entire value proposition), yet they're priced on access to features and account volume, not on the pipeline they actually generate.
You could run campaigns for a full year that produce zero qualified opportunities and still owe the full subscription. The vendor has no skin in the game. This is the model we've accepted as normal, and it's finally starting to crack.
Why couldn’t we do this before?
I often hear founders and other hyper-rational types ask why we haven’t always billed for outcomes. The answer usually boils down to technical limitations and risk.
Historically, vendors needed that predictable subscription check to survive. We couldn’t reliably measure or guarantee outcomes because there were too many uncontrollable variables. If we couldn't prove the software caused the "win," we defaulted to the safest model: pay us every month, regardless of whether you’re getting value.
I don’t think the future is a binary choice where "no one gets paid unless the result is perfect." That’s not how business works.
The real future is likely a hybrid: a base fee that covers the "controllable inputs" (the actual hours and effort) plus an outcome-based "kicker" for the win. This aligns incentives properly. The vendor is protected for their work, but they’re rewarded for the actual result. Everyone wins.
What the future of SaaS could look like
I recently invested in a company that does ABM called hyperGTM (which feels significant, because I so rarely invest in martech!), that is already putting this into practice. When you look under the hood, they aren't just selling a tool; they’re committing to a target. They use a blend of technology and strategy to say, "We’re going to hit the number we agreed on, and we’ll use our tech to bridge the gap."
For example, they commit to generating at least 9 meetings from an initial sample of 30 target accounts, or 3x more SQLs from conferences, trade shows, and events. Customers pay a base service fee of $650 per account, which covers the research, outreach, and gift delivery. Then there's a one-time outcome-based kicker of $4,000 per meeting when hyperGTM books a qualified meeting with an executive.
“With agents, startups can significantly reduce labour cost internally and with lower burn you can actually afford to meet customers where they are, prove and provide value, and then upsell significantly.” - Adit Trivedi, co-founder & CEO of hyperGTM
This is only becoming possible now because of AI agents. The "blend of things" required to hit a goal used to be too manual and expensive to scale.
With agents, that complexity becomes programmable. Agents can:
Run 100+ messaging variations across channels simultaneously and learn which combinations actually drive conversions
Reallocate budget and effort across tactics in real-time based on what's working, without waiting for a human to update a spreadsheet
Orchestrate multi-step workflows—prospecting, qualifying, nurturing—and optimize each step based on actual pipeline movement
This isn't a dashboard showing you lagging indicators. This is technology that can guarantee outcomes because it can adapt faster than humans ever could.
Opportunity to create category leaders
This isn't just a better deal for customers, it's a massive opportunity for the companies that move first. Today, everything is a race! And for the first time in a long time, it seems like the race is open to everyone.
Usage-based pricing created Snowflake and Databricks. Outcome-based pricing will create the next generation of category-defining companies in infrastructure and GTM.
The early movers won't just capture market share, they'll define entirely new buying categories. When you're the vendor selling guaranteed results while your competitors are still selling seats and hoping for the best, that's not a fair fight. You win the deal every time.
For investors, this means higher NRR when outcomes are achieved, better retention (customers don't churn vendors that deliver measurable results), and the chance to back true category creators.
I get that this isn't easy and changes won’t happen overnight. Revenue recognition gets complicated. Forecasting becomes harder when outcomes introduce variance. You need tech that can actually deliver on the guarantees, not vaporware. And you need customers who will partner on success, not just buy and disappear.
These are solvable problems, but they're real constraints. The companies that figure it out early will have a decade-long head start.
Conclusion
Software isn’t dying; it’s just getting more honest. We’re moving toward a world where we pay for the work (the base) and the value (the outcome).
Subscription-based billing won't remain the standard, and that’s actually a win for everyone… including the investors who got used to the predictability of the old model. They will eventually come around when they see that the companies betting on outcome-based models aren't just growing faster—they're creating defensible moats that seat-based competitors can't match.