For years, founders had a clear target to aim for. Triple-triple-double-double was ambitious, sure, but it gave you something concrete to build toward. If you could pull off that kind of growth, you were well positioned for a great fundraise.
The benchmarks weren’t easy, but they were at least clear. In the pre-AI era, SaaS companies would often demonstrate consistent and comparable growth patterns. Some might argue that investors were over-reliant on ARR as the singular indicator of a startup’s health, but it was at least a shared language across investors, startups, and sectors.
This consistency has since given way to a bar that is both higher and less clear.
The problem isn’t that standards have dropped. It’s that growth patterns themselves have become far less clear. Companies can now spike to millions of users in weeks, then decline nearly as fast. Revenue can look spectacular on paper while masking fundamental weaknesses. Traditional benchmarks still get measured, but they’ve stopped being a complete marker of a startup’s growth and fundraising prospects.
This has forced investors to completely rethink how they evaluate companies. The old question was “Did you hit the numbers?” The new question is “What do these numbers actually tell us about what the future might hold?”
Why Growth Patterns Broke
AI collapsed the cost and time required to build and ship products. A competent founder can get something real in front of customers in weeks instead of months. On the surface, this seems like pure upside. In practice, it makes the market much noisier and conviction more elusive.
A few years ago, if you had a product in market with meaningful traction, that alone told investors something. Building and distributing software was hard enough that getting to scale proved capability and demand. But in an era where we are seeing more companies than ever reach $1M, $10M, and $100M in revenue, other metrics become key factors in shaping a startup’s fundraising story. The elements that matter now vary nearly as much as the startups and sectors themselves.
The issue compounds because AI products can show usage growth that doesn’t necessarily translate to lasting value. Someone can rack up impressive engagement metrics while building something that’s fundamentally a novelty. Traditional SaaS had cleaner signals. If a company was paying you $50K annually and renewing, you were probably solving a real problem. Consumption-based models are murkier. High usage might mean genuine dependency, or it might mean people are playing around with a toy.
At the same time, what used to pass for solid early revenue has gotten less reliable:
-
- Pilot programs that once signaled real interest now feel more like experimentation budgets.
- Customer budgets have tightened in many sectors, making any given dollar harder to defend if the value isn’t immediately obvious.
- Competitive disruption is happening faster than it used to—the barrier to starting a new company has dropped dramatically.
- Buyers are more willing to try something new and innovative, which means yesterday’s traction can evaporate surprisingly quickly.
Early revenue growth used to be the gold standard metric. It’s not anymore. We’ve seen too many companies hit $5M or $10M ARR, then stall completely because they never built anything defensible. The growth came from being first or novel, not from being necessary.
The result is that investors can’t rely on pattern matching anymore. There’s no “if you have X ARR at this stage, you’ll raise a Series A at Y valuation.” Every company requires actual judgment about whether the growth is real and whether it will compound or decay, or whether it is poised to be disrupted by a new entrant or by expanded functionality from foundation model companies or legacy tech players.
What Investors Actually Evaluate Now
The companies that raise strong rounds today aren’t necessarily the ones with the best top-line metrics. They’re the ones that can prove their growth is durable and accelerating, not fragile and plateauing.
Here’s what that looks like in practice.
Clean metrics that prove real demand
Investors have stopped caring whether you hit a specific ARR milestone. They care whether your metrics prove that customers need what you’re building.
A company with $500K ARR and crystal-clear unit economics can raise a strong round if the data shows undeniable product-market fit. A company with $5M ARR will struggle if half of it came from one-time pilots or consulting work dressed up as recurring revenue.
The metrics that matter depend entirely on your business model. For some companies, it’s retention cohorts. For others, it’s consumption growth inside existing accounts. For others, it’s expansion revenue happening organically without the founder grinding through every upsell conversation.
What kills deals is ambiguity. If you’re not precise about how you define revenue, or if your numbers require lengthy explanations to make sense, investors assume you’re either confused or hiding something.
Momentum matters more than your current scale
A $1M ARR company growing 30% month-over-month will raise faster than a $10M ARR company growing 10% quarter-over-quarter.
Investors are trying to figure out where you are on the curve. Are you accelerating into your steepest growth phase, or have you already hit it and started decelerating? Your current size matters far less than your trajectory.
The tells they look for are straightforward. Is growth speeding up or slowing down? Are existing customers using the product more each month, or has usage plateaued? Is expansion happening because customers are pulling the product deeper into their workflows, or because your sales team is heroically pushing it?
If your last two quarters look flat, the burden is on you to explain why the next two will be different. “We’re about to launch X feature” rarely cuts it. Investors have heard that too many times from companies that never broke out of the plateau.
Economics that make sense for your business
AI companies often have gross margins hovering in the 50-60% range (and sometimes as low as 20% for pure infrastructure providers). While that would have been unimpressive in the SaaS era, it’s standard in today’s AI world, where COGS includes inference and GPU costs.
If you’re delivering value through inference and your costs reflect actual usage, lower margins early can be a healthy sign. The question isn’t whether your margins match Salesforce in 2010. It’s whether you understand your unit economics well enough to explain why they make sense and how they’ll improve as you scale.
The same logic applies to go-to-market spend. Product-led companies with lean sales teams might have CAC:LTV ratios that look strange compared to traditional SaaS benchmarks. That doesn’t make them wrong. It makes them different. What matters is whether you can articulate why your model works and what the path to profitability looks like.
A wedge that makes your expansion obvious
Here’s where most pitches fall apart. Founders either show metrics with no coherent story about where the company is going, or they pitch a grand vision with no evidence that the current product gets them there.
The companies that break through can explain their wedge clearly. They solve an urgent, specific problem today that customers will pay for right now. Solving that problem positions them to expand into something much larger. And their current traction proves they’re actually on that path, not just hoping to be.
When those three things line up, the story tells itself. Investors stop needing to be sold. They can see where you’re going because the product roadmap, the expansion strategy, and the long-term ambition all reinforce each other.
If you can’t explain that arc in a few minutes, you probably haven’t figured it out yet. The market won’t do that work for you.
Durability signals that prove this will stick
In a world where products can grow fast and collapse just as quickly, durability has become one of the most important filters.
Investors are calling customers earlier in the diligence process and weighing those conversations as heavily as the metrics deck. The question they’re asking, in various forms, is simple: what happens if this product goes away tomorrow?
If the answer is “we’d figure something out,” your round is probably dead. If the answer is “we’d be in serious trouble,” you’re in good shape.
The durability signals that matter are concrete. Are customers expanding usage without you pushing them? Are they embedding the product into workflows that would be expensive to unwind? Do they have proprietary data or processes locked into your system? Are they asking for features that make them more dependent on you, not less?
If you don’t know what your best customers will say when investors call them, that’s the problem you need to solve before you fundraise.
What This Means for Founders
The fundraising market in 2026 rewards companies that can prove their growth is real and sustainable, not just companies that hit arbitrary milestones.
The game has changed in the AI era. Investors are making actual judgments about momentum, durability, and trajectory instead of relying on whether you checked the right boxes.
For founders, this means you need to understand your business with a deep level of precision and use the numbers to help convey a narrative. Clean metrics. Clear momentum. A coherent wedge. Customers who would genuinely struggle without you. As always, the story is most important, and the metrics exist to help support that.
The question isn’t “are we big enough to raise?” It’s “if an investor spends two hours looking at our data and talks to our top customers, will they believe we’re building something that is truly differentiated and can compound for decades?”
Because that’s what closes rounds now.
The post The New Rules for Fundraising in the AI Era appeared first on Madrona.