OverviewSaying that investors won’t take a conversation with you until you’ve passed the AI gate may make it sound like expectations are high — that you have to prove AI revenue is mature enough to make a meaningful difference on your P&L today.
But the truth is that crossing the initial AI gate is easier than you think.
To get a foot in the door, you need a clear point of view on where AI fits in your business, what you’ve already built, and what the roadmap looks like from here. That’s enough to pass the initial AI gate today.
At our Boston event, one investor said that even 5% of contracts carrying distinct AI revenue can be enough of a signal that your business is worth evaluating.
“Today, there is no framework for valuation,” they said. “Let’s first cross that AI gate.” Once you satisfy that expectation, you can start to build the story that earns your multiple.
The signal that satisfies investors at the gate — a handful of contracts carrying distinct AI revenue — stops being the point once the conversation progresses. From there, investors want to know whether that revenue reflects something the business owns that a competitor can't easily copy.
One of our CFO Roundtable speakers said: "We care less about whether 5% of revenue is explicitly from AI. We care more about whether you have a meaningful moat that others cannot replicate."
The high-performing companies we’re seeing build those “meaningful moats” are leveraging:
AI features alone don't produce any of that. A competitor can ship a similar feature within a quarter. What can't be shipped that quickly is a customer base that has built its operations around a product, or a data advantage that compounds with scale.
Retention and adoption are what prove a moat is real. A customer who keeps paying for AI and keeps using it is telling investors the moat is holding. That evidence is what separates a company that crossed the gate from one that's actually built something defensible. And it's what determines the multiple your company can command.
Investors are going to test your AI moat narrative using the same quality-of-revenue scrutiny they’ve always applied to growth stories.
The following four questions can help you evaluate whether or not your AI story will hold up to that scrutiny.
1. Do Your Contracts Separate AI Revenue from Base?
If AI capabilities are bundled into your base contract pricing, investors have no way to test whether customers value your AI enough to pay for it specifically. Separate AI revenue means contracts where customers make a conscious choice to pay for AI outside of what they were already spending, reported as a distinct line item in your ARR.
A customer who chooses to pay for AI outside their base contract is doing exactly what a moat requires: assigning it value on its own. That choice is the first piece of evidence investors can actually test.
2. Are You Tracking Gross Retention at the AI Contract Level?
As companies add AI revenue through upsells to existing customers, the expansion side of NDR improves. That's expected, but it can obscure whether the base is healthy and whether those AI upsells will actually renew.
Gross retention strips out the expansion. It measures whether existing customers are staying at the same contract value, separate from any upsells.
One investor said: "Where I see the most important metric right now, it's gross retention — because you can play around and manipulate net retention. And it has to always include down sell."
A customer who reduces their contract value — even without churning — is a retention problem that a strong NDR will absorb without surfacing. Gross retention that includes downsell forces that problem into view before it becomes a diligence issue.
Adoption data is the companion metric. Low usage alongside a signed AI contract is a renewal risk that won't show up in your ARR until that contract comes up.
3. Have You Segmented AI Revenue by Customer Profile and Channel?
Aggregate AI revenue numbers are important, but the real evaluation lies in composition: which customers are paying for AI, how they were acquired, and whether the retention story holds across different cuts of the business.
Some cuts investors will ask about:
A moat that only shows up in one segment isn't a moat yet. Working through these cuts before a process starts is what gives you the answer before investors ask for it.
4. Is Your AI Pricing Model Tied to Measurable Customer Outcomes?
Seat-based pricing is under structural pressure as AI agents begin replacing the users those seats were built around. Beyond that shift, seat-based models give investors no clear way to evaluate how customers are actually valuing your AI.
The pricing structures that are working tie the cost of AI to a specific outcome customers can verify:
The pricing model matters less than the ROI story attached to it. Customers will accept meaningful AI price increases when they can tie the spend to something they can measure. Without that connection, AI gets treated like any other software line item, and it won't survive the next renewal.
Everything above — separated contracts, gross retention, segmentation, outcome-tied pricing — depends on the same thing: a revenue and customer data foundation clean enough to produce those answers on demand. One investor confirmed that “the businesses that get the fastest term sheets have amazing data.”
To meet investor AI expectations, you need a Revenue and Customer Cube that provides:
A CFO who walks into diligence with this foundation demonstrates the same operational discipline the moat itself is supposed to prove. And that has a direct impact on valuation.
Most finance teams are focused on proving they've crossed the gate. Whether the moat behind it actually holds is the harder question, and it's the one investors will spend the most time on.
If you want a clear read on how your business would hold up, we're happy to take a look.
