[Un]Churned · 28 Aug 2026 · From the week of 24 August
The VC Who Coined "AI Native Services" ft. Jake Saper (Emergence Capital)
These are notes on the conversation, checked against its transcript. The episode itself has the full discussion.
In brief
Jake Saper, General Partner at Emergence Capital, talks with host Josh Schachter, SVP of Strategy and Go-To-Market Development at Gainsight, about Emergence's thesis of being early experts in emerging business models, from on-prem-to-cloud SaaS to vertical SaaS and now AI native services (AINS). The conversation covers how Jake defines AINS, Gainsight's move from selling software toward selling outcomes, the cultural and financial headwinds of that shift, the metrics and gross margin he looks for in AINS companies, and the 'mirage product-market fit' pattern he sees in pitches. His central argument is that AI may make high-margin services venture-backable for the first time, and that software vendors may need to sell guaranteed outcomes rather than tools.
For founders
- An AINS business has to deliver an existing service quantifiably better, faster or cheaper than the incumbent to disrupt it, according to Jake Saper.
- Jake Saper says that in AINS the buyer is buying the team itself, so early domain credibility matters more than it does in software.
- Jake Saper says AINS companies should run an operating dashboard tracked quarter over quarter, with KPIs assigned to individuals in the company.
- Jake Saper says the goal for an AINS business is at least 70% gross margin, with many businesses likely to land between 50% and 70%.
- Jake Saper calls Mirage product-market fit, where revenue and happy customers mask human-delivered service, a low-margin business that should not take venture capital.
For revenue leaders
- Jake Saper says per-seat SaaS pricing is dying and that AINS businesses have to teach buyers to move toward outcomes-based pricing.
- Jake Saper describes a possible future where a vendor guarantees a renewal percentage and is paid more or less depending on whether it beats that guarantee.
- Jake Saper says moving from seat-based to outcome-based revenue moves a company away from predictability, which he calls scary for CFOs.
- Jake Saper says that in most cases, similar to Gainsight, a company keeps selling its existing software while a new business unit starts to sell more over time.
- Jake Saper says annualized revenue per service FTE is the leverage metric AINS companies should track, similar to a CS coverage ratio, and that it should rise over time.
What was said 25, most useful first
The likelier path to an enduring breakout company is building an AI platform from scratch rather than rolling up existing services firms. Listen
He distinguishes this from the AI roll-up strategy, such as buying accounting firms or MSPs and joining them together, where he says most venture capital dollars in this area have gone. He presents it as his and Emergence's belief rather than a proven result.
“My and our belief is that the likelier way to create an enduring, kind of iconic breakout company is to start from scratch, build an AI platform, and then go to market with that approach.”
The per-seat SaaS pricing model is dying and will go away, but he does not think SaaS itself is dead. Listen
He expects the enduring value of software to remain in regulated industries, using pharmaceuticals as an example, where he does not think companies will vibe code their software. For large mission-critical SaaS companies, he says survival depends on how well they shift what they deliver toward outcomes.
“The per-seat pricing model is dying and will go away.”
AINS gross margin includes human labor, tokens and web services in COGS, and token spend must be split between COGS and R&D. Listen
He says tokens used to serve a specific customer count as COGS, while tokens used to build a broader platform count as R&D. He says early gross margin is messy and will be a bumpy journey, since initial AI product investment is expensive and token optimization improves it over time.
“allocating the token spend and the human labor costs between COGS and R&D is a delicate but really important distinction”
The goal for an AINS business is to reach at least 70% gross margin, and that above 70% it looks like software. Listen
He expects many businesses to land between 50% and 70%, which he says is still well above legacy service providers. He says higher margins draw higher multiples and make these companies more likely to be valued like software companies.
“I think the goal for an AINs business is to get to at least 70% gross margin.”
Jake Saper calls the biggest AINS risk mirage product-market fit, where revenue scales and customers are happy but humans deliver most of the service. Listen
He says that is just a services business and a low-margin one that should not take venture capital. He says a 30% margin AI-based accounting business could be a great bootstrapped company, or it needs a credible path to higher margins.
“And that's just a service. That's not an AI native service. And that's a low margin business that should not take venture capital.”
Jake Saper's unproven hot take is that AINS businesses can build more data-driven defensibility than SaaS businesses because they own the full outcome. Listen
He says SaaS vendors often see only portions of a task's lifecycle, while an AINS company responsible for the full outcome, such as whether a deal closed, has the full data. He says that data builds an eval set, which is what allows training open-source models to a use case and is the defensibility.
“I think that ultimately Ains businesses will be able to build more data driven defensibility than SAS businesses. And the reason I think that is because they own the full outcome.”
Building an AINS business is much harder than building software because the software platform and services organization must be built at the same time. Listen
He says you have to build McKinsey and Stripe at the same time, and that the two must integrate seamlessly, which requires more intensity than software. He says he sees many people start AINS businesses because they think they are not that technical and it is a services thing, then realize it is harder.
“actually building a AINS business is much harder than building a software business”
Jake Saper describes Gainsight as the first major SaaS company to move from being a software business to an AI native services business. Listen
He treats Gainsight as a third path alongside building from scratch and roll-ups, and says the rest of the SaaS industry will look to it to see whether the move can be done. He says Gainsight's CEO Chuck was already putting the wheels in motion when they talked during a walk.
“Gainsight is the first major SaaS company to do this.”
Software vendors may need to shift from selling tools to selling outcomes because AI can now do much of the work itself. Listen
He frames software as a means to an end, like the wheel, and says selling a tool makes less sense when AI does the work. He puts it as wanting to sell the fish rather than the fish pole.
“We're now in a world where AI can do much of the work itself. And so selling a tool doesn't make as much sense anymore.”
Jake Saper suggests technology vendors may come to look more like insurance carriers that financially guarantee an outcome, though he calls this a wild future rather than a settled one. Listen
He describes a vendor guaranteeing a renewal percentage and being paid accordingly as the rate flexes up or down. He ties this to the value of a human being accountable for warranting the outcome and fixing things when they break, which he says is why he does not see human involvement as a passing fad.
“Instead, it could look more like I am an insurance carrier and I am guaranteeing financially some outcome you're hiring me to perform.”
The biggest headwind for a SaaS company becoming an AINS business is cultural rather than technical. Listen
He argues that a company built over 10 to 20 years to build and sell software differs from one that builds and sells a services platform in how it is built, how it goes to market and how it delivers the product. He notes legacy players may be able to cross-sell into existing customers, but delivery is still very different.
“I think it's cultural.”
Moving from seat-based to outcome-based revenue moves a company away from predictability, which has large implications for the CFO. Listen
He says this creates a fear that the market will punish a company whose recurring revenue line goes down during an uncertain outcomes-based transition. He calls the CFO implications huge.
“if you move to an outcome-based model versus a seat-based model, you're moving away from predictability, which is scary.”
SaaS operators have been trained to see services as the forbidden word, and that AI may let services carry high gross margins for the first time. Listen
He notes that software's high gross margins and multiples have been tied to market, margins and recurrence, and that professional services in SaaS exist to deploy the software product. He says the core of the AINS thesis is that AI will enable services at high gross margin, which makes them venture-backable.
“the core underpinning of the AINS thesis is that AI will enable us to deliver services at a high gross margin for the first time.”
A SaaS company moving into AINS will likely keep selling its existing software and open a new business unit that grows over time. Listen
He says in most cases, similar to what Gainsight is doing, a company will not stop selling the software it currently serves. He expects the new AINS unit to likely be a new business unit that starts small and sells more over time.
“It'll likely be a new business unit that you open up, and over time, that starts to sell more.”
A private equity sponsor may give a company room to shift toward outcome-based revenue, though he is skeptical most sponsors would support it. Listen
He says Gainsight's sponsor has accepted putting some revenue at risk by moving to outcome space, which he thinks may create more enduring value. He is skeptical most PE sponsors would understand this the way Gainsight's does, but says a success could become an example for other PE firms and public markets.
“Ironically, it's possible that being housed within a private equity firm could allow you to make this shift better.”
Jake Saper expects some AINS businesses to need a licensed human in the loop permanently where regulation requires it, using customs brokerage as an example. Listen
He describes licensed professionals who must assign tariffs to every imported item in a tariff regime that changes constantly. He says AI should be well suited to the work, but a licensed person is still required, which he sees as an enduring angle.
“you're required to have this licensed person kind of in the loop, and I think that is kind of an interesting, enduring angle.”
The early team of an AINS company needs domain expertise, though it does not necessarily have to be the founding team. Listen
He describes Hanover Park, an AINS fund administration business, which hired senior, credible people from legacy service providers early. He says those hires let the company go to first customers and show software built with people who know the work.
“you want the early team to have domain expertise, not necessarily the founding team.”
Hanover Park's CEO interviewed 150 fund CFOs before starting the business to build the depth needed to talk with domain experts. Listen
He says the CEO needed that depth because he is the face of the firm, and that Emergence's own fund CFO tested him with obscure accounting questions in the pitch meeting. Host Josh Schachter draws the lesson not to take shortcuts, and Jake adds that the no-shortcuts bar is even higher for AINS than for software.
“I interviewed 150 fund CFOs before I started this business, because I needed to be able to have that depth of conversation”
In AINS you are selling yourself as a service, so buyers need to trust the early team's domain credibility. Listen
He contrasts this with software, where a buyer may not care who built the product as long as it is good. He says he is biased toward AINS companies whose early team has domain credibility, citing Mechanical Orchard, whose CEO had started Pivotal Labs.
“In AINs, you're selling yourself. You're selling a service.”
Jake Saper outlines an AINS operating dashboard in which a North Star metric breaks each service step into parts AI can perform, tracked quarter over quarter. Listen
He gives an illustrative example where a renewal response that took four hours now takes two and a half with AI plus human, with a target of two next quarter. He says KPIs should be assigned to individuals, since otherwise the measurements get moved.
“basically you create a dashboard and you track it quarter over quarter and you assign those KPIs to individuals in your company, which is a really important point.”
Annualized revenue per service FTE measures how much leverage a company gets from each service provider, and should rise over time. Listen
He gives examples of the service FTE as a renewals person, coder, accountant or lawyer depending on the service. He says it is similar to the CSM coverage ratio that CS leaders track, and that investors and boards want to see it going up.
“The second layer is what I call your annualized revenue per service FTE”
Open-weight models can make AINS token spend, which he treats as labor, exponentially cheaper and let companies customize models on their own data. Listen
He says open-weight models can be hosted privately so company data stays private, and that they are a big boon for AINS businesses delivering services more cheaply. He mentions a letter Emergence co-authored with NVIDIA and Microsoft, which his partner Gordon helped write, supporting the rise of open-weight models.
“open-weight models can make that token spend exponentially cheaper than it was even a few months ago”
AI that tracks dump truck loads could make numbers harder to fudge in construction, where he says the mafia is often a competitor to companies he backs. Listen
He gives a hypothetical in which the mafia operates dump trucks and charges per truck emptied, and says AI that tracks the actual count would make numbers harder to fudge and increase accountability. He gives this as an example of unexpected real-world implications of AI.
“you now have AI that can track the actual number of dump trucks to get emptied, they can't fudge the numbers as much”
An AI native services company provides an existing service better, faster or cheaper than an incumbent by using AI to do it. Listen
Jake Saper says the company must be quantifiably, often exponentially, better, faster or cheaper to disrupt the incumbent. He adds that these businesses move the model toward outcomes-based pricing over time, and that some services businesses already sell on outcomes while many remain labor-based.
“So an AI native services company or an AINs company provides an existing service better, faster and or cheaper than an incumbent by using AI to do it.”
Being an early expert in one business model gives unfair access to later business models connected to it, citing Salesforce and Veeva. Listen
He says Emergence invested in Salesforce in 2003, then in horizontal software companies, then in Veeva, a vertical software company for pharmaceuticals that he says is now the world's largest vertical software company. He says Emergence is the only venture investor in Veeva and that its partner Gordon is still chairman.
“This is why being an early expert in a business model gives you unfair access to the subsequent business models that are often connected to the previous iteration.”