The AI-Enabled Services Playbook

5 Emerging Lessons for Building in this New Frontier

Greetings! We’re sharing a new article by General Partner Jake Saper and Associate Rishub Nahar, offering tactical advice to founders building AI-enabled services. We hope it’s both valuable and thought provoking.

---

While SaaS startup founders have playbooks and mentors to guide them, AI-enabled services founders are charting new territory. These pioneers combine AI with human expertise to deliver faster, better, and cheaper outcomes than legacy service providers. Take Mechanical Orchard – they're using AI to accomplish in months what traditionally takes years: moving mainframes to the cloud.

We wrote about this emerging model last year. Since then, we've seen an explosion of companies adopting it, many with compelling early traction. But here's the catch: founders are trying to force-fit traditional SaaS strategies onto these service-based businesses. There’s some key similarities, but also some critical differences.

Having invested in and studied this emerging space, we've identified 5 essential lessons for building an iconic AI-enabled service company. Consider this a contribution to the new playbook.

1. Bring on a domain expert early – they are even more critical than before

In traditional SaaS, you're selling a product. In AI-enabled services, you're selling yourself. As such, domain credibility isn't just important - it's existential. Early customers need to believe in your ability to deliver results, and established credibility goes a long way towards building that trust. Domain authority also unlocks access to high-quality talent channels, which enables the rapid staffing that you may need while your AI is still maturing.

That said, technical leadership is equally vital to prevent over-reliance on manual services revenue. A domain expert paired with an applied AI expert makes for the ideal founding team – balancing credibility with a relentless push for automation.

2. Beware Mirage PMF

PMF is a different beast in AI-enabled services. Strong revenue growth and NDR can mask a lack of true AI enablement. We call it "Mirage PMF".

Real PMF in AI-enabled services requires proving you can scale non-linearly relative to your costs. To get there, your AI must drive measurable improvements in cost, quality, or speed—or ideally, all three.

3. Develop partnerships early on – they can be a key growth accelerator

Partnerships with incumbents can be a major accelerant early on for AI-enabled services. Incumbents offer immediate market credibility, established distribution, and access to proprietary datasets, which can be crucial in the early days while your data corpus is small.

To take advantage of these benefits, service startups are exploring partnership models that go well beyond the traditional revenue-share approaches of SaaS. Some startups are exploring equity share relationships with incumbents, while others are fully acquiring existing service providers. Each of these approaches have pros/cons worth considering critically.

4. Leverage new pricing models – they can help unlock higher contract values

AI-enabled services come with different pricing models than classic SaaS. If used effectively, they can help you unlock higher contract values because you’re delivering both technology and labor. Two major models exist for how to price these contracts; you need to understand the benefits and risks of each:

• Labor-Based: Price based on the number of labor hours used to service the contract. This guarantees margin early on, but risks cannibalizing growth as automation increases.

• Outcome-Based: Price based on the value of final output delivered. This aligns incentives and scales margins long-term, but may hurt profitability in the early days when AI capabilities are still nascent.

We’ve found that it is generally best for AI-enabled service vendors to start with the market norm, which is typically a labor-based approach while they are learning how to efficiently deliver their service. At the same time, they should set clear timelines to transition to an outcome-based model.

Regardless of pricing model however, founders should look for opportunities to embed recurring revenue so that they can continue to monetize clients even when their initial projects are completed. Palantir is the best example of this at scale. They deeply embed themselves in initial engagements with heavily customized software and then support that software via ongoing contracts.

5. It’s the demo, stupid!

In SaaS, demos are used to showcase product value. It’s a tried and true method that works, particularly in an era where buyers are eager to see some AI magic 🪄

In AI-enabled services, the art of the demo has been forgotten. Instead, founders default to pitch decks and talk tracks. This is understandable since the AI product isn’t used directly by the customer; the service provider wields the magic internally to deliver their outcomes.

But it’s a huge mistake. Without a demo, it can be hard for a buyer to visualize what makes you different from every other service provider with a similar pitch. And if you’re a startup competing with established/trusted services brands, it’s even more critical to showcase your magic. Don’t hide it behind the curtain.

It can feel wasteful to build a customer facing demo for a product they’ll never directly use. But it’ll be the best GTM investment you’ll make. I’ve seen AI-enabled services companies halve their sales cycles just by adding in a demo.

So stop telling prospects about your AI. Show ‘em.

---

If you’re exploring AI-enabled services, we’d love to learn alongside you. Share your thoughts, pushback, or examples – we’re all figuring out this new model together.

Special thanks to Arjun Chopra, Medha Agarwal, Wayne Hu, James Currier, Zachary Bratun-Glennon, Wenz Xing, Nic Poulos and Kent Goldman for their great feedback and thoughts on this emerging model.

---