There's a version of the AI story that goes like this: software companies figure out agents, unbundle the workforce, and the $16 trillion services market quietly disappears into their P&L.
That version is wrong.
VixulCon 2026 made the case for why, with 110 operators in a room in Austin who had already stopped debating whether AI is real and started working on what actually comes next. If you missed out on being in the room, here are some of the major findings from our VixulCon 2026 Insights Report.
The line between services and software is gone. The firms with the clearest path forward are the ones that already know how to deliver bespoke outcomes.
The Big Flip
For thirty years, the accepted wisdom was that services companies aspire to become product companies. The keynote argument at VixulCon flipped that logic entirely.
The skills the AI-first era actually rewards — real discovery, bespoke value delivery, long customer relationships, the ability to translate a client's edge case into a deliverable — are already standard inside services firms.
They are deeply unnatural inside SaaS companies, which spent three decades deliberately optimizing them out. SaaS discovery means finding the feature that ships to everyone and ignoring the edges. The edges are exactly where enterprise customers are pushing back now.
AI just dropped the cost of serving those edges by an order of magnitude. The firms that reorganize around what is specific to each customer, rather than what averages across all of them, will win in the next decade.
The head start isn't the technology. The head start is thirty years of knowing how to listen.
The Pricing Problem
Pure T&M is broken. The math stopped making sense the moment a senior person with AI could solve in five hours what a junior person would take forty hours to complete.
The room's answer wasn't a replacement model. It was a stack:
- Base retainer: ongoing team presence
- Fixed-price capture: take the AI efficiency discount internally
- Outcome-tied fees: anchored to measurable results
- Usage-based with ceilings: per-action pricing, inference cost absorbed by the vendor
The most sophisticated operators aren't picking one layer. They're stacking them in a single commercial conversation.
The sharper observation underneath the pricing discussion: if a service isn't narrow enough to make outcomes measurable, no pricing structure fixes it. No invoice is clever enough to fix poor positioning.
The Limit Moved
Delivery used to be a limit. Sales armies and capacity planning were leverage points. AI compressed delivery. The limit moved. Reaching fifteen vertical conversations simultaneously without growing headcount requires:
- Association and industry network leverage: vertical communities sell in parallel; one introduction lands ten conversations
- Partner ecosystems: hyperscalers and foundation-model labs have one salesperson per 100,000 accounts; they need channel
- Content as inbound: the buyer arrives partially educated; thought-leadership reach beats quota-carrying reps
- Investability discipline: years of clean margin, linearity, and customer diversity to make sure your house is in order before capital is needed
That gap between how services firms currently go to market and what distribution actually requires in an AI-first era is the real work that should be your next focus.
Accountability Is the Key
What's holding the F500 back from agentic workflows isn't capability. It's accountability: Who approved the change? Who is responsible for it afterwards? Who monitors it?
That's because execution speed used to be scarce. It isn't anymore. Governance and assurance is what's scarce now. Whoever prices accountability in dollars wins the category. There's capital locked in problems enterprises cannot afford to get wrong.
What the Showcase Proved
The nine portfolio companies at VixulCon weren't pitching AI as the product. Each built AI around a vertical or a function they already owned.
- Stormatics paired 130+ years of Postgres expertise with agents that design, deploy, and tune databases around the clock.
- EZOps's Ace Dev operates inside client governance frameworks, handling ~80% of requests across 20 clients.
One pattern showed up across all nine: AI is an input into a product. It's not the product itself. The model is rentable. What they built around it is what they own. Whether the rest of the market converges upon this is an open question.
The line between services and software is gone. The firms with the clearest path forward are the ones that already know how to deliver bespoke outcomes — and are now figuring out how to do it at scale without the margins falling apart.
Join the conversation — May 21.
The field is changing fast. Business structures are changing with it.
To help founders stay ahead, Vixul is launching monthly roundtable meetups — built on the same small-group format that generated the sharpest conversations at VixulCon.
Join our first session:
Hiring & Team Model for an AI-Native Agency
May 21 · 10 AM Central
To RSVP, DM Vixul on LinkedIn.