The next trillion-dollar companies won't sell software—they'll sell outcomes, rebuilt from scratch with AI doing the work
The core thesis is that some of the most valuable companies of the next decade will not be software businesses at all—they'll be AI-native services companies in fields like tax, insurance, legal, mortgages, and healthcare, delivering outcomes directly to customers rather than tools for customers to use themselves. The opportunity is framed as software-level margins (50%+) applied to a market two to three times larger than traditional software, displacing incumbent vendors rather than asking buyers to change behavior.
The best target markets share four traits: low trust (work is already outsourced, buyers care about the result not the process), low per-task judgment (work can be decomposed without requiring constant human discretion), high intelligence threshold (hard enough that models plus humans are genuinely needed), and regulatory complexity that raises barriers and moats for new entrants. The "Sam Altman test" is offered as a key strategic check: as models improve, does the service get stronger, or does the model commoditize the company?
Founder attributes that matter most are domain fluency, model fluency, and operational rigor. Critically, the human in these businesses is the customer interface, not the product—the product must make each human non-linearly more productive. If headcount and revenue scale in lockstep, the model is broken. Variance, not speed or cost, is what causes customer churn.
Several execution traps are flagged explicitly: - The early demand trap: signing too many pilot customers overwhelms delivery capacity and kills the ability to build product - Zero or negative margin pilots: tempting for learning but dangerous as a habit - Acquiring legacy services businesses to bolt on AI—legacy culture, metrics, and expectations don't transform just because AI is added on top - Physical/equipment-heavy operations where software margin math breaks down
The bottom line: price on outcomes or per-unit (per return, per claim, per loan), cap early pilots ruthlessly, and treat the automation of the process itself as the product. The window to build category-defining AI services companies is open, but the playbook is distinct enough from SaaS that founders who treat it like traditional software will likely fail.
AI-native service companies that deliver outcomes directly—rather than selling copilots to human workers—represent a massive emerging category. The playbook differs sharply from traditional SaaS, demanding domain expertise, model fluency, and tight operational discipline. Getting the market selection, pricing, and early scaling decisions right is critical.
Some of the biggest companies of the next decade will be software businesses at all. They'll be services companies like insurance carriers and law firms rebuilt from scratch with AI doing most of the work.
The best markets for AI services have four new, pretty unique traits: low trust (work already outsourced), low judgment at task level, high intelligence threshold, and regulation can be good.
If you can break the work into pieces and every piece needs a human exercising actual judgment, you can't really scale.
The overall work has to be hard. Hard enough that models plus humans are needed to actually deliver an outcome the customer accepts.
Regulated industries have higher expectations and legal accountability that raises the bar and the moat for founders.
You need to know what frontier models can do today and design the product to ride the curve as they get better. There is no substitute for great tech here.
With AI services, the human is the interface of the customer, not the product. The product helps the human scale their work non-linearly.
Customers will fire you for variance faster than they will fire you for being a bit slower or bit more expensive than the incumbents.
If revenue scales just in line with the number of humans you add, you'll have major problems.
Pricing is harder than traditional software, cuz you're not competing with other software providers, you're competing directly with the cost of labor, internal or outsourced.
Traditional services firms top out around 30% margins. The bet on these services companies is that AI operating leverage gets you closer to software margins, say 50% plus, on a market that's two to three times bigger than software.
You just can't acquire product market fit. Legacy service businesses are, you know, legacy. They have different expectations on metrics, hiring, and performance. Adding AI on top of that doesn't immediately change any of those realities.
Anything involving equipment and on-site labor is where to be careful. The software margin math doesn't apply when you own and operate physical things.
Ask yourself sincerely, are you using humans cuz the work genuinely needs judgment or you compensating for product gaps?
The early demand trap: it's easy to sign up a lot of pilot customers when you're just starting out and have nothing, but it can quickly overwhelm your ability to serve them, and you won't be able to build the product to scale.
Be deeply suspicious of zero margin or negative margin pilots. They're fun to learn from, but it's really dangerous to get hooked on those.
There's a temptation to buy an existing services business, add some AI on top, short-circuit the revenue. This is generally a trap.
AI native service companies provide the outcome to the customer versus build a co-pilot that the customer uses internally.
Low trust means the work is already outsourced and the customer cares about the final product, not how they got there. You're displacing a vendor, not asking the customer to do something fundamentally different.
The Sam Altman test: as the models get better, does your service get stronger or does the model itself commoditize you?
For AI services specifically, there's three attributes that all the best founders share: domain fluency, model fluency, and operational rigor.
Pricing options include per unit pricing (per return, per claim, per loan) and outcome-based pricing. Cost plus pricing and straight-line undercutting are strategies to avoid.
Cap your first pilot customers to a small handful. Resist the temptation to sign too many too quickly.
“Automating the process is the product.”
Shared with Earmark · earmark-ai.com