The Next Wave of AI Companies Aren't Selling Tools—They're Selling Results

Deep thoughts on AI and aspirations —— ByteDance Deep Thinking Circle

There’s a company in the US called Zoca doing something that sounds almost audacious: it serves local businesses like nail salons, massage studios, and fitness studios, charging no subscription fees—only taking payment when it actually delivers bookings. According to publicly available information, one nail salon owner went from three or four clients a day to lines out the door—even having to turn people away—after 30 days. Since launching in 2024, Zoca has served over 1,000 local businesses, completed 120,000 bookings, and generated over $10 million in revenue.

These numbers need more time to be validated, but the direction they’re betting on deserves serious attention: the AI industry’s business model is quietly shifting from “selling tools” to “guaranteeing results.”

What’s Really Crushing Small Businesses Is That No One Takes Responsibility for Results

First, consider what these businesses are selling. Local service businesses sell time: a hair stylist’s empty afternoon, once gone, is gone forever—it can’t be stored, can’t be made up later. Every unsold minute is permanent loss. So marketing efficiency is literally a matter of survival for them.

But when you calculate the time and energy involved, they’re precisely the group least equipped to do marketing. Research by Fiverr shows that 70% of small businesses spend no more than 5 hours per week on sales and marketing. It’s not that they don’t know it’s important—it’s that a hair stylist working ten to twelve hours a day serving clients simply doesn’t have time to study SEO, social media management, and ad placement. And the combination of tools on the market—booking systems, customer management, social media scheduling, ad platforms—each one requires learning. Mastering a complete suite is almost like picking up a second trade.

So the traditional tool model contains an unspoken subtext: I’ve taken your money, you’re responsible for the results. Tool companies’ revenue has nothing to do with whether the shop survives or fails. The risk is entirely pushed onto the party least capable of bearing it—the small shop pays the money, time passes, and if there are no results, they can only blame their bad luck.

What “Guaranteeing Results” Really Guarantees Is Risk Transfer

Zoca’s model flips this subtext: it doesn’t promise features, it promises bookings; no bookings, no payment.

This change looks like just a different billing method, but it actually transfers risk from buyer to seller. The tool model is “I’ll rent you a fishing rod, whether you catch fish is your business”; guaranteeing results is “I promise you a certain number of fish, if I don’t catch them I don’t get paid.” The latter means the service provider must swallow the uncertainty of delivery themselves—only those confident enough in their own system dare play this way.

The incentive mechanism changes completely as well:

When the service provider’s revenue directly equals the client’s results, both sides’ interests align for the first time. The provider won’t pile on features that look good but don’t solve problems, because adding them doesn’t generate bookings.

There’s another easily overlooked prerequisite here: daring to guarantee results is itself an expression of technical confidence. Performance-based payment means every failed delivery is your own cost—this business only stands up when the system’s success rate is high enough to cover those failures. So the results-based model naturally weeds out players who can only demo but can’t deliver consistently, leaving the market to those who’ve actually closed the loop.

Why This Bet Only Makes Sense Now

Guaranteeing results isn’t new. Ad agencies’ performance splits and sales commissions have been around for decades. Why is it only now, in the AI industry, that it could become a viable model?

Because before AI appeared, guaranteeing results was often a money-losing proposition. Human delivery costs were high, non-standardized, and not replicable—the previous client’s success might have relied on a star salesperson, and taking on the next client might mean losing money.

AI agents change the cost structure: demand monitoring, content generation, ad adjustments, pricing recommendations, customer follow-ups—all can execute automatically 24/7. The marginal cost difference between serving 1,000 shops and 100 shops is minimal. Zoca’s system claims it can capture neighborhood-level demand fluctuations—when demand for a certain service in a certain area heats up, automatically adjust the shop’s visibility and pricing; raise prices when demand is hot, run promotions when demand is cold to fill gaps. Whether this execution capability can ultimately be delivered remains to be seen, but the direction is clear: only when delivery costs drop low enough does “pay for results” have positive gross margin space for the first time.

So I prefer to understand it this way: this isn’t business model innovation, it’s the first commercial realization of AI’s cost structure. The tool model is how software should be sold in the high human cost era; guaranteeing results is what AI services should look like in the era when delivery costs approach zero.

But Guaranteeing Results Isn’t a Magic Shield

The dark side of this model needs to be written clearly too.

First, results can be faked. Bookings don’t equal walk-ins; low-price promotions can inflate booking numbers. If the system only pursues booking quantity, it might fill a shop’s entire day with low-quality clients—the shop is busy but hasn’t made money. When choosing a results-based model, you must ask “what counts as a result.”

Second, individual cases don’t prove universal applicability. A single shop’s explosive growth has no control group and might be survivorship bias; the company’s stated revenue and booking numbers are one-sided; the real net growth for merchants requires longer observation.

Third, dependency risk. When a shop’s customer acquisition is entirely dependent on one AI system, an algorithm adjustment, platform policy change, or system failure hits directly at the shop’s lifeline. This risk doesn’t exist in the tool model—if the tool fails, humans can step in; if the system stops, the customer source is cut off.

So I give four prerequisites for the results-based model: results must be measurable and distinguish real value; marginal cost of delivery must be low enough; fraud can be identified; customer repeat purchase must be high enough. Missing any one of these four prerequisites, “guaranteeing results” isn’t a business model—it’s a gamble.

Assessment

The real significance of this case isn’t another company’s fundraising story, but that a competitive direction is quietly changing: the AI services industry is moving from feature competition to results competition. Whoever can bear the risk, whoever can prove their delivery, gets pricing power.

If you’re building vertical AI, ask yourself once before launch: do I dare promise results to clients? If you dare, it means your delivery loop, data feedback, and cost calculation are all working; if you don’t dare, most likely they’re not—so go back and get the loop running before rushing to scale.

The same logic applies to buyers: when selecting AI service providers, rather than looking at feature lists, see whether they dare tie payment to results. Those who dare at least demonstrate confidence in their delivery; those who only want to collect subscription fees while being vague about effectiveness probably don’t have confidence themselves.

The next wave of AI companies will split into two kinds: those still selling tools, and those who dare guarantee results. The latter is rare now, but almost all have touched the real pulse of AI’s cost structure.