SMB Software Is Dying. Winners Will Be Problem-Solvers.

01 /The Concept

Aesthetic medicine clinics and wellness centers – medspa businesses – are one of the fastest-growing markets in the US. But these clinics are losing roughly $300,000 per year on entirely avoidable problems: missed calls from prospective clients, late follow-ups with existing clients ready to book their next treatment, staff spending time on manual paperwork instead of patient care, and so on.

The root issue is that these clinics run on software originally built for basic beauty salons – businesses that offer consumer-level, not medical-level, services. A manicure appointment is transactional. Pursuing a facial rejuvenation program to slow the signs of aging is a medical service requiring planning, consistency, results tracking, and plan adjustments based on those results.

Tepali is building an "AI-native operating system" for aesthetic medicine clinics and wellness centers that need medical-grade operational software.

Three AI agents are already live in the platform.

The first handles inbound calls – explaining services, answering questions, booking appointments, and routing urgent or complex cases to a live staff member.

The second maintains a SOAP record (Subjective, Objective, Assessment, Plan) for each client – effectively a medical service history populated from visit records and provider notes. This history tracks treatment course adherence and check-in scheduling, enabling the agent to proactively reach out when it's time for a client to book their next appointment.

The third agent's job is to convert every inquiry into a booking. A caller who phoned "just to ask" becomes a contact the clinic can nurture with relevant advice and updates – until they're ready to start coming in.

Beyond these agents, Tepali includes the standard operational toolkit: CRM, electronic medical records (EMR), search and social advertising management, supply and injectable inventory tracking, staff scheduling, payroll, and more.

Tepali graduated from Y Combinator last month and is currently operating in early-access invite mode.

02 /Why It Matters

Aesthetic medicine is genuinely one of the stronger-growing markets in the US. The market reached $18 billion in 2025, and the number of aesthetic medicine clinics has grown from fewer than 5,000 in 2010 to over 15,000 today.

That kind of trajectory attracts founders. A [related review](/review/vygodnee-pomogat-professionalam) covered Moxie, which has built something similar – effectively an operating system for aesthetic practices – and raised $50.7 million in total, including $25 million last month.

Moxie's founder previously built Thinkful, an online coding education platform he sold to Chegg in 2019 for $100 million. His move into aesthetic medicine signals that he sees at least comparable upside.

But Moxie's real angle is that it does "more than software." Its mission is to help aesthetic medicine professionals launch and grow their own clinics. The target customer – a highly skilled practitioner – typically knows the medicine but not the business.

So Moxie layers in business experts: webinars, workshops, masterclasses, and one-on-one consultations with clinic owners. It doesn't just give tools – it helps people open, operate, and scale a business. The software is an instrument inside a larger service.

As a result, Moxie doesn't charge for the software subscription. It takes a percentage of client revenue, which it then works to grow – aligning its interests perfectly with those of its clinic owners.

This reflects a broader trend gaining clarity. After AI-assisted development made writing software dramatically cheaper, the value of software itself started approaching zero. The pricing pressure is structural: an abundance of tools for any use case compresses margins, while any company can now build custom software at near-zero cost – perfectly tailored to its own needs, with no vendor required.

This means startups need to stop selling bare software. They need to sell outcomes – of which the software is just one component.

This same trend appeared [recently](/review/cennost-ii-platform-teper-ne-v-ii) in a review of Rocket, which raised $25.8 million for what looks like a vibe-coding platform – but which opens with: "Most AI tools help you build faster. But none of them advise you on what to build. And none of them tell you how to win with what you've built."

Rocket's core value isn't the coding capability. It's two AI engines: one to discuss what's worth building, and another to monitor competitors and feed insights back into that conversation.

The hybrid model – AI plus human expertise – tends to perform even better. More effective help attracts and retains users more reliably, compounding the advantage over time.

03 /Opportunities

Tepali's tagline is: "your business deserves software built to 2026 standards, not 2006" – meaning AI-native software.

Hard to argue with. But the goalposts will shift again by 2027. The new framing will be something like: "your business deserves real outcomes delivered to 2027 standards – not just software built to 2026 standards"

The implication: selling bare software will carry no special value on its own. First, because the software itself won't be the differentiator. Second, because with AI, anything can be built – making the harder question what's actually worth building.

Services in 2027 will solve client problems – using purpose-built software, smart AI engines, human experts, and whatever else it takes. Software will become a background tool that, in many categories, people simply won't buy standalone. Why would they, if they can buy the solved problem instead?

This means the SMB software market – and consumer software in many categories – may effectively disappear. In its place: a services market where companies solve problems, and the software they use to do that is something they give away. The money comes from the outcome.

So: what problem do you want to solve, and for whom? What do your clients need beyond software? How do you deliver that at scale? What tools do you need to build for yourself? What people do you need beyond engineers? And how do you price outcomes rather than subscriptions?

Start answering those questions, and you'll be building for 2027 – while your competitors are still thinking about how to ship a better UI.