Загрузка видео...
Не удалось загрузить видео
AI service firms are commanding 30x multiples right now. Yes, thirty. That's why a16z, Sequoia, and YC are chasing services, not SaaS. Most agencies will see this and reach for the wrong move. They'll keep selling hours, bolt on AI, and cut headcount to pad the margin. But that's... show more
415,590 просмотров • 3 месяцев назад •via X (Twitter)
Комментарии: 36
the 30x makes sense when you realize what's being priced. SaaS sells access to a tool, AI service firms sell decisions already made. the moat isn't the model, it's the judgment layer that knows which output to ship and which to kill.
For more like this, level up your marketing with 14,000+ marketers and founders in my Leveling Up newsletter here for free:
why do you need VC investment if you can organically scale your service business? can't imagine why lawyers, insurance brokers and other service providers should seek predatory VC capital. all professional service providers that I know already use LLMs, Claude Cowork and other cool tools. the concept of services-as-software was outdated even on the day it was born. you tend to underestimate AI adoption
"Services are the new software" so every agency founder fired their junior team, called their retainer a "managed growth loop," and waited. Same spreadsheet. Different font.
Great read!
This is the part I agree with most. The opportunity is not “agency but with fewer people”. It’s rebuilding the full workflow: insight, strategy, creative, approval, execution, reporting. That’s where AI actually changes the business.
Selling hours was always the fragile model. AI just made that fragility visible faster. The ones with performance-based agreements are actually in a better position than ever.
Built one! Launching today.. :)
by definition, arent those people doing the services requiring various saas apps to do their job ?
30x multiples on a services firm? those compress fast once capital floods in
Yep
The theory makes sense. I want to understand more specifics on the implementation. Imagine a simple scenario, there are 3 engineers running a AI services firm. The total cost to the company is salary + token cost + general company cost. How should or rather on what basis should the customer be invoiced? How does one think of managing working capital here?
the 30x isn't really for services though, it's for the firms that productized delivery so the next client costs almost nothing to serve. selling hours can't get there no matter how much AI you bolt on.
@SegTax is the AI native Cost Segregation services firm
サービスがSaaSを凌駕する理由、サービス企業がAIに注力している理由。営利を拡大するために効果的な成長ループがなぜ重要か、そしてAIサービスビジネスを成功させる方法。サービスビジネスがソフトウェアとして機能するための方法について学びましょう。
@grok summarize this entire video and provide actionable items, throw away the fluff
The 30x isn't for AI. It's for the firm that productized its judgment layer so the output stops depending on which senior person is in the room. Bolting AI onto hourly billing just makes the same business faster at being cheap.
Fair point on bloated agencies. But I’d be careful pretending the agency was only the tools. Clients don’t just pay for output. They pay for someone to know when the idea is weak, when the work is risky, and when the shiny AI result is actually rubbish.
Services-as-Software 💯
the agencies that make it aren't leading with 'we use AI.' they're leading with outcomes. the system is how the outcome holds at scale. lead with the result. not the infrastructure.
Nobody is paying 30x revenue or profit for a services firm. 2.5x revenue is an outlier.
The agencies that win won't sell hours or AI. They'll sell the outcome and let AI carry the cost. Cutting headcount to protect margin is what you do when you're defending, not building.
are you stupid or is this a joke? services as a software is saas
My bias: the multiple only makes sense if the delivery loop survives a new client without a senior person babysitting it. Cutting hours is margin. Reusable intake → execution → approval → reporting is the software part.
This is the part most agencies miss: AI doesn’t make services valuable by cutting labor. It makes them valuable when the delivery process becomes repeatable, measurable, and tied to outcomes. The model is not “agency + AI.” It’s service delivery redesigned like software.
You’re ahead of how so many are approaching this. Would love to chat about this. I agree on the loops. I built the agent product for this. It’s called LatchLoop.
Huge trap and here’s why: The AI Independence Platform whose purpose is to ensure any forward-deployed engineering team is replaceable is the next spend ticket spend for enterprise. Core components: * agent registry * workflow registry * evaluation registry * observability * architecture documentation * governance layer * deployment specifications
Every AI cycle starts with "software is dead" and ends with people rediscovering that execution still matters
brilliant breakdown @ericosiu! very much aligned on ai & services as a model to empower employees as opposed to replacing them!
services are the new software
The wrong move is selling AI hours with better margins. The real move is execution accountability. FinalBossTech. DM me. 💜
agreed
30x for services is the market pricing an arbitrage: AI collapsed delivery cost but clients still pay output prices. Winners keep pricing the outcome and pocket the spread. Selling hours with AI bolted on is just deflation with extra steps.
A text post, a video and an article in a single post. Xmaxxing
Hot take: most "AI service firms" getting 30x multiples are just agencies with an AI pitch deck. The multiples will compress once investors realize "we use AI" ≠ "we are a software company." Real services-as-software means the AI does the work, not the humans. We're not there
Wild that the smart VC money is rotating out of SaaS into services. Total reversal of the last decade.
