Good morning 👋
I’ll preface this by saying nobody really knows what is going to happen to SaaS, but the next few years are going to be very entertaining for some and very painful for others.
As somebody who has spent my entire working career evaluating and investing in SaaS, I’ve been doing a lot of thinking about what all changes when we move from a world of software-as-a-service to a world of service-as-a-software.
AI makes existing software more efficient as a co-pilot
Expertise has moved from building software —> operating software
As it gets smarter, it will be able to move from co-pilot to autonomous operator
“Software” will become invisible, and core workflows will be optimized and improved automatically without the need for human involvement
Expertise will move from operating software —> controlling agents
Greg Isenberg and others have written about this. Right now, it’s somewhat of a taboo subject, but so was the internet 30 years ago.
We’ll break down what we think in today’s piece.
Today’s highlights
How the underlying dynamics of SaaS are changing
The future belongs to the obsessed
Ramp’s growth
Visuals for vibe coding
TOP
what happens to SaaS?
I’m betting my career that software investing will look VERY different in 10 years.
This article from Greg perfectly captured what I have been seeing and thinking through.
The Three-Step Implosion
Step 1: AI’s the Sidekick (Right Now)
GitHub Copilot for devs, Gamma for killer decks, Harvey for legal digs.
All of these tools are layered onto existing workflows, and existing players are not too concerned about any real disruption to their business (yet).
This still operates under the assumption that you as the user pay for the software and in exchange, you receive some sort of service.
Step 2: Agents Run the Show (12-18 Months Out)
We’re not there yet, but we’re moving towards the point where agents can work autonomously.
These agents won’t need a human babysitting them through software. Cleaning up presentations in Figma, fixing Meta ads, improving sales campaign funnels - all of this work used to require constant maintenance and human involvement.
Step 3: Software Goes Silent (2-3 Years)
Why bother with dashboards or buttons when there is no human review required.
SaaS sold us slick interfaces and workflows for humans. Agents don’t need this - they can just grab the data and use that to make decisions.
Where We Go From Here
Labor becomes the X factor: This all hinges on AI agents and usage-based products actually being able to replace the functions of given jobs. If AI agents can start eating jobs—analysts, marketers, ops grunts—the app market explodes. If not, cheap clones flood the long tail, and SaaS margins get torched.
Behemoths survive (but with a weaker offering): Enterprise giants like Salesforce won’t die fast—security, bundled licenses, and sticky workflows keep them alive.
Apps go niche or bust: Vertical apps and agents will carve out wedges—think industry-specific workflows—and either scale big or get copied quick.
VCs are forced to play both sides: Investors have to triage their existing portfolio while also while chasing AI-native companies. Expect this to create conflicts.
More winners; more losers: This shift’s bigger than the internet. The cash haul will be insane, but the graveyard of dead companies? Packed—starting a company’s a breeze, and so’s killing one.
GP / LP Mastermind
The “anti-conference” for fund managers 🌴
We don’t like conferences, so we created our own version that we would actually look forward to going to.
📆 Dates: April 24-27, 2025
📍 Location: Miami, FL
ℹ️ Other details:
Small group of 6-10 LPs, GPs, and partners
Golf and padel
Boat charter
Private dinners
No outside sponsors
We’ve already confirmed half of the guests, we only have space for a few more, and applications close in under a month.
If you’re interested, we’d recommend letting us know sooner rather than later …
TWEET
HEADLINES
Fintech startup Ramp nearly doubles valuation to $13B in secondary share sale (TechCrunch)
General Catalyst loses three top investors as the firm expands beyond venture, contemplates IPO (TechCrunch)
Anthropic triples valuation to $61.5B as AI growth soars (Pitchbook)
ICG shatters its record with $11 billion secondary fund (Pitchbook)
MEMOS
Chima: Interoperability for AI agents
Superpower: The all-in-one health membership for elite performers
Documenso: The DocuSign killer
LINKS
➕ The Complete Quantitative Guide to Judging Your Startup: A crash course in the metrics used in the quantitative analysis of startups by early-stage investors
🧾 0-$5M—How to Nail Founder-Led Sales: First Round partner Mika Asonye unpacks 6 specific tactics for learning to embrace founder-led sales instead of avoiding it
😇 As an Angel Investor, Which Qualities Do You Look for in Early-Stage Companies?: An audio clip from the CTO at Hubspot
🪟 The 8-Hour Window Most Startups Are Missing: Instead of asking users to come to their products, the most successful companies bring their product to their users
⏩ Accelerated Learning—Learn Faster and Remember More: FS discusses why we’re bad at learning, what are the sources of learning, and suggests 2 proven techniques for improving your learning
MEME
BOARDROOM
Get a FREE audit of your business 🆓
Honest take: I’ve made hundreds thousands of mistakes building this business.
GTM mistakes that slow down sales
Marketing / paid advertising mistakes that literally light money on fire
Ignoring platforms or not learning how different algorithms work that results in not getting enough attention
Not delegating or using automation that cost me hundreds of hours of my time
And a thousand others
But after making tweaks for 5+ years, you start to learn what works and what doesn’t.
Lucky for me, 99% of this knowledge of what works is now stored in my head.
Unlucky for you, most of this can’t be shared online, and it has to be applied to specific situations.
That’s why I’m doing FREE business audits for any reader who owns a business and is looking to grow.
You talk, I listen, I come up with a game plan after the call, and you decide what you want to do with that information.
Fair enough?
Apply below to book some time with me.
Thanks for reading this far and giving us a little bit of your attention this week.
Feel free to unsubscribe whenever this stops becoming valuable to you.
- Clay
(Founder @ Confluence.VC | GP @ Outlaw)
