SaaS Metrics Benchmarks: thesaascfo.com/saas-metrics-…
Nine leading benchmark studies from 2025-26, reconciled into one set of numbers
An intro to startup anti-patterns: itamarnovick.com/intro-to-st…
"Studying repeatable patterns of startup failure (startup anti-patterns) is more useful than studying non-repeatable strategies for startup success.
How universities should prepare founders: paulgraham.com/prepare.html
The median B2B SaaS company now generates $193K of ARR per employee, up 29% from $150K the year before: thesaascfo.com/arr-per-emplo…
Top-quartile ARR per FTE is $279K
How to Build an Outcome-Based Pricing Plan:
thesaascfo.com/how-to-build-…
"customers pay when your product delivers a defined result.
Applications are open for the Community Foundation of Louisville’s #VogtAwards. Early-stage businesses can apply to the 26th Vogt Invention & Innovation Awards to accelerate your business. Learn more and apply at vogtawards.com by June 5th
Tax implications of LLC vs C-corp, including QSBS & Section 1202: dwthecounselor.substack.com/…
"If you hold qualifying C-corp stock for the required period, you may exclude a significant portion of your gain from federal capital gains tax when you sell. The SAFE is not clearly stock.
How to account for AI expenses in COGS vs OpEx:
thesaascfo.com/what-should-b…
"The COGS rule has not changed. The costs to deliver your product or revenue belong in COGS. Costs to build your product belong to R&D under OpEx. Other than capped R&D amort in COGS.
How ASML designs and builds the world's most complex machine, and dominates the high end of semiconductor photolithography:
worksinprogress.co/issue/the…
Why your startup should be a C corporation: substack.com/home/post/p-194…
"In 30 years of practice, I have yet to encounter a venture-backed startup that was well-served by a different choice.
The Startup Framework: A Better Pitch through Storytelling:
seobrien.com/startup-framewo…
Code Is Cheap Now, And That Changes Everything:
perevillega.com/posts/2026-0…
"I am advocating for focusing on system definition as the primary engineering activity. The skill set needed to use AI agents well turns out to be a mixture of product manager and development team manager.
Greg Langdon retweeted
Don’t forget! The Student Venture Capital Summit is this Thursday, March 27 at @SparkHaus in Covington, KY. The event is designed for students eager to learn the ins and outs of venture capital.
Free to attend, register here ➡️ luma.com/b5de3cz2
Eight moats for sustainable software companies
With all the continued talk about the SaaS apocalypse and the challenges facing software in the age of AI and “vibe coding,” last week on the podcast 20VC, startup investor Gokul Rajaram shared his eight moats for sustainable software companies. Similar to other frameworks around what makes a business durable, these are presented in the context of software and cloud.
Let’s take a look at the eight moats for software:
1. Data
Software companies with proprietary data have a clear advantage. This data can come from reaching a critical mass of customers or from exclusive agreements with third parties. Even in the age of vibe coding, where building software has less friction, some products are 10x better or only possible because of proprietary data.
2. Workflow
Workflow refers to software that is mission-critical to running a business. Examples include accounting systems or e-commerce platforms. The more deeply embedded the software is in daily operations, the harder it becomes to replace.
3. Regulatory
Companies in industries like financial services and healthcare often face heavy regulatory requirements. These include things like money transfer licenses or approvals to integrate with government systems. In some cases, only a limited number of players are allowed. In others, approval requires significant time and capital.
4. Distribution
Getting software into the hands of customers is often expensive and time-consuming. Some companies build strong distribution advantages. For example, Apple’s App Store controls how billions of devices access software. Distribution is one of the hardest moats to build and one of the most durable once established.
5. Ecosystem
Ecosystems emerge when a product becomes a platform and third-party developers build on top of it. Well-known examples include Salesforce .com and Shopify, each with thousands of integrations and add-ons. This moat takes significant time to develop and typically only happens once a platform becomes the clear leader in a large market.
6. Network
Network effects are especially powerful in marketplace-driven software platforms that connect buyers and sellers. These often combine software functionality like ratings, reviews, and pricing with mechanisms to solve the chicken-and-egg problem. Over time, the value of the network compounds faster than user growth and eventually becomes the de facto standard.
7. Physical Infrastructure
While software is often viewed as asset-light, some companies rely on physical components such as devices, equipment, warehouses, or data centers. If a product requires hardware and you have a million devices deployed, switching costs and customer stickiness increase dramatically.
8. Scale
Some software companies achieve scale across multiple dimensions including geography, employee expertise, and customer base. This makes it difficult for new entrants to compete. This is especially apparent when startups try to displace incumbents or expand into adjacent markets with well-established vendors.
When thinking about these eight moats, one approach is to assign a point or partial credit for each category and then total the score. Moats are a critical component of building a sustainable business, especially for software companies.
Entrepreneurs would do well to evaluate their ideas through this lens, recognizing that most of these moats take significant time, effort, and success to build. Achieving one or two is difficult. Achieving three or four is rare. Companies that reach four or more are best positioned for long-term, durable success.
That said, entrepreneurs should not limit themselves only to ideas that check multiple moats upfront. Instead, they should understand how software is evolving in the age of AI, where barriers to entry are falling, and be intentional about how they build defensibility over time.
The American Healthcare Conundrum:
github.com/rexrodeo/american…
"The American Healthcare Conundrum is an investigative data journalism project. Each issue identifies one fixable problem in the US healthcare system, quantifies the waste, and recommends a specific solution.
A guide to thinking about competitive risks to existing B2B SaaS companies from AI: saas-capital.com/blog-posts/…
"We developed the three candidate Rating Dimensions that we believe are close to the “principal components” of B2B SaaS AI risk as we understand it.
A more rigorous test of LLM logic and reasoning using the Car Wash Test: opper.ai/blog/car-wash-test
"I want to wash my car. The car wash is 50 meters away. Should I walk or drive?"
KY Startup Scene Navigator:
keyhorse.notion.site/Startup…
from Rron Thaci at @keyhorsevc




