Hey - It’s Nico.

Welcome to another Failory edition. This issue takes 5 minutes to read.

If you only have one, here are the 3 most important things:

• 80 Acres Farms, a startup building vertical farms, has shut down — learn more below.

• The emptiest distribution channel with 800M users

• Founders are using AI to quickly make tiny SaaS and sell them for very cheap — learn more below

This Week In Startups

🔗 Resources

The emptiest distribution channel with 800M users

10 ways Clay's GTM engineers use AI to accelerate sales

Meet your biggest competitor (👋 Claude)

Tech’s Search for Legitimacy

📰 News

OpenAI says it slowed Astra model development over security concerns

Cloudflare launches Kitesurf, a browser built for AI agents

Claude will now add invisible watermarks to text

Mark Zuckerberg wrote a controversial essay about the future of AI

💸 Fundraising

HappyRobot, an AI platform for automating enterprise operations, raises ⁠$150M Series C funding 

Ambrook, an accounting and payments platform for independent businesses, raises ⁠$30M Series B funding 

Buzz Solutions, an AI platform for electric grid inspections, raises ⁠$20M Series A funding 

Valar Atomics, a nuclear startup developing mass-produced microreactors, raises ⁠$1B Series B funding 

Fail(St)ory

80 Acres Farms

80 Acres Farms built indoor farms that could grow lettuce, herbs and tomatoes all year, close to the supermarkets selling them.

By August 2026, it had raised more than $350 million, supplied thousands of stores and claimed several farms had profitable unit economics. Then a planned sale fell apart, the company ran out of funding, and operations stopped almost immediately.

What Was 80 Acres Farms:

80 Acres grew produce inside warehouses using stacked growing beds, LED lights, sensors, automated handling and tightly controlled water, temperature and humidity. Crops could grow every week of the year without waiting for the right weather or season.

That gave supermarkets a fairly simple benefit: predictable local supply. A retailer in Ohio could buy greens grown nearby instead of bringing them across the country, with fewer weather shocks and less time between harvest and shelf.

80 Acres sold herbs, microgreens, tomatoes and salad blends, then added salad kits, dressings and prepared meals. Kroger became a major customer, and products eventually reached chains including Whole Foods, Meijer and The Fresh Market.

The farms also came with a sustainability pitch. 80 Acres said they used 95% less water per pound than conventional farming, ran on renewable electricity and grew without pesticides.

Behind the produce sat a large technical operation. Its GroLoop system tied together lighting, climate controls, crop data, forecasting, inventory and automated equipment. Running one of these farms meant managing biology, software, machinery and a food production operation under the same roof.

The farms were expensive to build and run. Indoor growing meant paying for lighting, climate control, automation, maintenance and labor every day, so each facility needed high, steady output to make the economics work.

80 Acres said it had reached that point at several farms. By 2024, it reported selling about 2,000 metric tons of produce and claimed profitable unit economics at the farm level.

The company then grew far beyond those individual farms. It added new facilities, acquired other businesses and merged with Soli Organic, expanding to more than 17,000 retail locations.

That growth brought a much larger cost base around the farms, including corporate staff, R&D, technology and a national operating network.

The Numbers:

• 🏗️ Founded: 2015

• 💰 Funding: $350M+ raised

•  🥬 Produce sold: ~2,000 metric tons in 2024

•  🏪 Retail footprint: 17,000+ locations after the Soli merger

•  👥 Jobs affected: 800+

Reasons for Failure: 

• Farm profits could not carry the company: 80 Acres said several farms had profitable unit economics, but the company still had to fund R&D, software, sales, headquarters and other central costs. By March 2026, management acknowledged that the overall business remained unprofitable.

• Expansion kept adding fixed costs: New farms meant more staff, equipment, maintenance and working capital, whether each site was running near capacity or not. Expanding into Georgia, Texas and Colorado increased the amount of cash the company needed to keep the network operating.

• The business became too broad: 80 Acres added salad kits, dressings, prepared meals, plant genetics and farm technology, then merged with Soli Organic and added roughly 1,000 employees. That left profitable farms supporting a much larger and more complex operation.

• The company ran out of financial room: By summer 2026, 80 Acres expected a sale of the company to provide the cash needed to keep operating. The buyer withdrew on August 2, and WARN filings say there was no other funding available, forcing the shutdown the next day.

Why It Matters: 

• In capital-heavy businesses, “profitable units” can be misleading if each new unit still depends on a growing layer of central R&D, software, sales and management that never gets cheaper.

• Buying distressed assets can look capital-efficient, but cheap capacity is only useful when you already have enough local demand to keep it full.

Trend

Micro-SaaS Flipping

AI has made it possible to build a tiny software product in days. A growing number of founders are taking the next step: get a few users, some revenue or a burst of traffic, then sell the whole thing for a few thousand dollars and start again.

Why it Matters

• A $5,000 exit means more when the product took a week to build. Founders can run many more shots, recover money from the ones that show some demand, and move on before turning every experiment into a company.

• Buyers increasingly pay for the head start around the code. Users, search traffic, a working payment setup, customer feedback and a clear niche can save months of uncertain work. AI keeps reducing the time required to rebuild software, which makes those other assets more important.

• This creates room for a new type of founder. Someone can specialize in finding small opportunities and getting products to their first signs of demand, then hand them to operators who are better at sales, support and long-term growth.

The Signals

The example that pulled me into this was Livestamp.

Its founder says he had no app-building experience. He saw a photo effect going viral, used Claude to figure out how it worked, ChatGPT for the visual assets and Codex to build it. Three days later he had a working app.

Then his demo went viral, the app passed 10,000 downloads and roughly 1,000 daily users, and a Vietnamese CEO offered to buy it. Six weeks after starting, he sold for $8,000 with zero MRR.

Once I started looking for similar cases, they were surprisingly easy to find.

In June, an AI blog generator doing $100 MRR sold for $2,000 in 12 days on TrustMRR. A few weeks later, an AI virtual try-on app doing $300 MRR sold for $6,000. Around the same time, somebody sold a simple Stripe tool for recovering failed payments for $1,300 despite having zero revenue.

These are tiny exits from tiny products. That is exactly the behavior I’m interested in. Founders are getting products to a very early proof point and finding someone willing to take over from there.

This is Happening a Lot

The marketplace data is moving in the same direction. Flippa says SaaS transactions grew 73.5% in 2025, while 37% of its buyers made multiple acquisitions.

Buyer demand is still growing. In H1 2026, Flippa had 123,022 active buyers, up 18% from a year earlier.

And a new generation of marketplaces is forming around much smaller deals. TrustMRR only launched in October 2025 and now says it facilitates more than 20 acquisitions a month. It already gets around 200,000 monthly visitors, with many products on the marketplace making only a few hundred dollars in MRR.

That is the part I find interesting. There are now enough tiny software products for sale, and enough people willing to buy them, that a marketplace dedicated to micro-SaaS can clear several deals every week.

The Trend

I’ve started calling this micro-SaaS flipping.

The model is pretty simple: build, or occasionally buy, a tiny software product, get enough evidence that people want it, sell it to another operator, then move on.

People have been buying side projects and tiny SaaS businesses for years. SideProjectors was doing this in 2013. MicroAcquire made small startup acquisitions much easier starting in 2020. Founders were even running “build and sell a SaaS in 30 days” challenges before vibe coding existed.

AI is accelerating the cycle. A founder can now get from idea to working software in days, run more experiments and reach the point where there is something transferable with much less money upfront. On the buyer side, AI also makes an unfamiliar codebase cheaper to understand and maintain.

I think that combination is creating a proper second-hand market for software.

The useful mental model is that AI is turning software into inventory. A founder can manufacture a small product, prove some demand and sell the asset to someone who wants to handle the next stage.

Most of these exits will stay small. That is part of what makes the model possible in the first place.

Help Me Improve Failory

That's all for today’s edition.

Cheers,

Nico