The flower-shop roll-up has already been attempted. In the late 1990s, Gerald Stevens set out to acquire hundreds of leading local florists and create a national network. It grew to more than 300 shops. In April 2001, it filed for Chapter 11. Most people read that history as proof that flower shops cannot be consolidated. I think it may prove something narrower: Gerald Stevens consolidated ownership before it had an operating system capable of improving the businesses it bought. I entered this category accidentally. In 2018, I took over my family’s flower shop in Los Angeles, operating sinceredactedMy uncle was selling, I came from consulting, and I knew near nothing about flowers. During my first week in QuickBooks, I discovered that we were spending $13,000 a year on printer ink. Four printers ran throughout the day, printing every order, photo of the arrangement, card message and route sheet (one printer for each one). The business operated as though paper were one of its primary inputs. That turned out not to be an isolated quirk. Since then, I’ve reviewed (heavily) I think over 250 independent flower shop's P&Ls. The businesses are different, but the same exact handful of structural leaks appear repeatedly: - Wire services (likeredactedFLOWERS) taking 27-40%+ commission of orders the local florist still has to design, fulfill and deliver. - Flowers passing through way too many intermediaries before reaching the shop, let alone the end-client. - Purchasing driven largely by instinct, with spoilage and stockouts accepted as unavoidable. - Marketing held at 1% or 2% of revenue rather than treated as a capability that can create demand. - The owner serving simultaneously as head designer, production manager, salesperson, buyer and institutional memory. These are not minor efficiencies. Together, they determine whether the shop is a scalable company or simply a demanding job for its owner. They are also way more addressable today than they were twenty-five years ago. Ordering, production, purchasing, customer data and delivery can now run through one beautiful operating system. Local demand can be deliberately developed and measured. Forecasting can inform stem purchasing. Farms are connected way more directly to stores. Marketing, technology, training and customer service can be shared across multiple locations. But the operational inefficiency alone does not prove a roll-up thesis. The stronger evidence is what happens when those capabilities are applied inside a real flower shop. Today, one of our mature company-operated locations generates more than $7 million in annual revenue at approximately 19% EBITDA. Across the broader French Florist system, we now have more than 20 locations open. I do not present the $7 million location as a typical result. It has decades of local history, strong market density and years of operating-system development behind it. I would not underwrite a newly acquired shop to reproduce that performance immediately. And having more than 20 locations open does not, by itself, prove that a local acquisition strategy will work. But our very first shop we did this with in another state did over $900k in profitable revenue in Y1, and that unit was built from the ground up. I feel those facts do move the thesis beyond theory. They demonstrate that a flower shop can become a substantially larger and more profitable business than the category’s traditional unit economics might suggest. They also show that a common operating platform can be deployed across multiple markets and owners. The part we are now most interested in is the ownership transition. Many florist owners know their businesses extraordinarily well. What they often lack is the time, capital, infrastructure or desire to rebuild the company while continuing to produce orders six or seven days a week. We have found a change in ownership creates a rare opening to change the system itself. That suggests a different model from the national roll-up attempted in the 1990s: - Acquire a respected local florist with real customer relationships and underlying demand. - Preserve the people, craft and local goodwill that made the business valuable. - Replace the structural leaks with stronger technology, purchasing, marketing, training and operating infrastructure. - Build professional management so the economics no longer depend on the former owner working sixty hours a week. - Once the first business is stable, acquire nearby florists and share production, purchasing, delivery, customer service and marketing across a dense local cluster. In other words, do not acquire hundreds of stores and hope an operating system eventually emerges. Prove the operating system first on a small scale, then continue to scale what's working. For anyone underwriting a florist acquisition, I believe the most important questions are: - How much demand belongs to the shop itself rather than wire services, marketplaces or a handful of referral sources? - Does design quality survive without the owner personally touching every major order? - Is there enough geographic density to share production and delivery across future acquisitions? - What is EBITDA after paying market compensation to replace everything the owner currently does? Disclosure: I am not neutral about this. We built French Florist’s operating system and now franchise it. We have more than 20 locations open (doubling this year, predominantly from existing owners expanding), and we are looking for a small number of acquisition-minded operators in A-level markets who want to lead local florist roll-ups with our infrastructure behind them: technology, purchasing, marketing, training, operational playbooks and an established consumer brand. This post is an attempt to get an honest read from experienced searchers: do you see the same consolidation opportunity I do, or am I missing something fundamental? I am especially interested in the skeptical cases. The strongest objections are often the most useful because they expose weak assumptions, overlooked risks and places where the model needs to become stronger. We are actively building in this category, and I would rather pressure-test the thesis now than protect it from criticism. All perspectives are genuinely welcome, skeptical to bull case, or somewhere in between.