Jeremy Horowitz | Buying $10-100m eCom Brands

@drewfallon12 This is a fascinating dilemma that every role is facing. All of my engineers are saying the same thing. How much are you vibe coding that at this point you don't understand? We either have to have faith the agents are actually smarter then us or stop using them.

this is basically the way it goes *ask claude for something sophisticated “oh wow! *validate data “oh wow” the thing that makes me nervous is that i don’t think people are validating. how could they? anyone who tells you they are is lying like look at this formula claude made 😂 =IF(OR(IF($CR$1=1,ROW()-316>29,(ROW()-317)*$CR$1+1>30),IF($CR$1=1,ROW()-316+COLUMN()-5+1,(ROW()-316+COLUMN()-5+1)*$CR$1)>30),NA(),IF(IF($CR$1=1,SUMIF($BN$3:$BN$200,ROW()-315,$BP$3:$BP$200),SUMPRODUCT(($BN$3:$BN$200>=(ROW()-317)*$CR$1+1)*($BN$3:$BN$200<=(ROW()-316)*$CR$1)*$BP$3:$BP$200))=0,NA(),IFERROR(SUMPRODUCT(IF($CR$1=1,($BN$3:$BN$200=ROW()-315)*1,($BN$3:$BN$200>=(ROW()-317)*$CR$1+1)*($BN$3:$BN$200<=(ROW()-316)*$CR$1))*IF(INDEX($F$3:$BM$200,0,IF($CR$1=1,ROW()-316+COLUMN()-5+1,(ROW()-316+COLUMN()-5+1)*$CR$1))>0,IF(INDEX($F$3:$BM$200,0,IF($CR$1=1,ROW()-316+COLUMN()-5+1,(ROW()-316+COLUMN()-5+1)*$CR$1))<$BP$3:$BP$200,INDEX($F$3:$BM$200,0,IF($CR$1=1,ROW()-316+COLUMN()-5+1,(ROW()-316+COLUMN()-5+1)*$CR$1)),$BP$3:$BP$200),0))/IF($CR$1=1,SUMIF($BN$3:$BN$200,ROW()-315,$BP$3:$BP$200),SUMPRODUCT(($BN$3:$BN$200>=(ROW()-317)*$CR$1+1)*($BN$3:$BN$200<=(ROW()-316)*$CR$1)*$BP$3:$BP$200)),NA()))) if you haven’t built the model from scratch by hand and know what the numbers are supposed to be then you’re not going to catch it it’ll get there, just needs some time

@zachmstuck Congrats Zach. Awesome to see. And great group to be joining.

I started a little marketing agency 7 years ago called Homestead, named after the street I grew up on. I had the vision to build something special and looking back now, I think we did exactly that. Today I’m excited to announce that Homestead has been acquired by Verndale (also recently acquired VAAN Group) in an effort to create a DTC growth platform built to support brands across the entire commerce lifecycle. I’m extremely grateful to my business partner @_RileyTrotter for all of the time he’s given this company and all the incredible work that he’s done. Riley took over as CEO a few years ago and Homestead wouldn’t be where we are today without his leadership and pursuit of greatness. I can’t say enough about my other partners @kellybird__ and @jsappington. These two have shown a level of commitment to building this company that is unmatched. Watching them both grow as leaders over the last few years has been nothing short of inspiring. To all our current and past team members, thank you for always giving Homestead and our clients 100% effort. Will forever be grateful to you all for helping us get to where we are today. To all of our clients, thank you for trusting in us to help grow your businesses. This is not an easy task and we have never taken it lightly. As for me, I’m officially stepping out of the agency business, but as most of you know I haven’t been involved in the company for some time as Riley/Kelly/Jacob have been and will continue to lead Homestead for years to come. Expect to see more updates on the brands I’ve been building in weeks to come.

@RomanEcom Why build a CRM when you can build a system of Agents? Also don't think they all suck. You haven't given Coco a chance yet.

I want to incubate and build a WhatsApp first CRM that can be as high quality as the Postscripts of the world for SMS. Everyone sucks. Like every single platform. No one has solved this DM if you’re a founder who wants to build this. I can probably get you the first ~50 large enterprise clients

@Seanfrank As someone actively looking to buy eCom brands forward looking Rev/earnings won't happen. And while SaaS has gotten hit the $$$ went to AI winners/ cash. With wars, tarriffs, inflation Consumer (Esp SMB/MM) is still considered a huge risk and PE/real money is waiting it out.

Everything you need to know to sell your brand: Ecom valuations are going to rise again in the next 24 months. Here is everything you need to know if you want to make money during this window. 1- How are companies valued? Almost everything will be sold on a multiple of earnings. If you are small, you will use "Seller discretionary earnings". The money you as an owner can take every year. If you are bigger, it will be a multiple of EBITDA. Right now, a mid growth brand (growing 20%), in an average category (clothing, not beverage), with a solid team and clean financials, with 10 million in trialing 12 month EBITDA could get 8x-10x. That means they will sell for 80 million to 100 million. What makes the multiple go down? - not at all time high revenue (shrunk one year) - weak margin / too discount focused - bad channel mix (too amazon, too wholesale) - platform risk (tik tok shop) - tariff risk (all china) What makes the multiple go up? - good product mix (multiple hero items) - very diversified revenue (omni) - best in class margin - very strong LTV - HIGH MER But this brand will not trade on REVENUE. If someone told you that, they lied to you or are old. PE buys profit. The only things trading on revenue will be beverage or the most elite CPG. Gruns could get a rev deal done. Ridge cant. But valuations, multiples, will continue to go up. why? 2- Interest rates are coming down. SAAS is being crushed. All this money needs to seek a return and ecom/brands/cpg have been suppressed for a while. It is a natural ebb and flow- Dollars are going to come in, valuations are going to go up. Covid highs we would see this appeal brand go for 10-14x EBITDA. In 2022-2025 the lows got as bad as 4x. So we have already recovered a lot from the low, but we are still 50% off the peak. Do I think we will see 20x EBITDA deals again? Or 3x revenue deals? no. But I think the asset class is still underpriced and will turn out to be pretty ai resilient. 3- Understand timing and terms. Part of the rise in valuations will be doing deals on forward looking projections. The past 4 years every deal was being done on trailing 12 months. I suspect in the next 12 months, more deals will be done on forward or current year numbers. SO doing a deal in June, but getting credit for the projected ebitda for the entire fiscal year. The flip of this, in the past 4 years most deals have been very preditorary. Lots of coupons, dividends, participation snuck in. I saw deals being done at 10x EBITDA, but its a minority deal where the buyer basically is guanreteed a 4x on their money... Someone was buying 25% of something but getting so many protections its like they owned th3 whole thing. Thats all going to go away 4- Hire help. If you are small use a broker or a lawyer. For 50k a lawyer can save your ass on a 3 million dollar deal. If you are bigger, 3-5m in ebitda and up, hire a banker they will charge 1-5% of the entire deal, but every banker pays for themselves. they will find more buyers, tighten the screws, and guarantee a deal closes also- if you want to sell, dont miss the window. the windows are open 2 years every 6 years. Eventually everything goes to shit and you gotta wait out the bad times anyway- this is actually a series and this is part 1 of 5 suckers see you soon

@stuartchaney Preach for all the normies groaning in the back.

a reminder that we are in a bubble on this app most people have never heard of Anthropic, OpenClaw or an MCP Julie from HR is not about to vibe code a custom CRM on a mac mini and cancel Salesforce technology moves fast but changing the daily workflows of a workforce takes many, many years

@drewfallon12 But what if you don't work anymore and Claude Browser does everything for you?

every single time you do something for work, you should document step by step how you're doing it, record a loom video, and save the transcript and SOP for it to be an openclaw. As you work, write detailed requirements on how to automate that task, and youll never do it again

@OisinO Especially in the public markets.

When everyone rushes to AI I can’t help but wonder how many non AI opportunities are being overlooked.

@c_gro They don't need to leave their jobs to do this. As an employer I'd rather see how they're using it at their current job.

I've now spoken to 3 extremely smart high-performing people this week who all mentioned leaving their jobs in April to study AI all summer before their next gig.

@BillDA @nateliason But my mission control looks so sick! I cloned Linear's UI!!

@nateliason I just hooked mine up to Trello - each task is a card, they comment on the cards, pass between agents with tags Why do people need custom mission control again?

Yes if you’re spending time building a Mission Control / dashboard you’re not using OpenClaw properly imo

@codyplof @TaylorHoliday This is an "and" answer. The entire team should be able to build to solve their problems and scale themselves. You'll need a new version of an Ops role that will be responsible for building, maintaining and scaling infra for everyone so you don't have 100 diff set ups.

Interesting. Idk if they should be coding tools for me. Here's how I currently think about it, curious what you think: 1. I am going to build stuff because A) I love it, B) I think I have a lot of useful knowledge I want to permeate through the org via Claude skills and there is tremendous leverage in doing it and C) sets the culture from the top 2) I expect probably the top 5% to seriously vibe code nd build stuff. My Director of Data and AI was up with me until midnight building stuff last night. My IT systems manager built a shopify app that will save us $15k/year in Netsuite licences his 2nd day of having Claude access. 3) I do think it will be the role of one person or a few who will build stuff for others as part of their job. I think there is leverage in it at big enough company size. For example, my guy above building a retail forecasting model for our VP of Retail. Or I built a JD generator and recruitment engine for Head of People. So I have one guy currently, but going to bring on another. I expect a few others to build some smaller things. But idk if I expect everyone to use claude code for actual software dev. 3) I do expect our entire entire team to use Claude/ AI seriously and become high leverage. Anyone that doesn't might not have a place here in the future. Curious how you are thinking about "people building stuff for me"?

Ok I've decided I'm gonna hire 1-2 people to vibe code stuff for me full time. I'm still gonna do it but the opportunity is too big and I have too many ideas. Want someone young and hungry. Can start part time but i'd love it to become full time. Hit my DMs if you are interested.

The through-line this week? AI tools crossing from "interesting" to "how I operate." If you're still treating these as experiments, you're already behind. Follow for the weekly DTC Twitter pulse ⚡

5/ Davie's debate on what a talent moat means in the age of AI. x.com/daviefogarty/s…

5 posts from DTC Twitter this week that every brand operator needs to see. 1/ Bill's "TikTok/Reels are weapons-grade dopamine hijacks that will break your brain over time" is the perfect description. x.com/BillDA/status/…

@ron_ecomm Scrape every checkout on the internet. ID the platform. use a mix of signals to estimate Rev. Not perfect in every use case, but directionally accurate at scale.

@Jeremy_Horowitz How do they get the data?

7k+ Shopify stores that do $10m+/yr. The Breakdown: 1) US: 82% 2) UK: 5% 3) India: 3% 4) Canada: 2% 5) Australia: 2% Germany, France, Italy, Hong Kong + the Netherlands all trail @ 1% each. The US is still the only real Shopify market. Source: Storeleads

This will go down as @Google's greatest miss in history. Now @OpenAI has a legit B2B use case. Gsuite will become the infra everyone runs agents on. But won't be the agent.

Stitch Fix ($633m) is dead. Rev declining for 4 straight years. Losing $120m+/yr. -95% from ATHs. The ""Fashion is tech"" biz model a complete lie. They shut down their Individual item program. Marketing spend is flat. They don't disclose active customer accounts anymore.

The Affordable DTC Brand got slaughtered in 2025. The CAC + COGS + Fulfillment math doesn't math. 10 yrs ago, affordable luxury was the DTC play. MVMT Watches, Pura Vida, Casper. Non-premium biz gets crushed. Luxury/Premium brands are the only ones crushing it.