Building a $200M Software Company in an Industry Nobody Wanted: Todd Saunders (Broadlume)

Todd Saunders spent a decade turning Broadlume into the software backbone of the flooring industry. He and his co-founder did not start there. They launched an ad tech company called AdHawk in 2015. By 2019 a third of the revenue was coming from flooring retailers they had barely noticed. So they bought the website company feeding them those customers, pivoted hard into flooring, rebranded as Broadlume, acquired eight companies along the way, and sold the business in 2025.

Flooring is about as unglamorous as software markets get. The buyers are family retailers who still run on pen and paper. The systems they replace were written in COBOL thirty years ago. Nobody in Silicon Valley is racing to serve them. That is the whole point. The Physical AI thesis is that the real openings sit in the industries nobody wants, where trust is the moat and labor is the bottleneck. Flooring is a clean test of it.

Todd sat down with Jay to talk about what actually wins in these markets, and his answer is not the product. It is brand, relationships, and showing up in person year after year. They also get into roll-ups, the build versus buy math in an AI world, and why he thinks AI go-to-market tools are Twitter hype for vertical software.

The conversation

Jay: You built this over ten years, sold it last year. Seeing all the AI tools now, what would you do differently?

Todd: Almost everything. We sold it right before the true AI era. Three places I'd have used it. First, M&A. We bought eight companies. Not for a customer-facing product, our customers are still coming from pen and paper. I'd have used it for internal transformation. It took us forever to get off multiple Salesforce instances, RevOps platforms, payroll. AI could have done that faster. Second, budgeting. I'd sign only annual contracts and re-look at every tool each year, because today you're integrated with X and tomorrow Y ships a better model. Third, our ERP. It had been around thirty years, written in COBOL. Moving a customer from another ERP onto ours took four or five months. AI could have taken that down to weeks. So for us it would have been a very internally focused AI organization. Our customers are not looking for the cutting-edge AI solution.

Jay: In a legacy industry your competition is not the next sexy tool, it's some person doing it in Excel. So how do you get thousands of buyers who ran their own system onto one platform? How do you build that trust?

Todd: 51% of the reason people signed up for us was our brand, and 49% was our product. My co-founder and I were non-technical, but we were a go-to-market machine. We were the people who brought the flooring industry together and gave local retailers a voice. Our number one product was a website platform. Could you get a website from Squarespace for $49? Sure. Or you come to Broadlume and pay $1,000 a month. Why would a million-dollar flooring store pay that? Because they believed we were going to build their ecosystem and be their partner. It was not about what we were that month. It was about where we were going to take them. We realized early that being in person was everything. Their parents were in this industry. They want to shake your hand. The real deals get done at the bar at the end of a conference.

Jay: How did you recreate that in-person magic?

Todd: We started the largest Facebook group in flooring, 7,000 retailers in one place. You moderate softly, you're not jamming sales messages in there. That became owned distribution. Then we ran bootcamps twice a year at a $250 million flooring store in Columbus, who was also an investor because we'd bought his CRM. One day was a tour, the next he taught his process, then we taught retailers things that had nothing to do with our business. How to use ChatGPT, how to use Canva. When COVID hit, my dad, an accountant, did an hour of PPP Q&A with retailers across the country. Nothing to do with our software. We repeated that until Broadlume became the trusted advisor to the industry.

Jay: How do you measure ROI on that?

Todd: Extremely difficult. I came from Google, where you spend $5 on clicks and get leads. It took me until the end to understand the power of the brand we built. Take our big conference in Vegas, Surfaces. Most people set up a table and hope. We treated it as a team-building event, plus all our customers there to shake hands, plus we brought a video team to capture testimonials and B-roll that would otherwise cost tens of thousands. Could I hand you an exact annual ROI number? No. But there's a reason our retention was so high on what was basically a commodity website and CRM. When we sold, the buyers put a premium on what our brand meant to the industry.

Jay: As someone who makes content for a fund, the tools to do this are so much cheaper now. People still think it costs thousands.

Todd: Right, but it's about doing it repeatedly. My favorite hack: we had limited-edition Broadlume socks, new ones every year, given out only during show season. We emailed every customer to come to our booth for them. Our booth was packed, 98% our own customers, and people walking by wondered what was going on. Then, talking to a prospect, I could turn to a customer standing there and say, Debbie, can you give her two seconds on why you signed up? I had the booth selling each other. A lot of guys go to collect leads. We always closed at the show.

Jay: Voice AI is getting good. Can I use it as my SDR, re-engage customers with AI emails? Believer, or Twitter hype?

Todd: In vertical SaaS, these AI go-to-market tools are Twitter hype. The way you win is you become vertical with them. If you're a behind-the-screen company blasting out agents and videos that look like you, you'll never scale inside an industry. I only had 15,000 retailers. If I send slop at them, it's the same as a cold call from overseas telling them their Google Ads are bad. When ServiceTitan tried to come into flooring at scale, the mass approach never won, because they didn't know Debbie, or that her uncle passed away last week. I did, because I was at the bar with her.

Jay: Take me back to the AdHawk to Broadlume pivot. Why flooring?

Todd: We launched in 2015 as ad tech, doing about $12 million by 2019. A third of our business was coming from flooring retailers, and we didn't really know it. They came to us because one website company was sending us their business. So we bought that company and went hard into flooring vertical software. It was the luckiest thing we ever did, because come March 2020 our ad tech business would have gone to zero fast, and instead flooring went up the escalator. Home improvement boomed. I also made my biggest mistake in 2019, my first reduction in force, terribly executed. I told everyone if we ever did it again they'd know months ahead. Then COVID hit and within a week I had to cut people again. That ate me alive. It taught me to never say anything definitively again. We went from $12 million down to $3 million and back to $12 million by 2021, raised a growth equity round, bought eight companies, and sold in 2025.

Jay: Why do most investors get roll-ups in legacy categories wrong?

Todd: The first thing people get wrong is being one foot in, one foot out. You either commit to the roll-up or you don't. Our board had both VCs and PE. The VCs wanted organic growth, the PE guys wanted the roll-up. At some point we just committed, and your resources go to integration, not other things. We also had an identity crisis. Were we an all-in-one platform like ServiceTitan, or an ecosystem of products? We admitted we could not rewrite that COBOL ERP and stitch everything into one UI. So we committed to being an ecosystem. Log in once, data passes through, separate systems. The integration is the hardest part. We made a promise to everyone we bought: we will never recut you. There are only about ten software companies in flooring and they all talk. Recut one and you lose your reputation. What I'm proudest of is that every founder stayed until they personally chose to leave.

Jay: My skepticism about VCs doing roll-ups is that most don't have the operational muscle to extract value at a price that makes sense.

Todd: After eleven or twelve years I wouldn't feel comfortable doing it. Some of them are being forced into it, because these ERPs are so entrenched. If you want a full-service product you also have to be the ERP, and that migration is a two or three-year job. We could build a better ERP than the one on COBOL. Great, that's two years, three million dollars, execution risk, and you still have to convince customers to migrate. The hardest part for a flooring store is that their employees are trained on where the buttons are. Your UI can be prettier, but their button is here and yours is there. About 25% will move, 75% will fight you.

Jay: Almost weekly someone pitches me a system of record where the data lives on a legacy IBM system and their strategy is to build a wrapper. So the customer keeps paying the ERP and also pays them.

Todd: Right. Unless your agentic layer is so valuable they'll happily pay IBM and pay you, forget it. At some point we realized we were creating enterprise value, and we had to ask whether a project actually created enterprise value. A new ERP would have been a better product experience, but there was essentially no enterprise value in it for our next three years. A year out from selling, I told my team: we cannot raise more money, we need to grow EBITDA and revenue. I do not care if it runs on COBOL. When you're backed by PE doing $25 million in revenue, the goal is different. I was protecting the downside and limiting my upside on purpose, because getting everyone to a good outcome mattered more than swinging for the fence.

Jay: Everyone's realizing AI is eating white-collar jobs and turning to manufacturing, construction, logistics. Is it still possible to build a vertical business today, with more software options than ever?

Todd: Yes, and it's still go-to-market. Every vertical software company's product is basically the same. Website, CRM, ERP, payments, transactional revenue. There have always been horizontal options like Shopify. The hard part is getting the right message in front of Mrs. Jones, who runs a flooring store and gets a thousand calls a day, and getting her to trust you and sign up. If you think you're going to out-product a horizontal company, you're lying to yourself. Our approach was to sell you a website first. We had 4,500 website customers. If you trusted us there, my ERP pitch was two minutes: it's like all the others, but you already use our website and you like it, and everything integrates, and we'll do whatever it takes to keep you happy. Guys would sign up without even demoing it.

Jay: How do entrepreneurs ask the right questions of such fragmented customers, from mom-and-pops to big multi-location operators?

Todd: I see founders from BCG say, I researched this industry, the TAM is $16 billion. For us it was about being the most authentic. When we bought companies, I was buying reputation as much as revenue. And we lived and breathed flooring. I worked in flooring stores for almost a month. Every single person in the industry has my cell phone number. My favorite thing was the Facebook group. The downside is people complain publicly. I loved it. When someone complained, I responded manually, no assistant, gave them my cell. Salesforce and ServiceTitan are not in the flooring Facebook group answering complaints. Half the time the complaint was a misunderstanding, and the customer would go write their own story about how they talked to the CEO and now the company is amazing. We turned complaints into magical moments.

Todd: So let's turn to the investor side. You've taken VC and PE money and invested yourself. How are VC incentives changing the approach to legacy industries?

Jay: How do you see it?

Todd: Once you understand the incentives on the other side of the table, the relationship gets better. Everyone wants to raise VC, but you have to know you're generally a binary outcome. One or three companies return the fund and the rest do not matter. So when a founder tells me they raised at $40 million, I ask: if someone offered you $75 million right now, would you sell? They say yes. Then you're playing the wrong game. You signed up with a partner playing baseball and you're playing soccer. We sold for a couple hundred million. How many buyers buy at that price? A small percentage. How many buy a $50 million business? Many times more. If we'd raised less money, taken less execution risk, gotten to $6 or $8 million in revenue but kept our optionality, selling for $50 million could have been the same financial outcome for me, with less risk and more buyers.

Jay: And two more rounds of dilution is meaningful for a solo founder. PE optimizes for the protected downside and will take a 3x or 2x.

Todd: Exactly. Understand the game your partner is playing. What you're worth on paper is the least interesting thing to me. Look at how most companies actually get acquired. The majority of exits are tack-ons to private equity. That's your most likely scenario. You can still raise from VC, just don't give up your optionality.

Jay: Five years out, what's a consensus view about software for legacy industries that you think will be totally wrong?

Todd: That these businesses are small potatoes. When you talk to these customers and see how much of their business is tied to labor and how badly they're stretched by labor shortages, there's a massive automation opportunity people are overlooking. There's a giant shortage of skilled labor in the trades. The interesting part is that people are still necessary, but we get second brains and co-pilots. ChatGPT has made me a better handyman, because I take a picture and ask if it's right. It's been impossible to find skilled flooring installers. But now you take someone reasonably handy, put some glasses on them talking to an assistant, and that person can install flooring. So the TAM is labor, but not by replacing it. The pool of skilled labor grows.

Jay: It's about giving owners the tools so they spend everything hiring field ops people, not overhead.

Todd: The thing that changes all of this, and I don't know if it's five years away, is whether the robots show up. If the robots are here, we're talking about something completely different. But otherwise I totally agree. Asterisk, if robots are not here.

Jay: I'll have to have you back in a year to do a whole conversation about robotics and legacy industries.

Todd: Let's do it. And the first robot flooring installer, please call me. This was fun, thank you for having me.

Pull quotes

1. "51% of the reason people signed up for us was our brand, and 49% was our product."

2. "In vertical SaaS, these AI go-to-market tools are Twitter hype. The way you win is you become vertical with them."

3. "Our business was built on relationships and being at the table shaking hands. AI is not going to replace that."

4. "If you buy an ERP and plan to build a new one on the side and migrate people, you had better be ready to maintain two ERPs for a decade."

5. "You signed up with a partner playing baseball and you are playing soccer. It is not bad, you just do not know the game your partner is playing."

Source

From CLIMB Episode 089 with Todd Saunders (Broadlume). Transcript cleaned from the published episode. Watch the full episode: https://youtu.be/cA-a4F5LeQQ