Billion Dollar Learnings: Serial Entrepreneur Dane Atkinson on Serving Coffee Shops and SMB Tech Strategies

As the founder ofOdeko, Dane Atkinson has raised $100 million to scale the company that streamlines small businesses’ inventory deliveries and backend software. He’s also launched six other companies and spent time as the CEO of Squarespace and SumAll.

I sat down with him to discuss how Odeko has found an angle in vertical integration. Below, learn how Atkinson built a network of real-time trust with distrusting customers, why his ADHD is a strength, and how AI will ease small business operations.

You mentioned on Turner Novak's show that the food system has not changed since the Babylonian days. What's wrong with the food system, and how are you helping to solve it at Odeko?

Well, for many restaurants and stores, the model right now is for someone in the city to make a phone call to a local farmer through another distributor, through another distributor. No one's directly communicating. It's all driven by wagons, so to speak, into the city. But companies like Starbucks have leveled the playing field. A truck magically appears at night and loads their store. They're procuring from the source. There's not a variety of middlemen or obstructions.

Odeko wants to bring that same infrastructure to the rest of the physical world. If you're building your independent coffee shop, you should have the same backbone as a Starbucks. Just tell your own story without worrying whether the milk will be in the fridge or your croissants will be delivered.

Why did you want to focus on this after SumAll?

A lot of founders get obsessed with an idea. I try to get obsessed with a customer, especially small businesses. They're why there's color in the world. They create new, diverse ways of thinking about things. It’s why I’m always trying to find a way to enable their world.

Due in part to my ADHD, I've also owned bars, restaurants, and coffee shops. They're incredibly locked in an old model with very little technology. A small business owner is a distinct archetype of a person. It's someone who is nurturing and guarding a dream. It encapsulates their vision of how they want to live in a free world and how they want to better their community. It's different from a consumer's behavior and very different from an enterprise's behavior. Having served that community for a lifetime, you start to understand how amazing they are and the limitations in the choices they face.

Speaking of choices, one of the most challenging decisions a small business owner can make is how they spend money. Spending on startup products like Odeko—especially if they’ve never heard of them—is never an easy win. Is it easier to sell revenue to them, or is it easier to sell savings?

It's easier to sell reality versus potential. You can sell revenue if it's true revenue, and you can sell savings if it's true savings. Selling hope on either front is very, very hard for a small business to swallow. Small business owners also work on intuition. If someone walks in the door and says, I can make you tons of money, their hair stands in the back of their heads.

That’s why our angle is to provide irrefutable proof and demonstrate to them how to reallocate existing spend to try it out. If you do this, they'll do it all day long.

Iteration is also no longer efficient. Making a minor improvement on something will not get the switch-over-cost behavior. It needs to be revolutionary. There needs to be no doubt that it's worth their time. They usually end each year with a net of three to fifteen points. You need to be adding three to five points to that equation to make them listen to you. If you walk in saying you're going to save them 10 points, they won’t believe you and no one's going to care.

Really?

Everyone lies to them all the time. When we started, we thought the coffee shops would lock, stock, and barrel move over to Odeko because it was such a clear sell. You give us your keys to your door. We’ll drive around at night and load your store. We’ll take a picture of the inventory sitting on your shelf, so that when you wake up in the morning, you know it's there. But it turned out they still didn’t trust us. They would say, “Sure, I bet it’s going to be great, but you’re going to add a gas tax, there's going to be a software fee, or I bet you didn’t tell me the whole story.”

That’s why, when we walk in the door to show them Odeko, we say it’ll be the same or cheaper than the outdated systems they currently have. We’ll talk about how it comes in one vehicle, how they can use one piece of tech to do multiple things, and that there’s a guaranteed price that costs less than the price of the product they’re already using. That gets them behind it because it’s primary data.

From there, you just have to earn more trust. It can take a month or two to prove we won’t upsell or upcharge them. Some coffee shops will have us bring a bottle of milk to them just to prove that we’ll do it.

So how did Odeko get its first customers?

We had to put ourselves out there for our first customers, which were Birch, La Colombe, and Joe. We started with them simply because they're much easier to find. They made a huge bet on us, which is why you need to make sure the flywheel of your product works. That turned into a very good strategy for this industry. In Odeko’s world, when Pura Vida buys you, your trust factor goes up. Our vendors will even do ride-alongs and vouch for us in person, telling a small business owner, “Hey, these guys are great.”

Once others come in, you cannot brute force a business model. The founder cannot sell to scale efficiently to make a venture-funded business. Now that we're in a lot of cities, we try to build councils of customers that can tell us when we're screwing up. That way, we have some feedback loops. But quite quickly after you get that first group, you need to see if the product will sell without any kind of founder magic attached to it.

Were there any hacks along the way or fun things you guys tried that made you think, “Damn, that really worked?”

Always. We were early in the new wave of podcasts and content shows. We would sponsor Leo Laporte and Kevin Rose and all these early podcasters. We gave away an iPhone a day on Twitter and that blew that up. We’ll go to national competitions for latte art. ESPN covers it at the end of the year. It's spectacular. We’ll sponsor baristas and give away a prize. We even sat on the roof down in Spring Street and Kevin and I threw iPads off the roof. You just try everything to see what sticks. It’s what makes startups magical. Your mental space is constantly breaking out of the framework of what exists.

It's this rare breed of analytical and creative that I find makes for the best entrepreneurs. Those two skills drive real results when it comes to problem-solving. To pivot, let’s talk about unit economics. I love vertical integration, especially with SMBs given how hard it is to acquire them as customers. How do you compare your unit economics on a per-customer basis from when you started to today? Have you unlocked more revenue that allows you to spend more to acquire new SMBs?

Because we picked up a heavy lift in our customer base, the unit economics are probably similar from when we started to today. We make a lot more per door, but we've had to invest more too. A coffee shop will spend twenty grand with us a month, which you would never get for a technology piece. With our margins, we're making thousands of dollars a month, which again would be impossible through technology alone.

Having a vertically integrated solution means the whole system works a lot better. Not only did we get into a buying power position, but we also reduced nine trucks into one. We removed the payment rails from splintered systems into one conduit. That means we reduced the cost burden and didn’t have to steal margin from either the farm or the coffee shop. We could just have a healthy margin and both of them see a benefit.

If a young founder building an SMB tech company came up to you, would you tell them to vertically integrate or sell software?

A lot of young founders are still looking at building UX, which seems crazy because UX is kind of going with this new iteration of technology. I would encourage them to own the problem and not just try to pitch the most business-effective model as a pure SaaS. In many cases, it would be fully vertical and owning the solution, but in some cases, it's making sure you're solving the whole thing, you're doing the payroll for the staff, you’re scheduling the team, and so on.

The long-term gain may be us suddenly becoming Starbucks. Customers will say, “I don't want to go to the old chain, I want to go to my local space and embrace that owner.” We'll have 30,000 doors, and an opportunity to maintain being vertical and owning the solution for all of them. That’s why I think people haven't quite caught on to how vertical SaaS can be in the long term, where they end up owning the entire industry.

Should VCs be playing in this space?

It's early. I bet you in five years they will, because Odeko will keep scaling and others will want to create better partnerships and funnels from the rails we’re building. They’ll just need to be better intertwined for new companies to raise the kind of money necessary to start.

You’ve brought up the topic of ADHD. What role has ADHD played in your life?

Central. It's my neurology. It persists in every aspect from start to here. It's made it very hard for me to succeed in certain parts of my life and very easy in other parts. I've come to appreciate the power it's given me. I can context switch incredibly fast. I tend to see things that others don't. I’m well suited for the speed of being a CEO.

Those two things I feel are additive that I can embrace them and work them, but I suck at so much else. I am bad at so much stuff. Being a CEO led me to imposter syndrome. I thought I had to be great at everything. I had to get to a place where I'm okay with having slightly different neurology. Coming to terms with it is helpful.

How will AI impact the next generation of SMB SaaS companies?

Well, I drank the AI Kool-Aid, because I think it's very aligned with many small business use cases. With the stress of building a small business, you have no time. You are all alone. You have a huge amount of communication overhead that you don't have tooling for. AI is great at navigating communication. So as an example, for a long time, we looked at acquiring staffing tools. It's one of the problems our customers face because scheduling is a nightmare. One of our small teams built a natural language processor in just two weeks for this. It will call someone and say, “Hey Jason, can you work Saturday? No? No problem, I'll call the next person.”

What will that look like in 10 years?

It's going to be so awesome. The owner is going to walk their way to their ice cream store and ask their AI what the day's going to look like. It’ll help them understand that there’s going to be a rush at two o'clock, but they've scheduled Jim who has some float hours to come in to help. The AI also increased their inventory by 25 percent for that, and it was delivered last night. “And oh, by the way,” it will say, “the trash can still looks full. We told Jill about it three times. We're going to have to give her a notice if she doesn't clean the trash can. But when you go in, if you want to point it out, that'd be great. We also messaged our customers we're doing an ice cream test on six of the new flavors. 15 people signed up to do it.”

In other words, AI is an operations partner that manages all the different data streams and infrastructure. It's going to take a while, and there will be resistance, but it’s a bright future.