Glen Coates Shares Why B2B Commerce Failed at Handshake and How Shopify Became the Perfect Home

Glen Coates spent nine years building Handshake, a wholesale e-commerce platform that grew to millions in ARR before hitting fundamental market challenges.

Now VP & Head of Core Product at Shopify, Glen shares the hard lessons about why B2B commerce struggled as a standalone business, the painful decision to sell to Shopify, and what he’s learned about building at massive scale during his nearly six years there.

He also reveals what makes Tobi Lütke such an effective leader and why so many acquired founders choose to stay at Shopify long-term.

The full discussion with Glen Coates follows below…

Mark MacLeod:

In this episode, I sit down with Glen Coates, who was formerly the founder and CEO of Handshake and now is VP of Product for Core at Shopify in their enterprise division. Glen and I go way back. I actually helped sell his company to Shopify and it was really great to catch up with him. We walked through the Handshake journey, the whole process of selling to Shopify. He’s almost on his sixth anniversary as a Shopify employee. And so we talked about what makes Shopify such a great company, keeping so many great founders around.

We explore lots of different aspects of Shopify. This was really a fascinating discussion. If you run an app in the Shopify ecosystem, you’re also going to want to listen to Glen’s advice. I hope you check out the episode and I hope you enjoy it and subscribe to get future episodes. Thank you.

Glen, welcome to The Startup CEO Show. It’s a true pleasure to have you here. How are you doing today?

Glen Coates:

Good, man. It’s good to see you again, Mark. For those listening, Mark and I are prior friends and acquaintances. Mark was instrumental in my company Handshake and handing it to Shopify. So Mark and I actually know each other pretty well from the trenches back in the day.

Mark is much cooler looking now than he used to be. Mark was kind of a blazer and shiny shoes man back in the day. And now he’s very enlightened, as you can see if you’re reading this.

Mark MacLeod:

Thank you. The last time I saw you in person, I think we had a coffee somewhere down in Manhattan. Very shortly after you joined.

Glen Coates:

Yeah, right after. But honestly, those times are all kind of a blur. There was a lot of liquid medicine around that time, so I can’t remember all the various recoveries.

Mark MacLeod:

Yeah, that’s fair. Well, let’s get into it. First of all, everybody knows who Shopify is. And we’re going to spend a bunch of time talking about Shopify, but maybe let’s take a step back and tell folks what Handshake was, what you were trying to achieve. Just walk us through that journey. And then separately, we’ll talk maybe about the decision and process to sell it. But let’s just start with what Handshake was.

Glen Coates:

Yeah, well, Handshake started in about 2010. So I had a sort of computer science background. Worked in video games for a while. Had this weird thing where I… So I got an offer from an ex-high school teacher to move to San Diego when I was 26. And he was like, hey, I’ve got this company with my wife, and we sell eco-friendly shopping bags. Do you want to move to San Diego and run our office?

So I ended up moving to San Diego, not really knowing anything. I knew how to code, didn’t know how warehouses worked, didn’t know how retail worked, didn’t know how the supply chain worked. And the business that I was running happened to have a very big wholesale component, which led to me being at trade shows a lot. At trade shows, you happen to have to write a lot of orders out by hand back in 2008.

And so I’m doing a lot of writing orders by hand. And I think the iPhone had come out. I think, wouldn’t it be great if there’s an iPhone app for this? I ended up writing an iPhone app that was originally an order entry tool for a salesperson, right? Like punching my orders when I’m talking to my customers. And that grew from being an iPhone app to an iPad app.

And it was very salesperson-driven for the kind of wholesale industry. And then, eventually, of course, like everything, it went online, and customers wanted to place their own orders without going through a sales rep. So we added e-commerce, but it was wholesale e-commerce.

So, as you can tell, the thing that I’m describing is starting to sound a lot like Shopify, except only for wholesale.

And so our whole mission was just how do we take these pen and paper, old school ordering wholesale businesses into the kind of post-Internet, post-mobile world.

Mark MacLeod:

I love it. And sort of coming, is this fair to say, from the opposite end: the bulk of retail is still in a physical store, even though online keeps growing. And so Shopify started with pure online. Now it also does in-store. The wholesale, by definition, the end retailer was probably selling in a store. Is that fair to say?

Glen Coates:

Exactly. And so Handshake was really like, okay, you’re a wholesaler selling to all these retail stores. You’ve probably got salespeople who are either at trade shows or out on the road meeting with their accounts, and they do a lot of their work on pen and paper or in spreadsheets or something. That feels very old school.

How do we give you a great iPhone app to do that work a little better? And of course, because I’m an insane masochist, I picked the most old school, brutal, change management people to build an iPhone app for because I thought life’s too easy, and how do I make it really, really hard?

Mark MacLeod:

Well, you succeeded. I can only imagine, you’re dealing with overseas and antiquated ERP systems. People whose systems are in Cobalt are basic, and you’re probably dealing with all of that.

Glen Coates:

Yeah. I would say the beginning of Handshake, the thing that made it possible, or the thing that made it feel like a thing worth doing, was the tailwind of mobile arriving. Because, prior to that, it wasn’t like anyone thought that pen and paper was a good idea. It was just like, well, how do we do this mobile? And the answer was Palm Pilots and a whole bunch of crap that was not very fun to use. And the iPhone’s arrival was really the sort of disjoint in the curve of, okay, this is clearly the way this should be done. Who’s going to do it?

And so there was a tailwind of everyone buying their first iPhones, and then the iPad came out and, I mean, every salesperson in the world was like, obviously I should be using this iPad thing. So there was a bit of that mobile tailwind, which made the ordinary sort of brutal change management of how do you get all these old school people to do work in a different way easier than it would have been otherwise, but it was still hard. There just was that there was that moment in time.

Mark MacLeod:

Interestingly, your timing is not dissimilar to Shopify itself. Like, Shopify was in market and I was there in 2010 part-time, but I was there. Still a very small company then. Were you aware of them, or did they not come into your universe yet?

Glen Coates:

Just barely, but I mean, you’ll remember this, Mark, back in 2010, I don’t think it was a bit more of a competitive race. There was Shopify, there was BigCommerce, there was Magento, there was even, what was it called, Volusion? There were still players kicking around, and it wasn’t really clear who was coming out on top at that point.

Mark MacLeod:

Let’s not forget Yahoo Stores.

Glen Coates:

Oh, yeah. I mean, who could forget?

Mark MacLeod:

So, walk us through the journey. The highs and lows, what stood out for you when you could do a retro on Handshake.

Glen Coates:

Yeah, I mean, it was nine years in total. You know, to be very honest about it, I felt like the first three years were the kind of fun builder times where I was still spending most of my time coding. Like, I wrote the original version of the app myself, and then we eventually brought on the first few engineers.

But the first three years, I would say were super fun, coding all day all night, which I really enjoyed. You know, I made tons of mistakes, but I really enjoyed that. I was very builder-y.

And then we were successful enough that we raised a seed and then a Series A. And then the Series A was around 15 people. And that brought us into era two. And I would say era two was the let’s scale sales and marketing and let’s have Glen become a real CEO time, AKA let’s stop having so much fun.

But that was when it was like, okay, pull back from spending so much time coding and being in the product. Let’s start building out sales, let’s start building out marketing, let’s start having an org chart, let’s start building a real company, all that kind of stuff. And I would say during that period that was less fun, but we were still growing quite aggressively, and it was still up on the right, and it felt really good.

Although at the time, I was doing things that I found very challenging. I don’t know if I found them as fun as I did the building part, but I was in this sort of like, this is the right thing for the company, and I’m going to be a real CEO. I’m going to be a real boy. Look, I’m a real boy now.

And then I would say that the last third was probably the hardest part, and to be transparent about it, like many startups, Handshake started off mainly serving smaller companies like Shopify, although not quite as small as Shopify. Because Shopify is like, I have no business at all. I want to sell a candle on the Internet kind of small.

A small wholesale business is they might be doing a million in revenue or something small like that. Sort of similar to the beginning of where Shopify Plus starts today. And so Handshake grew up on a lot of these smaller companies and then so many SaaS companies, right? It’s like, oh, we’re going upmarket because the ACVs are higher and the LTV is higher.

Mark MacLeod:

Was that your decision or was that the VCs pushing you?

Glen Coates:

I guess it was both. They were definitely like, hey, it looks like you should go upmarket. The economics look better. And I definitely, I mean, look, this is the thing that Shopify got so right and Handshake didn’t quite get right enough, which is Shopify committed so deeply from the very beginning to being a product that is so easy to use that it doesn’t use human support people as basically a crutch for the inability of people to figure out the product on their own.

Now, again, easier problem because it’s like, oh, I want to sell three candles in my store. Well, it’s not that hard to set up three candles through a website, but when an old-school wholesaler comes in and says, well, here’s my 10,000 products and I exported them from QuickBooks in this crazy format, you can just take that format, right? And you’re like, well, not really.

And now we get into this, how do we help you do it? And so it was always hard. I would say the thing that was true about Handshake was that it was just as hard to onboard the $100-a-month customer as it was to onboard the $1,000-a-month customer.

Mark MacLeod:

That’s definitely an argument for moving up-market then.

Glen Coates:

Exactly. You can see how we got there. The mistake I made was that we sort of extrapolated that curve one or two steps too far. Where I think the right place for Handshake was in the mid-market somewhere. The companies that would pay us $1,000 a year were probably too expensive to serve.

The company’s paying us $10,000 to $50,000 a year was probably the sweet spot. But then we tried to go above that, right? We were like, let’s just draw this curve out a little bit further, and let’s go way up. And we were just not ready for that.

And at the time I didn’t really understand exactly why. I now understand that Handshake was just not really architected truly as a platform, right? It did not really have the APIs, the SDKs, the extension points, all the things that you need in order to let…

Mark MacLeod:

Those enterprise customers probably are very custom environments, right?

Glen Coates:

Yeah, they want to customize everything, and you need to have the building blocks there so that they can do that. I mean, if you go into the Handshake code base, I think it still exists on GitHub somewhere, there are certain places you can go in there, and you’ll find a bit in our code that says if customer name equals blah, do this special thing. Which is, of course, like the cardinal sin.

And so we were just not ready to be an enterprise-ready platform and we pushed very hard in that direction. And I made the classic first-time dumb startup founder mistake of going out over my skis on sales and marketing to the enterprise. And the problem with that of course, as you know, is that when an enterprise go-to-market fails, it fails very slowly because sales cycles are very long.

Mark MacLeod:

It’s funny, they both grow and fail slowly.

Glen Coates:

Exactly. So, when you’re doing SMB, if you’re failing, you find out within a month. When you’re doing enterprise, when you’re failing, you find out in 18 months. And if your runway is 18 months, that’s tough, right?

Mark MacLeod:

And then you’ve built so much because you’re serving three masters there, right? What the SMB needs is different from what the mid-market needs and then different from the enterprise. So like you’re an inch thick and a mile wide.

Glen Coates:

Exactly. So that push didn’t work, and we burnt a bunch of runway on sort of a big enterprise go-to-market that ultimately just did not go anywhere. And then we ended up having to scale back the company. I think we cut headcount by some pretty significant amount, something like 50% and we refocused on that kind of mid-market segment where we really did have good product market fit but I mean, as you know, startups are a thing where your margin for error is not usually one entire fundraise.

Mark MacLeod:

No, no, it’s not.

Glen Coates:

And so, this is kind of when you and I met and we got to the point where we’re like, okay, cool, here we are. We know what the business is now. We know where we do have product market fit. And look, by that point, we had been at it for a long time.

And I think I learned something late in the game, which, I mean, I’ll tell you this story. I’m not sure if I’ve told this story publicly before, because the thing about B2B e-commerce versus D2C e-commerce, which was basically the investment thesis for all investors who invested in B2B e-commerce companies, was that the market size of B2B is so much bigger than D2C. It depends on which Gartner or whatever report you read. But people have always sized B2B at 2 to 3x as big as D2C as a market.

So you’re like, great. Huge TAM, amazing. So much opportunity. And this prior art of Shopify, Magento, Demandware, and Hybris going into these markets and being successful. So, okay, great. Looks like an amazing market.

And you remember Russ Jones?

Mark MacLeod:

Of course.

Glen Coates:

So Russ was an advisor to Handshake. I think after he left Shopify as the CFO, he had taken Shopify public. And we got connected with him at some point, and then he became one of my advisors in that sort of last third of the company’s lifetime or so.

Mark MacLeod:

What a coincidence.

Glen Coates:

I remember doing a prep call with Russ before a board meeting, and we were going to go out and do a fundraiser. And this is in, I want to say, 2017 or 18 or something around there. And I’m doing a prep call with Russ and I’m like, hey, this is basically the fundraising story. You know, here’s the market, here’s the product, here’s the competitive set. You know, here are the things that we figured out and kind of got wrong, but we’ve now fixed. This is the story. It’s a huge market and we’re well-positioned for it.

And I’ll never forget what Russ said, he heard me do the pitch, and I’m in a startup founder tap dancing look at the stars kind of mode, right?

Mark MacLeod:

Jedi mind tricks.

Glen Coates:

Exactly. I’m saying all the words. And Russ says to me, he’s like, Glen, everything you just said to me, I totally believe you. I believe that the market is huge. I believe that you’re the best company to solve this. You’ve got the best product. You know, I believe everything you just said. I believe the things that you figured out that were not working, and fixed them. And he was like, if all of that’s true, why is this company so fucking hard to grow?

Mark MacLeod:

Sometimes the truth’s hard to hear.

Glen Coates:

Exactly.

Mark MacLeod:

And I can totally picture him delivering that as well. Like, in a compassionate but direct way, you know?

Glen Coates:

I know, he was very kind about it. Yeah. I think maybe the time after that, he bought me a $2.50 breakfast at the Capital Diner. Maybe he treated me once for that. But at the time, I actually did not have a good answer for why it was so hard to grow.

I later got to a classic douchebag product two by two answer. I think the real answer was that D2C and B2B commerce are similar in some ways, right? And they even share some customers in some ways. But D2C e-commerce typically has customers with high gross margins because they’re eliminating, and that is literally the premise of direct-to-consumer. I’m taking the retail margin and the wholesale margin at the same time, right?

So Shopify had customers with extremely high gross margins and extremely simple needs, right? I just want to sell five candles on my website. I just want to sell Allbirds. And I only have the first version of Allbirds, there were only a few of them, right? Simple needs, high margins. B2B e-commerce is the exact opposite of that. It has extremely complex needs and razor-thin margins.

Like, people who sell like, I don’t know, candy and cigarettes to all of the gas stations in America, these companies operate on almost no margin at all, right? And so, yes, the market is 2 to 3x as big as D2C, but if you adjust it for gross margin, it might not be. And then if you adjust it for complexity and cost-to-serve, okay, it’s even smaller than that.

And anyway, fast forward through all of this, sad, sobby, crappy story, in a way, when you and I met, part of the thesis for why is it better to be part of a company like Shopify is because once you are solving both of those problems together… Like, if you as Shopify can come to the merchant and you can say, okay, you want to sell online and you want to sell wholesale and you want to sell retail, you want to do all these things.

The bad news is that’s sort of a big problem to solve but the good news is if you do it all in one place, you can amortize all those costs and simplify it down and just put it into one system that does it all. That’s such a better deal for the customer.

Because otherwise, the other version of that is I buy a D2C e-commerce thing and I buy a B2B e-commerce thing and then I buy a point of sale thing and then I have to integrate them all together and of course they all fight with each other and I’m going to spend the next five years in IT hell, and it’s just going to be awful.

Or you can just be like, I’m just going to do it all in one place. It’s actually going to be not-so-hard to do. And it might be challenging in some ways, but I get a lot back in exchange for doing that. And so, to go all the way back to the top, like B2B e-commerce on its own is actually not a great business. But B2B e-commerce as part of a unified stack is a great business.

And that’s kind of what we’ve seen in the last five, six years that I’ve been to Shopify, among other things. But it’s just one of those things you learn the hard way as a founder. It’s such a weird thing where it’s like there are some businesses that make no sense on their own, but make incredible sense as part of another thing.

Mark MacLeod:

That’s very true. What was your state of mind when the board was like, you know what, we should find a home for this?

Glen Coates:

What was my state? I mean, it was a stressful time, no doubt. Like, if you go through a company layoff, that’s like not an easy thing. I think if I’m being honest, one of the things that founders sort of have to have in them a little bit is not asking if this is going to work out, don’t question it ever.

Like, you just have this mindless, “I’m going to smash my head through this wall,” but the second you ask yourself, hey, is your head actually going to go through this wall? Like, that’s the second that your head doesn’t go through the wall, right? And so I had that very mindless, like, don’t ask why, just keep going thing.

I think I read an article. I’ve tried to find this, and I can never find it, but I read a newsletter back in 2011 or something, and Mike and I were back in projective space at the time, and some VC funded that had done some analysis of, like, okay, we interviewed our 50 most successful CEOs, who all IPOed, and we tried to figure out what makes a great CEO, and the conclusion at the end of the newsletter was that the only thing we could find in common between all these people is that none of them would quit.

Mark MacLeod:

Well, that could be a feature or a bug, but is probably most often the time a feature.

Glen Coates:

Exactly. And so I basically approached the job that way for a very, very, very long time. And I think it was it was like conversations like that with Russ Jones where I had to just ask. And it was not just for me, I had to look at my team. I looked at Mike, I looked at Brandon, I looked at Aaron, I looked at Mandy.

I looked at all these people who’d been with me for the better part of a decade in some cases, and I had to really think about them and say, even if I want to keep doing this, even if I want to keep smashing my head through walls for this thing, can I really, in good faith, ask Mike to give me another five years, Brandon to give me another five years if I can’t really give a super good answer to Russ’ question or the variations on that question, can I ask these people to keep doing this?

And so I think eventually I was just like, let’s look at what the options would be because I didn’t have a super good answer at the time. I didn’t have the pithy, like, oh, low-margin, high-complexity, yada, yada. I didn’t have that at the time. I would eventually get that. But I knew at the time I didn’t have a good answer.

And so it’s really hard. I mean, I’m an atheist, but I almost think of it if you have someone who’s very religious, and then you say something that makes them realize that God doesn’t exist.

Mark MacLeod:

Their whole world falls apart.

Glen Coates:

Like, it’s almost a feeling of like, oh, my God, God doesn’t exist. What does that mean for what I’ve been doing for the last eight years? Was I just praying at a false altar? Like, what have I been doing?

And I mean, Handshake was an amazing journey. Like so many amazing people who worked with our company. Our customers loved us. You know, there are so many businesses that provide incredible value to their customers and build great cultures that just happen to be not super good businesses.

Mark MacLeod:

Yeah, totally agree.

Glen Coates:

I mean, I listened to the Scott Galloway podcast, and he did a bit about this recently. It was about airlines. He was like, airlines are such an incredible… Like, if you think about what airlines have contributed to humanity over the last 50 years…

Mark MacLeod:

Horrible businesses, huge fixed costs, razor-thin margins.

Glen Coates:

Yeah. Think about what life would be like on planet Earth if there were no airlines. It’d be not as good, but they’re just not that good of a business.

Mark MacLeod:

Let’s switch gears and talk about Shopify. So how long have you been there now?

Glen Coates:

In May, it will be six years.

Mark MacLeod:

So you’ve been there almost six years. The other company I was involved with going to Shopify was Tictail with Carl. I think he’s still there.

Glen Coates:

Still there.

Mark MacLeod:

Dan was there for a while, Rohan, you just got Ray ready.

Glen Coates:

We’re collecting them.

Mark MacLeod:

Yeah. What is it about Shopify that keeps great founders? Because I can’t tell you how many companies I’ve helped sell where the founders don’t even make it through the earnout. Like, they’re just putting a bullet in their head. You could start something, you could go anywhere. Like, there’s something really compelling here. So I’d love to know what it is.

Glen Coates:

I can give my answer to this. And I think the answer would be different for different people. But Shopify, I mean, it sort of comes from Tobi. Like, Tobi is a founder’s founder, I guess I don’t really know how to say it other than, like, he’s still as hardcore as he’s ever been. He’s like, never happy. He’s always, “What’s the next thing?”

I mean, obviously, right now, AI is the thing, but the thing that happens internally is something new comes out and Tobi’s like, why haven’t we implemented this yet? When passkeys came out or when crypto’s exploding or all the LLM stuff, or even just some new CSS thing that’s like, oh, there’s a new CSS effect that all the browsers support. Like, why haven’t we adopted it yet?

Mark MacLeod:

Is he asking this the morning after it came out?

Glen Coates:

Literally, there’ll be a link in a Slack channel that’s like, look at this. And then as soon as you see that link show up the clock is ticking to like, where the fuck is this thing? Why haven’t we implemented it yet? And it’s just a matter of when. It’s not a matter of if.

And so there is this sort of relentless drive to keep improving, stay on the cutting edge, innovate. It’s such a lame, meme word but Tobi really is living in the future and is perpetually dissatisfied with the present, which has its pros and cons. And I think he’d be the first to admit this, one of the downsides of living in the future and being angry at the present is that you are sometimes too early to things. And Shopify has examples of that where we’ve been too early on a thing. But I guess I’d rather the version where we’re sometimes too early than often too late on stuff.

So, I think that helps for founders. Like, founders feel like they are at courtside waiting for what’s next. I think in a way that you might not have if you ended up at a more traditional acquirer. And so there’s that, there is this sort of drive to do new things that is relentless.

I would say the other thing that I was maybe not expecting about Shopify that I have learned that I quite enjoy is when you’re a small founder, and like, Handshake had millions of dollars of ARR, but not tens of millions of dollars of ARR. When you are playing on the big stage it is sort of exhilarating and sort of interesting in a way.

You’re thrilling your small customer base is very gratifying in a sort of very personal way. Like you’re talking to your customers and they’re like, man, I love that thing you guys shipped. Or like, it has that sort of personal quality to it. But when you do something at Shopify and you change something in the checkout and the internet changes the next day or you create you make lending products available that literally would be impossible otherwise to like millions of businesses.

There’s something about impact at scale, I guess is the thing that I’ve found quite rewarding that I don’t think you get a much smaller business unless you strike gold and you turn out to be Instagram. And then you suddenly become a 10-person team that is controlling a very large percentage of the Internet. But outside of rare cases like that, you don’t.

Mark MacLeod:

That’s so rare. A 10-person company is being bought for a billion dollars.

Glen Coates:

Yeah, exactly. So it doesn’t happen too often. Although I feel like that’s the opening bid for most AI. I mean, I feel like I could start an AI company and just…

Mark MacLeod:

I don’t know why you haven’t started an AI company.

Glen Coates:

I know I should start Glen AI and then just like take bids starting at a billion.

Mark MacLeod:

I’ll do that deal for you, no problem.

I had Tobi on the podcast last year, and my observations were that he’s still the same Tobi, is super intense and works as hard as ever. He responded to an email when I reached out to him at like 4 in the morning. Maybe it was in Iceland or something, I don’t know. I was just like, the dude’s busy and I really admire that.

Glen Coates:

Yeah, I mean the thing that you just brought this memory to mind. But like last year we did a sort of internal employee summit around June and it was at Enocare in Toronto and we had everyone there in this huge stadium hall thing. And so we do this sort of the day of all the talks and everyone’s talking about what’s coming. And then we had hack days, which is this kind of Shopify tradition of two or three days of people just hacking on cool side projects.

And I think on the last night of the week, it was a three or four-day event, maybe five, I can’t remember. But it was the last night and we had people DJing, like, Harley DJd, Mike DJd. Everyone was partying up in the front of the room, really fun. And you’re like, where’s Tobi?

And then you look and he’s still at the desk. We had all the desks in the room as well. And he’s just at the desk with a couple of engineers, heads down, headphones on. He was probably some LLM thing at the time, but he was fully in flow mode. And I was just like, yep, that’s Tobi.

Mark MacLeod:

Yeah, he told me he’s not about to step down as CEO, but he’s like, sometimes I kind of just wish I was the CTO. Like, I just love new technology so much.

Glen Coates:

What do you mean sometimes? 99% of the time.

Mark MacLeod:

My understanding is that you’re responsible for the app ecosystem. Is that right?

Glen Coates:

Among other things, yeah.

Mark MacLeod:

Yeah, I knew that was an add-on. So I’ve had many clients, either as coaching clients or banking clients who are in the app ecosystem. I helped sell a company called Privy in the email marketing space.

Glen Coates:

Yeah, I know Privy.

Mark MacLeod:

And hey, I’m just wondering what advice you might have. There are probably lots of folks reading who are in the ecosystem. Like, any advice you have on how folks can stand out and thrive and build a really big business on your platform?

Glen Coates:

The most sort of timely answer, I think Shopify’s developer platform has gone through a pretty major sort of upgrade over the past two to three years. So, Shopify hasn’t had an app store and a developer platform, quote unquote, for a long time. I mean, I actually don’t even know when it first started. Was it 2010? It was before 2015 for sure.

Mark MacLeod:

I don’t think it was 2010 but yeah, afterwards.

Glen Coates:

It’s been around for a long time. But if you sort of peeled back the curtain on the Shopify developer platform five years ago, for sure, 10 years ago, I think what you would find is a pretty basic set of APIs and sure, the ability to embed your app’s UI in that sort of iframe within the Shopify admin. But like, I’m being a little bit sort of reductive here, but it was kind of like just a bunch of APIs to read and write data from Shopify, and then just like an iframe you could stick your website into, right?

When I got Shopify a few years ago if you looked at a Shopify app, if you actually installed it and started using it, you would find that it didn’t feel like the app and Shopify really fit together in the way that most iPhone apps really feel like they fit in with your phone, right?

Like, when you pull up Instagram, Uber, Signal, or whatever, these things really feel like they’re part of the device. And I don’t think Shopify apps felt like they were part of Shopify. If you go back five or six years, they felt like a website that was yoinked in.

Now, fast forward we’ve actually done all this work in the last few years to create extension points and functions and flow extensions and Polaris has gotten better and the net of all of that is that you can now build an app that really does truly feel like it’s part of Shopify, right?

I mean, I always think of this as, like, mobile web existed before the iPhone, right? I don’t know if it was huge, but it existed. So you could be building some SaaS apps on the website and you could have made your mobile web version that would fit on the screen of a Blackberry or whatever.

But when the iPhone came out and there was the iOS SDK and you could build natively for the iPhone, you really saw the companies split into two camps, right? There was the sort of like, well, we’re just going to try and keep doing it the old way and we’re going to sort of mobile web it a little bit, and yeah…

Mark MacLeod:

Kind of like Facebook when they went public.

Glen Coates:

Exactly. So there was sort of the, like, we’re going to try and cram the old web thing into the iPhone and hope it works camp because they had lots of huge stacks, and man, it’s going to be a lot of work to redo all of this.

And then you had the new school people, one of which was Handshake. Like, this was literally my strategy was like, oh, I’m going to do it iPhone-first. You know, I’m going to build fully native on this thing and I’m going to be bespoke for this.

And I think Shopify is sort of going through that now. The dev platform now lets you build natively for Shopify in a way that before it was just kind of a website joined into Shopify. And I think there’s just a huge opportunity there for app developers to sort of quote-unquote build natively.

And I think because five to ten years ago, Shopify wasn’t quite the dominant app ecosystem that it is today. And so for us to go to an app developer in 2015 and say, hey, we want you to build just for Shopify, we want you to write a bunch of code that’s going to work just on Shopify, not anywhere else. Maybe a bit of a tall order. Today it probably is a bit more worth it. And so I think the people who make that choice can sort of reap those rewards.

Mark MacLeod:

It’d be self-serving if you were to say this so I’ll say it for you. I think the old advice from boards might be to diversify your platform risk and don’t put all your eggs in one basket. But I think that’s bad advice now. I think Shopify’s just unarguably won the platform battle. It’s the best place for potential distribution. It’s the best place for your end customer. The combined solution is just going to be better. Like, that notion of diversifying to spread platform risk, I think is just an old concept.

Glen Coates:

Yeah. And I mean, not to keep harping on Instagram, but imagine if Instagram had said, oh, we’re going to hedge our bets. It’s going to be sort of a mobile web thing so that we can get on all the phones. It wouldn’t have been the same thing.

So I mean, that’d probably be my most timely piece of advice. And I think what you just said makes sense as well. It’s like doing the A/B test in your head of like, what happens if I go all in on this thing versus what happens if I hedge my bets across multiple platforms? Which of those seems like the more exciting business?

Mark MacLeod:

So you talked about how different it is to operate at a real scale but you remember what it was like to be a founder kind of dreaming of that scale. From that vantage point, what advice do you offer to CEOs? Like, you now understand what it means to kind of operate at scale. If you were now Russ mentoring the next Glen, what nuggets of wisdom would you share?

Glen Coates:

Stay out of B2B commerce would be the first one. The thing that I’ve learned, just in terms of the way that I did the job, if I could do Handshake again, one of the things that I would do is I would stay much closer to the product than I did.

And one of the mistakes that I made was I did need to step back from doing quite so much coding. That was true. Like, I needed to spend more time on other parts of the business. And I mean, I don’t know if boards still give this advice, but at the time, this was sort of the playbook, right? It’s like, oh, you need to be a real CEO now and start building the business and let the R&D team do R&D things and don’t spend so much time on that.

The thing that I’ve learned at Shopify that has been surprising is that… So, I don’t build anything myself anymore. The engineering team has thousands of engineers. I don’t do any of it myself. I don’t design things myself. The thing that I do now that I did not do at Handshake, which is super weird, is that I still review everything insanely closely now.

So, I don’t do anything, but I review everything, down to the pixels, right? I think the mistake that I made at Handshake was as I pulled away from the doing, I also pulled away from the reviewing, and I shouldn’t have done the second piece. I should have been reviewing like a hawk the entire time, even as I let people do more and more of the build. So that’s something that I would say I didn’t get right.

And then I think you touched on this earlier, which was I think Handshake did become a little bit spread too thin. Like, it was sort of trying to serve too many different types of customers. And so it sort of became this thing that was in pursuit of the enterprise, we didn’t put so much emphasis on, well, how does a small company get set up without us spending tons of time with them?

But then we still had lots of small customers. And you kind of get caught in this thing of trying to serve too many masters and not being great at any of them. And I think Shopify served small and mid-sized businesses up until post-IPO.

And sure, Shopify now supports enterprises because we’ve got the depth and breadth of a platform to actually do that in a scalable way and have it not turn into a bunch of if statements in the code that’s like if Supreme or if Staples.

But Shopify was able to hold on to that focus for a really, really, really, really long time. And I think some of that is just that Tobi’s very focused. But I think there’s also sort of the spiritual home of Shopify that has always been the entrepreneur. And so I think Tobi just gravitates towards how you help someone start a business.

Mark MacLeod:

I was only there part-time, but at that time, I remember large merchants. I remember Mozilla years ago wanted to set up a merch store and they had a huge reach. This is like maybe 2009, 2010. And Tobi was just like, no, they wanted an SLA, they wanted these things, why are we even having this conversation? Just go away. Like, laser focus.

Glen Coates:

Yeah. Extremely principled and I think I was not as principled. And I think if I could do it again, I would be a bit more ruthless about staying focused on how do I become insanely good at serving these people.

And then I would say, make sure that whoever your target customer base is, don’t become American Airlines. You know what I mean? Sometimes you can become insanely good at serving a customer segment that just isn’t a good business and just be a bit rational about that.

It’s a really hard balance. You have to be rational about the business but you also have to be irrationally walking through walls all the time and it is a hard balance to strike.

Mark MacLeod:

Yeah, yeah, for sure it is. That’s why most of them fail. And so the point on focus, I think we’ve more than covered it. I completely agree with it. And this notion of not giving up the review and being in all the details. There’s a term for that now, it’s called Founder Mode. It’s like. And Tobi actually personifies it and always has. So I love that.

You know, so staying on the notion of kind of lessons from this other side now, this whole other scale, you’ve touched on some of this already and it probably could be an entire podcast but what else have you learned from Tobi?

Glen Coates:

You sort of learn something from Tobi every day because he’s such a voracious learner. Like a couple of years ago when I would say Shopify was not a very AI ML-centric company three years ago. And Tobi is known for being a very strong engineer. And I think when he looks in the mirror, he probably sees an engineer. But I wouldn’t say he saw an ML engineer in the mirror three years ago.

Then GPT3 came out and it was just like crazy to watch him go from being someone who probably didn’t know that much about ML to just reading every paper when it came out and just trying every open-source model on his local machine and just becoming like, oh, if you prompt it like this, this is going to happen. And if you say the words in a slightly different order, this other thing’s going to happen.

And he was just like, so close to the metal and such a voracious learner. And I think it’s sort of a reminder that if you want to build a company like Shopify or even a smaller one that stays right there on the cutting edge, and I think that’s a lot of what has made Shopify successful is that it’s never really fallen behind the times.

And some of that is just being ruthless about always making sure you know what the latest technology is and learning. And he sets a high bar in that way. And I think if there’s anything, that’s been one of the things. I mean, there are lots of other smaller things. But I would say especially lately because it’s like every week something’s changing with even small examples.

Like we’re playing with things now where there’s some migration thing that’s happening at Shopify right now where we need merchants to move from an old version of a thing to a new version of a thing. And it involves some code changes. And you and I have been around long enough to remember how these things usually go. And it’s like, oh man, like Magento one to Magento two, and like oh my God, this is going to be a years-long death march.

And it’s even crazier to watch. Like, we tried getting GPT4O to help with this migration. You know, the evals passed 0% of the time or maybe one out of a hundred. And then we think, oh, this isn’t going to work. Shit, I guess we’re going to have to do the more manual version of this, bummer.

And then you come back a week later and you run the same thing with reasoning set too high and you’re like, oh shit. It actually passes 85% of the evals, maybe this can work. And I bet you next week it’s going to be 95%. And like, right now, the premium on staying up to date and learning and making sure the thing you thought was impossible a month ago, is it actually still impossible? The premium is extremely high and this is again like one of Tobi’s greatest strengths is just being relentlessly right on that cutting edge.

And so that actually makes me feel, I don’t know if you went out on the street and said like, is Shopify an AI company? I don’t know if many people would say yes. But at some level, being an AI company is sort of about how quickly you are ingesting and learning and applying the stuff. And in that sense, I think Tobi’s culture and rate of learning make you the best kind of AI company.

Mark MacLeod:

I mean, if you think about it, Shopify just celebrated its 20th anniversary, I think from incorporation this past summer. That would be old by some standards. And at the time SaaS was reasonably new. I was, you can tell from the color of my beard. I’ve been around for a bit. Like I was there for the birth of the Internet. I was there for mobile. I was there when I forgot what the acronym was, but it became SaaS.

And this whole paradigm, SaaS is essentially a database and you have to do work to get value from it. And now the software is just going to do shit for you. Like it’s completely changing. And I have some companies that are clients that are just pure AI companies and they are disrupting themselves. Like, they haven’t even recouped the dollars they’ve invested in V1 before they’ve disrupted V2. And it’s just like a race, so it’s just fascinating to see where this is going to go.

Glen Coates:

I’m actually curious what you think here Mark, where if you went back a year, it’s funny how quickly it went from the frontier model companies like OpenAI, Anthropic, etc. Like, it was sort of like, okay, they have the secret sauce, they have the magical thing, they have the thing you can only get if you throw billions of dollars into training.

And then it moved into like, oh well actually Llama 405B sort of does it open source. And then it’s like, okay, now it’s deep seq. and the inference cost is radically lower. And I mean not to bring it back to American Airlines, but do you think it’s possible that some of these frontier models may turn out to be not so much this rare unique gem? They may turn out to be more like Linux, this thing that’s incredibly valuable but distributed almost at near zero marginal cost and they become almost like the human commons and super duper valuable, but maybe not the sort of like airlines.

Mark MacLeod:

It’s a great question. I feel like I would need to be a lot more technical to have the canonical answer. Like, I need to be the person doing the research like you and Tobi are. I think DeepSeek, open sourcing, and then Perplexity being able to stick that model in the next week, those kinds of developments are massive.

If I was Sam Altman, I probably wouldn’t be too happy about that. But AI is just now going to be the thing, right? And so when it’s everywhere, there’s different ways, different paths to success I guess is the point. Maybe some of your direct end-user monetization will go away in this world of really compelling open-source models or costs coming down. But then embedding the model like the deal that OpenAI has with Microsoft and then providing all the enterprise support, that’s a great business. So, different ways to skin the cat, I guess.

Glen Coates:

I just came from watching a demo video on this topic that I think is just one of these things that happened in the last week that’s so crazy and it’s such an amazing application to what we do in terms of entrepreneurship. The thing we’re trialing at the moment, we have Sidekick, which is the thing in Shopify where you can talk to it, and it can help you out with doing stuff inside of Shopify.

We’re trialing out this thing at the moment, which is basically a voice in a screen share mode where you can log into Shopify. You can say, like, start a call, and you start a call with Sidekick, but it’s sharing your screen, so it’s like actually reading the pixels and you can talk to it in voice mode.

And I saw a recording of one of some of the people who we did the trial with, and it was a recording of a woman I think in Sri Lanka who starts talking to the system and saying, hey, how do I use this system? I’m trying to save up to buy a laptop, how do I do this? And she gets confused because she’s not a native English speaker, English is her second language.

Halfway through it the agents are talking back to her, and then she says she can’t find the words, and she says, oh, can I talk to you in Hindi? And the agent says, oh, do you want to talk to me in Hindi? And she says, yes. And then it just switches into Hindi, and it just keeps going and then she figures it out. And it’s just things like that where you’re just like, oh, my God.

Mark MacLeod:

Honestly, the future is here. I’ve thought about that a few times, but I really think it’s fascinating to watch. And I’ll say I have a very vested interest in Shopify’s success. It’s by far the single largest business portion of my net worth. And many times over the years, my financial advisors have been like, hey, dude.

They were probably not as informal, maybe it’s time to diversify. And I’m like, no, Tobi’s a pretty special founder. As long as he’s motivated, I think I’ll stick around. And I would argue that’s why you’ve stuck around by all these amazing founders. And so I’m fascinated by your journey. And I’m really thankful that you took the time to chat today and it’s awesome to see you virtually.

Glen Coates:

Yeah, you too, Mark. It’s great to see you.

Mark MacLeod:

All right, thanks so much.

Hey, thanks for listening to the Startup CEO Show. If you’d like to connect with me, be sure to visit my website at markmacLeod.me, or follow me on LinkedIn at The Mark MacLeod, or X account @markmacleod_, and if you want to tune in again next week, be sure to subscribe on YouTube, Spotify, Apple, or wherever you get your podcasts. We’ll see you next time!