Should you buy or build? Hire an engineer? Go headless? These are the questions that can consume DTC brands, and there’s no one-size-fits-all answer. We were joined once again by Ben Kochavy, who helped shed light on the questions above. The key takeaway is this: there’s a time for brands to live on a platform like Shopify, and a time to go custom. There’s also a middle ground, where you can build custom apps for your Shopify-based business. What you don’t want to do is drown your early-stage company in engineering costs, or hobble your growth with systems that prevent agile pivots.

Show Topics

  • When to buy versus build
  • The pros and cons of custom
  • Why Shopify is the best bet for new brands
  • Tech doesn’t always add value
  • Invest in custom apps
  • Aim for fewer failure points

Show Links

Show Notes

3:02 - When to buy versus build

A company’s decision to buy or build should ultimately come down to what will provide them with the greatest agility.

BK: “I've worked in organizations which are like, founders are both engineers. So the solution to everything is build. It's never buy. Because they went into engineering, and good for them. I’ve been in other organizations where it's like a middle ground, product-led, and other organizations that was marketing-led. A lot of it's affected by the organizational tone in terms of, the way the organization operates and thinks about running a business, affects how they make these decisions. But then on the other side, from a growth perspective, the way I look at it is I think agility is the ultimate goal when you're doing growth. You want to be able to pivot and change little pieces of that experience, large pieces of the experience as fast as possible. Whatever lets you do that best is the right answer.”

4:26 - The priority problem

A team of in-house developers could certainly be the way to go. But consider if what you build will slow you down, such as this example from Ben Kochavy.

BK: “I work with a brand that has to use developers to integrate everything. So they integrated MailChimp. They now have this thing where they want to integrate Klavio but they can't, because it's going to cost them $50,000 and they need the devs to do something first. And then on the other side of it, they haven't even hit SMS yet as adding it as a marketing channel. Because of the same level of integration that you have to do for Klavio you have to do for SMS to do it well. And so you get hit against the priority problem. And then like, not cashflow, but resourcing problem. In a situation where you don't have unlimited cash or unlimited resources, you start to make trade-offs. And I think for growth people, at least in my experience, agility slash momentum above all else will get you very far. But then if you don't have the developers to support, you've got to buy it and then customize to the best that you can without the developer support.”

5:28 - The pros and cons of custom

A custom solution can set you apart in a good way. But if you want to move quickly, be prepared for lags.

BK:“I worked with a men's fashion brand. They are selling curated underwear to men. The whole tech stack is built on, I believe it's called Spree. It's all custom, right? They need to dev to do basically anything. They have a really custom experience. They have an amazing way of cataloging the different types of fashion they have, understanding people's fashion tastes, and then selecting for them products at a level it's probably more sophisticated than your average online quiz that you see from D2C brands. Because they actually have to understand and respond to style. And they can't just do ‘if this, then that’ for two or three selections. So they got pretty deep in there. But as a result, they can't pivot on anything. Anything becomes a feature build. It's like three weeks QA, bugs. You have to make sure everything's working. Check out goes down, you have issues, all that stuff's custom built. And I mean, custom isn't always bad. But if you're moving very quickly, custom can be bad.”

9:45 - Why Shopify is the best bet for new brands

If you’re an early-stage company, there’s security in situating yourself in the Shopify ecosystem. Don’t let early engineering costs drown you from the get-go.

BF:“The agility argument makes a lot of sense to me. And it seems like there's certain circumstances where the answer is obvious. If you're a small, early-stage company. Yes, you need to be agile. Yes, you don't have many resources. Yes, there's a high likelihood that you're just building stuff and throwing it away. You should just be starting from a platform like Shopify. You know, maybe at most you invest in building some custom app functionality. But the pros and cons of building your own thing, or even let's call it the risks and the costs of what goes wrong if you choose Shopify, versus what goes wrong if you decide to do a build, it's completely different economics. Because what I've seen happen is that if you're a new brand and you decide to build, the problem is that you might not ever launch. You might burn all of your money in engineering.”

11:22 - Tech doesn’t always add value

If your goal is to sell your company, the cost of maintaining an expensive in-house tech team may drag you down later.

BF: “There are reasons to perhaps invest in engineering and invest in building a tech team. But it comes at a cost. And that cost, I think, is felt down the road. If you're a brand and your ambition or your goal is to sell for 25, 50, 100 million dollars, it might be based in part off of how quickly you're able to iterate and test and stuff. But the fact that you have a tech team is not going to improve your evaluation or your resell value. What it does do is it introduces a lot of potential issues. It introduces issues around technical due diligence. If a brand's acquiring you, whether it's like Colgate or a private equity firm, which is more likely I suppose, if your stuff is custom built or built on headless, fairly nascent tools, the risks then become well, what are the maintenance costs? What's the ability of this product to be able to integrate into our portfolio of brands? What are the costs required to even do proper due diligence to understand what we're acquiring?”

15:54 - Invest in custom apps

Your best bet isn’t necessarily in going full custom. It’s in customizing what you can, like apps, while your business still lives on Shopify.

BF: “My thesis is that you can accomplish a lot. You can get really, really frigging far building and scaling a business on Shopify and Shopify Plus, if what you do is you invest in custom functionality or boutique functionality in the places that matter. The typical argument is, I don't want to rely on 15 different apps that may or may not talk to each other to grow my business. My go-to in general would be once you get to the place where you can afford to invest in engineering, build a custom app that integrates. Basically builds the functionality of seven of those apps that are otherwise interconnected, right? Find the natural interconnectedness of some of these apps and either build integrations yourself, or more likely what can happen is you can build a customer private app that accomplishes a lot of those different things that aren't really integrated properly. And you can do it on top of the Shopify platform, right? It's not that you have to rely on Shopify’s existing app ecosystem. You should be investing in building your own custom apps that solve your own relatively specific problems.”

17:29 - Get as big as you can

Ben Fisher’s take is that you should grow your business as much as possible on Shopify before you start running up again problems that you can throw cash at yourself.

BF: “There are still most definitely some trade-offs of being on a platform like Shopify. A lot of what's been said has been around like control of the checkout page. A lot of people want a one-page checkout. Shopify is not going to let you have a one-page checkout. And when you upgrade to Shopify Plus, you certainly have a lot more control over the checkout experience. You don't have full control. I wouldn't be building a brand because I don't know enough about all this. But if if I was a part - I've done fractional CTO going for a subscription eCommerce brands. That's what I did after CartHook for a bit. In general, I would advise people to get as big as you possibly can on Shopify. And in the event that you get so big and you have so much money that you can afford to pay several engineers $170,000 a year, or maybe less depending on your market, then you know, maybe you do that. But I'm not really sold on this idea of all of these trade-offs that people attribute to not building on Shopify.”

19:29 - Pay attention to the 1%

Once you reach a certain success rate, small incremental inefficiencies can start adding up. That’s when it might make sense to scale your in-house team.

BK: “It's that middle ground. I think checkout is something that came up a lot is that if you mess that up, and even one line of code that doesn't work on one browser and one like little phone that no one has, that's worth a lot of money at scale. And solving for that without an army of developers, an army of product managers, and like a designer or two is very hard. So you can cover 80%, 90%, 99% of the use case. But that 1% eventually becomes worth a lot of money. It just depends on what scale you're at. And when it becomes worth a lot of money is when you should hire that army I just described. But it's not even just engineers. It's a product manager, probably some like UI, UX designer to make sure it's working the way you want. So it's like one, two, and then maybe two or three engineers. That's like a five, six person team full-time. Not cheap each person, right? That problem better be a lot more expensive than that combined salary count.”

22:25 - The definition of headless

A headless operation can be defined as an ecosystem of multiple products where a different individual company tackles a different thing to make your overall business work.

BF: “When you say headless, you're thinking of Shopify headless. When I think of headless, I'm thinking of no Shopify at all. I'm thinking of either using, I don't know if Bolt is technically headless. But when I think of headless, I think of we’ll call it an ecosystem of multiple products. You have a page builder, you have a CMS, which would probably be like Contentful or something. You basically have a suite of independent companies, all of whom focus on a specific portion of the constellation of jobs that need to be done to support an eCommerce business. I'd say, the cynic in me is: when you're relying on 12 different products, even if they're directly integrated, won't you still fall into the problem that people complain about in the Shopify app store” Of, I have 12 apps, they all kind of do something. None of them necessarily talk to each other correctly. It's requiring me to have 12 different vendors to run my business. And it's exhausting...with headless, at least you have people specializing in certain aspects of the constellation.”

25:20 - Aim for fewer failure points

The ultimate goal for the way you set up your business should be efficiency. Explore methods where you can streamline and integrate.

BF: “Part of the problem with apps is that you're also building on top of Shopify. So you as an app developer actually don't have visibility into the foundation of the home. You are simply a veneer. In some cases, you have little holes in the roof that you are able to stick your little wire into. I do think that in headless, these brands were able to have much deeper integrations in theory. But you still do have this coordination issue, where people can potentially start blaming each other. And then again, it might just come down to in headless, you don't want to have 15 different vendors. You want maybe three. And from what I understand, Shogun has done a really good job. Because they're not just creating the landing page, they're doing multiple things related to the page building. And so what ends up happening is it is a benefit. Because rather than having seven apps, you have effectively one app. Which isn't so different from what I was advocating for within the Shopify place, which was: build a custom app that does the job of five apps that would otherwise be having to integrate with each other. Because when you have one app, there's fewer places that it's going to fail.”