Build vs buy in the AI era: should brands replace their Shopify apps?

October gave Shopify brands a hard lesson. A checkout-hijacking campaign came in through third-party apps. Apps were pulled from the App Store, and merchants got security notices about tools they had trusted for years.

For a brand, that's about as bad as it gets. A brand is built on trust and reliability. When your checkout is hijacked, that trust breaks in front of your own customers.

So it's no surprise that brands are now asking a sharper question: do we need all these apps, or should we build it ourselves?

And with AI, building yourself looks easier than ever. A capable team, or even one person with a good AI assistant, can put together a working reviews widget, a quiz or a video carousel in a weekend.

Brands are right to ask this question. Fewer third parties really does mean fewer ways in. But the honest answer depends on one thing most "build it ourselves" plans leave out.

Building is the easy part. Maintaining it is the hard part.

The first version is never the expensive part. The cost comes in the months and years after. When you build your own tool, you also take on everything the app team used to handle for you.

1. You own its security, forever

Every custom app you run is now part of your store's attack surface, just like a third-party app. Someone on your team has to:

• check every request it receives, and make sure it trusts nothing anonymous;
• make sure nothing typed into its settings can ever run as code on your storefront;
• keep its dependencies patched as new vulnerabilities come out;
• lock down the internal tools and admin panels around it;
• keep up with Shopify's API changes and security requirements.

That isn't a one-time job, and it doesn't stop when the person who built it moves on. October showed what happens when one layer gets missed, and it happened to app teams whose full-time job is this.

2. You own the roadmap

Markets move. Shopify ships new features, customers expect new formats, and your competitors keep improving. A tool that was good enough this year needs new features next year. Every hour your team spends maintaining and extending internal tools is an hour not spent on your products, your marketing or your customers.

3. You own the product decisions

When you build, you decide which features you need and which you don't. That sounds great, but it comes with a real risk of judging wrong, and it's an easy mistake to make. You're not an expert in what the app is supposed to do, at least not compared with a team that spends day and night obsessing over that one problem. That team knows which features matter most and which matter less.

So build only when you're sure about the features you're building, and even more sure about the hundreds of features you won't be building. Otherwise development costs keep climbing until they're higher than the app would have cost you for the next several years.

4. You lose the view from thousands of other stores

This is the part that's easy to underestimate. A good app learns from the data of every store it runs on: which formats convert, which placements work, what changed when shopping behavior shifted. Those insights come back to every merchant as better defaults and smarter features.

A tool you build yourself only ever learns from your own store's data.

5. You take on the hidden complexity

Some apps look easy to build on the surface. Then you start building and discover the deep work underneath that made them look so simple and smooth.

Shoppable video is a good example. The shoppable part was never the hard part, even before AI. The hard part is playing videos on your website without slowing your store down. Even images weigh a store down significantly. The level of engineering work needed to load potentially hundreds or thousands of high-resolution videos across a website without hurting its speed or the video quality is no less than rocket science itself. That means more than just how fast the page first loads and settles (LCP and CLS). It also means how fast the page responds when shoppers tap, swipe and scroll (INP). Honestly, many full-time shoppable video companies still haven't cracked it after years of trying.

And speed is money. Even a one-second delay on your store can cost you enough sales to wipe out all the savings you expected from building it in-house.

6. Do the real math

Before you decide, work out what building will really cost you, not just the first version:

• The one-time build: the development hours to design, build, test and launch it.
• The upkeep: the development hours every month for security fixes, Shopify API changes, bug fixes and new features, for as long as you run it.

Development hours are expensive, and the upkeep never stops. Over a few years it often costs more than the first build.

Then compare that total with what apps cost over the same period. If you find an app that costs about the same, less or even slightly more, give it a serious look, along with everything above: security, roadmap, product decisions, data and the hidden complexity you won't have to solve yourself.

So when should a brand build?

Building has one real advantage: no app can ever be as customised as the one you build yourself. It fits your brand, your workflow and your customers exactly.

Build when the tool is truly core to what makes your brand different, when no app does what you need, when no good app in the market fits your current or upcoming scale, when you're ready to own its security and upkeep for as long as you run it, and when the real math still favors building.

Buy when the problem is common to many stores, when it benefits from data across the market, or when your team's time is better spent on your brand.

Most brands will end up with a mix, and that's fine. The real goal is fewer, better-chosen dependencies, not zero.

What brands should ask every app they keep

If you're keeping an app, hold it to the standard you'd hold your own team to:

• Security: How do you protect my store and my customers' data, and how fast will you tell me if something goes wrong?
• Roadmap: How often do you ship improvements, and what's coming next?
• Insights: What do you learn across other stores and market that I couldn't learn on my own?
• Engineering depth: What's the hard part of what you do, below the surface, and how have you solved it across the whole product? (For a video app, for example, keeping a store fast with hundreds of videos on it.)
• Real cost: What will you cost me over the next few years, compared with building and maintaining it ourselves?

A good app team will answer clearly. If they can't, that's your answer.

What we app teams owe brands now

Brands are right to want fewer dependencies, so the apps that stay have to earn their place. That means taking on the security and maintenance burden so brands don't have to, being open about how we protect their stores, and bringing insight they couldn't build alone.

We took a hard look at Reelv this month with exactly that in mind. You can read what we changed in how we upgraded Reelv security this October.

If you're weighing build vs buy for shoppable video on your store, we're happy to talk it through honestly, even if the answer is to build. Write to us at team@reelv.live.

— The Reelv team