When a Shopify Plus store should build custom instead of installing an app

We sell apps on the Shopify App Store and we build custom for Shopify Plus merchants. That puts us in a strange spot when someone asks “should we build this or install something?”, because the answer costs us money either way. So here is the test we actually use, written down.

Is the process yours, or everyone’s?

Every store needs product badges, delivery estimates, a returns flow, gift cards. Those are everyone’s problems, which is why there are forty apps for each of them and why an app is the right answer. The app has seen a thousand stores, it has handled the edge cases, and it costs less per month than one hour of custom work.

The trigger for custom is a process that is specifically yours. A workwear retailer whose corporate clients fund safety gear for their employees. A B2B seller whose buyers need per-person caps and a manager’s approval. A brand whose pricing rules have three “except” clauses that no settings screen can express. No app will ever be built for that, because there are not a thousand stores with your process. There is one.

The tell is how you describe it. If you can name the category (“we need a loyalty app”), install one. If you have to tell a story (“so what happens is the client sends us a list, and then Dana…”), that is a custom build, and Dana is the person it will give their week back to.

Does real money move through it?

The second test is stakes. An app that shows a badge on a product can be swapped next quarter with no harm done. A system that holds $35,000 of a client’s store credit, or produces the number you invoice them for, cannot. It needs a database with backups, monitoring, someone on call, and a report that gives the same answer twice.

That is the reason we only take custom work from Shopify Plus merchants doing more than $2M a year. It is not gatekeeping. Below that size, the cost of running the system properly is bigger than the program it runs, and one of our apps plus a spreadsheet is honestly the better deal. We say so on the first reply.

Native first, custom on top

The mistake we see most in custom Shopify work is building a parallel system. The agency has its own ledger of balances, its own customer table, its own invoice store. It works until the agency relationship ends, and then a program worth real money is stuck in a database nobody can touch.

The better pattern, and the only one we build, is native first. Store credit is Shopify store credit on the customer record. Companies are Shopify B2B companies. Rules run as Shopify Functions inside checkout. The custom app supplies what Shopify does not have (pools, rosters, product rules, approvals, reconciliation, integrations) and nothing else. Switch the app off and the balances and the history are still in your store. Read our custom Shopify Plus development page for how that shapes an engagement.

What “custom” should cost you in time

A written spec in your own words within a week of the first call. A fixed price from that spec, so the number does not move. Working software on your store from the first week of the build, not a demo at the end. Live in three to six weeks for the core. Anything that is not needed on launch day, like a client self-serve portal, pushed to a second release on purpose so the launch date never depends on it.

If a proposal you are looking at does not have those five things in it, ask why.

A quick way to decide

Write the process down in one paragraph. If the paragraph contains a category name, install the app. If it contains a person’s name and the word “then” more than twice, and there is real money in it, it is a build. If you are on Shopify Plus and over $2M a year, send us that paragraph. If you are not, install Bulk Store Credit and keep the paragraph for later.