Here is a pattern I have watched three times now, from inside and from outside.
A retailer has been on Salesforce Commerce Cloud for years. The site works. The license is not cheap, the release cycle is slow, and every quarter someone on the business side asks why the brand across the street gets features in a week on Shopify. So the team builds a proof of concept. They stand up a Shopify store, load a slice of the catalog, theme it, and show it to leadership. It looks great. Everyone agrees it looks great.
A year later they are still on Commerce Cloud. Nobody made a decision. The proof of concept is, oddly, part of the reason.
What the POC proved
It proved the storefront moves. That was never in doubt. A catalog of a few thousand products, a theme, a checkout: a competent team gets that running in weeks. The POC answered a question nobody was really asking.
The question people were asking, usually without saying it out loud, was: what happens to everything else?
The everything else
I counted the integrations on one Commerce Cloud storefront this month. Nineteen. Order management, the ERP feed, tax, payments, fraud, search, reviews, loyalty, email, SMS, site personalization, the returns portal, the store locator, the gift card provider, the product information system, analytics, consent management, the help desk, and a custom piece for same-day delivery that someone built in 2021 and nobody has touched since.
The storefront is the part the customer sees. The nineteen integrations are the business. Every one of them has a contract, an owner, a data shape it expects, and a quiet set of exceptions that only matter at the end of the quarter.
A POC that loads a catalog and a theme touches two of the nineteen. The other seventeen are where the cost, the risk, and the timeline live. And they are where the decision gets made, or more often, not made.
What to test instead
If I were scoping a proof of concept today, I would skip the theme entirely and spend the time on five things.
The order handoff. Place an order on the new platform and watch it arrive in the order management system with the right line items, the right taxes, the right promotion codes, and the right store attribution. Then cancel it. Then return half of it. If this does not work end to end, nothing else matters.
The catalog feed in reverse. Not loading products into Shopify, but getting inventory, prices, and availability back out of the ERP into Shopify at the frequency the business expects. Most Commerce Cloud sites have years of custom job logic here. It does not come along for free.
Loyalty and customer identity. Who is this person, what tier are they in, what have they earned. The customer record is split across three systems on most retail stacks. Pick one customer with a complicated history and make them whole on the new platform.
The one custom thing. Every mature storefront has a piece nobody will admit is custom. Same-day delivery, a configurator, a B2B price list, a store pickup rule. Find it. Rebuild that in the POC. If it cannot be rebuilt, you have your answer.
Search and merchandising rules. The merchandising team has spent years tuning sort orders, boosts, and redirects. Ask them to recreate one week of their work on the new platform and time it.
Do those five and you will have a real number for the migration and a real list of what breaks. You will also know whether staying and modernizing the Commerce Cloud build is the better move, which it sometimes is.
What I would tell the board
I have no stake in which platform wins. I have built on Commerce Cloud for most of my career, and I have moved brands to Shopify when that was the right call. The storefront is the cheap part of either answer.
What costs money, and what nobody wants to put a number on, is the nineteen things wired into the storefront today. Can they move? What does each one cost to move? Who on the team owns it on the other side? Those three questions have real answers, and a two-week proof of concept aimed at the handoffs will get most of them.
Run that version instead of the pretty one. You will come out with a migration budget you can defend, or with a clear reason to stay and modernize what you have. Both beat another year of not deciding.