Shopify vs Custom Build for an Online Grocery Store

We build custom e-commerce for a living, so take this as it's meant: most new online grocery stores should start on Shopify. The interesting question is which ones shouldn't, and why.

Where Shopify wins outright

• Live in days, not months, for a few hundred dollars plus theme cost.
• Payments, tax, PCI compliance and fraud handled by someone else.
• Checkout tuned by a team with more conversion data than any agency will ever see.
• An app for nearly everything, including basic delivery slots.
• No maintenance burden and no security patching on your plate.

If you sell 200 to 500 packaged products at fixed prices to one delivery area, stop reading and go set up Shopify. You'll be trading this week.

The four things that break it for grocery

1. Variable weight items

Standard e-commerce assumes one unit equals one price. Grocery doesn't: a customer orders 1.5 kg of chicken, the picker weighs out 1.62 kg, and the price must change after the order is placed. Shopify can approximate this with weight-increment variants or a post-purchase adjustment, but it's a workaround, and it gets uglier as more of your catalogue is sold by weight.

2. Capped delivery windows

You can only deliver 30 orders in the 5–7pm slot. Once 30 are booked, that slot must vanish from checkout. Several Shopify apps do this adequately; none do it as well as purpose-built logic, and app limitations around slot capacity per zone per day are a common reason grocers eventually migrate.

3. Catalogue scale and search

At 3,000+ SKUs, browsing stops working and search becomes the store. You need typo tolerance, synonym handling (yoghurt/yogurt, aubergine/eggplant, coriander/cilantro), and filters that stay fast. Shopify's native search struggles here; a paid search app helps, and the monthly cost starts to approach real money.

4. Substitutions and picker workflow

Items sell out between order and pick. You need the customer's substitution preference captured at checkout, passed to the picker, and reflected in a partial refund. You also need aisle-sorted pick lists so staff don't cross the shop forty times. This is operations software, and it's where Shopify's model runs out — it ends at the order, and grocery's hard part starts there.

Cost over three years

The crossover usually lands somewhere between year two and year three — earlier if you need several apps, later if you stay simple. Cost alone rarely justifies custom in year one; capability is the real reason to move.

A straight decision rule

Choose Shopify if most of these are true:

• Under ~1,000 SKUs, nearly all sold as fixed-price units.
• One store, one delivery zone, and slots you can manage loosely.
• No dedicated technical person on the team.
• You need to be live this month.

Choose custom if most of these are true:

• A significant share of revenue is sold by weight.
• Multiple stores or zones with separate stock and capacity.
• Strict slot capacity tied to your actual driver count.
• You need picker workflows, substitutions and refunds handled properly.
• You're already paying for four or five apps to patch gaps.

If you do migrate, protect your rankings

The most expensive migration mistake is changing URLs without redirects. Shopify forces paths like /products/item-name; a custom build may use /shop/category/item-name. Every changed URL needs a 301 redirect from the old address, or you lose the rankings you spent years earning.

Export your top 200 URLs by traffic from Search Console before you start, map each to its new address, and verify every one after launch. This is an afternoon of work that protects years of traffic.