Ran into this pattern enough times building e-commerce sites that it's worth writing down.

Most custom storefront builds don't blow up because of code — they blow up because scope was never actually fixed. The client says "just build me a clean, modern store" and the dev says "sure." Three weeks later there's a Slack thread with 40 messages about whether the cart drawer should slide from the left or right.

The fix that's worked every time for us:

1. Lock the page list before any design starts. Home, PLP, PDP, cart, checkout, account. If a page isn't on that list on day 1, it's a phase 2 conversation, not a "can you just add" conversation.

2. One round of design approval, in writing, before dev starts. Not "looks good in the call" — an actual message that says "approved." This is the single biggest timeline killer we've eliminated.

3. Cap revisions up front. Two rounds. Not because more feedback isn't valuable, but because unlimited revisions means unlimited timeline. Clients respect this more than you'd think — it forces better feedback, faster.

4. Ship the checkout flow first, not last. It's the part that actually makes money and the part most likely to have technical surprises. Everyone builds the homepage first because it's fun. Don't.

If you're an e-commerce brand about to commission a custom build — ask whoever you're hiring these four questions before you sign anything. If they don't have a clear answer, you're going to have a bad time.

— Haecom (we build custom storefronts, fixed scope, fixed price, 3-week turnaround)