Most online stores start on Shopify, Wix or WooCommerce, and that is the right call. A template builder gets a catalog live in days and costs a few hundred dollars a month. The question that actually matters is when to stop treating that as permanent and bring in an ecommerce web development company to replace or rebuild parts of the stack.
That moment usually arrives earlier than founders expect, and it shows up as a conversion problem before anyone calls it a technical one.
The checkout math most stores never run
Baymard Institute tracks online shopping cart abandonment across checkout studies and puts the average documented rate at 70.22 percent, averaged across 50 separate studies. That number has barely moved over the years it covers, which suggests abandonment is not mainly a marketing problem. It is a checkout design and performance problem, and template builders cap how much of it you can fix.
A stock checkout from a builder gives you a handful of layout options, a fixed set of payment and shipping fields, and little control over what loads before a shopper can pay. Past a certain order volume, every percentage point of abandonment you cannot touch is real revenue, and it is usually the first thing a serious rebuild targets.
The speed number that changes the ROI conversation
Google commissioned Deloitte Digital to study what happens when mobile sites get faster, in a report called Milliseconds Make Millions. Deloitte measured retail, travel, luxury and lead-generation sites across Europe and the US over four weeks, tracking the effect of a 0.1 second improvement in mobile load time.
For retail sites, that 0.1 second gain lifted conversion rate by 8.4 percent and lifted average spend by 9.2 percent. For travel sites, conversion rose 10.1 percent and bookings rose 10 percent. The catch in the methodology: the gain only showed up when every page in the journey improved together, not just the homepage.
That is the part a builder struggles with. You can compress one hero image on a landing page. You usually cannot touch the app bundle, the third-party scripts a theme ships with, or how the platform renders the path from product listing to checkout. An ecommerce web development company that owns the full stack can.
What actually changes when you move off the builder
The rebuild is rarely the whole site. It is usually three things, in this order of how often they get requested:
- Checkout. Custom field logic, saved payment methods tuned to your actual customer base, and a path that drops unnecessary steps for repeat buyers.
- Catalog and search. Builders handle a few hundred SKUs fine. Past a few thousand, with variants, bundles and real-time stock across locations, template search tends to fall apart and shoppers bounce before finding what they want.
- Integrations. ERP, warehouse management, loyalty programs and marketing automation tools that a builder's app marketplace only half-connects, leaving someone on the team reconciling spreadsheets every week.
None of these need a full platform replacement on day one. A competent web and mobile app development partner can often rebuild checkout and keep the rest of the storefront running on the existing platform while the harder pieces migrate.
Builder vs custom build: a quick way to tell which side you are on
- Under 500 SKUs, one warehouse, standard payment methods: stay on the builder. The cost of a custom build will not clear in revenue terms for years.
- Multiple warehouses, B2B pricing tiers, or subscriptions mixed with one-off purchases: you are likely past what the builder's logic can express cleanly.
- Checkout abandonment above the Baymard benchmark with no visibility into where shoppers drop off: that is a sign the platform is hiding the problem, not that the problem does not exist.
- Mobile load time regularly over 3 seconds with no engineering capacity to fix it inside the current platform: this is where the Deloitte numbers above stop being theoretical and start being a monthly revenue figure you can estimate.
Questions worth asking before you sign a contract
Most vendor pitches sound similar until you ask specifics. Ask exactly what checkout and payment gateways they have shipped before, not just that they "support payments." Ask how they handle catalog data migration without losing search rankings on existing product URLs, since a careless migration can quietly erase a year of SEO work. Ask for one reference client whose order volume is close to yours, not their biggest logo.
Also ask what happens to your data and your domain if the engagement ends. A dedicated build should leave you owning the codebase and the hosting account, not locked into a vendor's proprietary platform with a login you do not control.
What a rebuild usually costs in practice
Pricing varies widely by scope, but the pattern is consistent across vendors: a checkout-only rebuild on top of an existing storefront is the cheapest and fastest piece, a catalog and search overhaul costs more because it touches data migration, and a full custom storefront replacing the builder entirely is the largest commitment and the one that should come last, not first. Most stores never need that third option. They need the first one, and sometimes the second.
A short discovery phase before any contract is signed should tell you which of the three you actually need. If a vendor skips straight to quoting a full rebuild without first asking for your current abandonment and load time numbers, that is worth treating as a flag rather than confidence.
The decision in one sentence
Keep the builder until a specific, named limitation - checkout flexibility, catalog scale, or an integration the app store cannot cover - is costing more in lost orders than a rebuild would cost to commission. Past that point, the Baymard and Deloitte numbers above are the argument, not a sales pitch.