Adding a shop to a website sounds like one more page. It isn’t. It’s the point where your
website stops being marketing and starts being operations — and where the gap between “we can
take orders” and “this actually works” gets expensive.
Most of what follows is invisible to customers. That’s the problem with it. A store either
feels effortless or feels like a hassle, and the reasons are almost entirely behind the
checkout button.
Here’s the honest list of what a real build involves, so you can tell whether a proposal has
accounted for it.
Start with the uncomfortable question
Before any of the technical decisions: who is doing the fulfilment, and how many
orders a week before that person is overwhelmed?
We’ve seen more online stores fail on operations than on technology. The site works, the
orders arrive, and nobody has worked out who packs them, where the stock lives, what happens
when something is out of stock, or who answers the email asking where an order is.
Build the store to match the operation you actually have. A store that can take two hundred
orders a day is worthless if you can ship nine.
Products: more work than anyone expects
This is the single most underestimated part of every e-commerce project, and the reason most
of them run late.
Each product needs a name, a description that answers real questions, a price, a SKU, weight
and dimensions (shipping needs them), a category, and photographs that are consistent in
lighting and background and cropped the same way. Variants multiply it: three colours in five
sizes is fifteen combinations, each with its own stock count and possibly its own image and
price.
Eighty products is a serious data project. If it isn’t scheduled as one, it becomes the thing
everyone is waiting on while the build sits finished.
Worth saying plainly: product photography is usually the highest-leverage money in an
e-commerce budget. Customers can’t pick the thing up. The photographs are the product.
Payments
Taking cards is the easy part. The decisions around it are where it matters.
You’ll want a reputable processor, and you want to understand the fee structure — per
transaction, monthly, chargeback fees, payout timing. Payout timing in particular: a few days’
delay changes your cash flow in a way nobody mentions up front.
On the build side, the non-negotiables: the checkout is served over HTTPS, card data never
touches your server, and your integration is current with the processor’s requirements. Keeping
card data out of your own systems is the single most important architectural decision in the
whole project — it moves the hardest compliance burden to the people equipped to carry it, and
it means a compromise of your site is not a compromise of your customers’ cards.
Then offer the payment methods people actually use. Cards, the digital wallets, and —
depending on your price point — the buy-now-pay-later options. Every method you don’t offer is a
small percentage of people who don’t complete.
Shipping, which is where the margin goes
Shipping is the most common place a small store quietly loses money.
Decide and then implement: flat rate, free over a threshold, real-time carrier rates, local
pickup, local delivery. Each has different consequences. Flat rate is simple and you’ll lose on
heavy items. Real-time rates are accurate and add a step that costs you some conversions. Free
shipping converts best and has to be priced into the product.
Then the operational layer: packaging, label printing, tracking numbers flowing back to the
customer automatically, and a returns process that exists before you need it.
Shipping thresholds are also one of the few reliable levers on order value. “Free shipping
over $75” when your average order is $55 does real work.
Sales tax
Not glamorous, genuinely important, and the thing most DIY stores get wrong.
Selling from Alabama to customers in other states means sales tax obligations that depend on
where your customers are, not just where you are. Thresholds vary by state and they change.
Your store needs to calculate the right rate at checkout, by destination.
Automated tax calculation services exist and are worth their cost. What you should not do is
guess, apply one rate to everybody, or plan to sort it out at the end of the year.
And this is a genuine “ask your accountant” item, not a web development question. We’ll build
whatever they tell us the rules are. We won’t tell you what they are.
The checkout itself
Everything above gets you a store. This is the part that determines whether it sells.
Carts get abandoned for a small number of repeated reasons, and most of them are fixable:
- A surprise. Shipping or tax appearing at the final step after the
customer has formed a view of the price. Show costs as early as you possibly can. - A forced account. Guest checkout matters. Offer the account after the
purchase, not before it. - Too many fields. Ask for what you need to fulfil the order and nothing
else. Every optional field costs you something. - No visible trust signals. A real address, a phone number, a returns
policy, a visible padlock. Especially for a business the customer hasn’t heard of. - Slowness. Particularly on a phone, particularly on a weak signal.
Then: abandoned cart recovery emails, which are among the highest-return things you can
switch on, and order confirmation and shipping notification emails that reliably arrive — which
means proper email authentication, or your confirmations land in spam and you spend your
afternoons answering “did my order go through?”
Being findable as a store specifically
Product discovery has its own requirements beyond ordinary SEO.
Product structured data describing price, availability, reviews and SKU, so
search results and AI assistants can read your catalogue properly rather than inferring it from
page text.
Unique product descriptions. Manufacturer-supplied copy is on every other
retailer’s site selling the same item. There’s nothing for you to win with it.
Clean URL and category structure, and correct handling of filters and sort
options so you don’t generate thousands of near-duplicate indexable URLs.
A plan for discontinued products. Deleting a page that has earned rankings
and links throws that away. Redirect it to the nearest equivalent.
Speed. Store pages are heavy — many images, more scripts, a cart that can’t
be fully cached. Performance work matters more here than anywhere else on your site, and it
affects both rankings and conversion simultaneously.
Security, raised a level
An e-commerce site holds customer data and sits next to payment flows. That makes it a more
attractive target than a brochure site, and raises the cost of getting it wrong.
Which means the maintenance discipline stops being optional: prompt patching, a web
application firewall, malware scanning, tested backups, two-factor authentication on admin
accounts, and a staging environment so you are never testing a change on the thing currently
taking orders.
That last one deserves emphasis. If your store is live and you have nowhere to test, every
plugin update is a gamble with your checkout.
Should you just use a hosted platform?
Sometimes, yes — and we’ll say so.
If you sell a modest catalogue of simple products, you want it running this month, and you
don’t need the store integrated with anything else, a hosted platform is a perfectly good
answer. It handles payments, tax and hosting for you, and the monthly fee buys real work.
Where a custom build on WordPress and WooCommerce earns its keep is when you need the store
to be part of a larger site rather than a separate thing, when your products or pricing don’t
fit a standard model, when it has to talk to your existing inventory or accounting system, when
content marketing is a significant channel, or when you want to own the whole stack and not be
subject to someone’s pricing changes.
The wrong reason to choose custom is that it sounds more serious. The right reason is a
requirement that the hosted option genuinely can’t meet.
What we’d tell you in a first conversation
Three things, usually.
First, that the build is maybe half the project. Products, photography, shipping decisions
and fulfilment are the other half, and they’re mostly yours. Knowing that at the start is the
difference between a launch and a stall.
Second, that a smaller store launched is worth more than a complete one planned. Thirty
products live this quarter beats three hundred live next year.
Third, that building the store and growing the store are different jobs. We handle the
build and the
hosting. Our sister studio
PG&J handles the ongoing
marketing — product visibility, content, paid, social. If you’re already with us, your site and
hosting stay exactly where they are.
If you’re thinking about selling online, the most useful first step is a conversation about
your operation rather than your website. We’ll tell you honestly whether you need a custom
build or whether we’d put you on something simpler.