# When a shop should sell online and when it should not

> We build both kinds of site: full ecommerce, and sites where the money is made in person and the site's job is to get somebody through a door or onto a call.

Canonical URL: https://vancouverwebsitedesign.com/insights/ecommerce-or-enquiry/

---

Retail When a shop should sell online and when it should not
We build both kinds of site: full ecommerce, and sites where the money is made in person and the site's job is to get somebody through a door or onto a call.
6 min Reviewed July 2026 Dave Curtin / Founder and creative director
We build both kinds of site: full ecommerce, and sites where the money is made in person and the site's job is to get somebody through a door or onto a call. Which one suits you comes down to staff time and margin per order more than to software. What follows is the arithmetic we use.
01 The number that matters is your cost per order
A webshop build with us usually lands between $18,000 and $35,000, depending on how many products you carry, how complicated the shipping gets and whether your stock lives in a till system that has to talk to the site. That is the visible cost, and the one that decides whether the shop works sits underneath it: what a single order costs you to fulfil.
Add up the packaging, the courier, the card fees and the twenty minutes somebody spends picking, packing and printing a label, then the returns, the emails asking where the parcel is and the one in thirty that arrives broken. For plenty of businesses that comes to $12 to $25 an order, before anyone has been paid a wage.
On a $35 basket at forty percent margin you make fourteen dollars, and most of that goes on getting the thing out the door. That is back of envelope arithmetic, and it will tell you more than any agency proposal will. If average order value times margin comes in under about $30, the shop needs volume or a bigger basket. Know which one you are relying on.
02 Somebody has to work in the shop
Owners plan for the build cost, then meet the staffing afterwards, because an online shop needs somebody in it on the evenings and Sundays when the rest of the business is shut. Stock counts drift out of true. Somebody has to answer the customer who ordered the wrong size, chase the courier who lost a pallet, refund a double charge and reshoot the photos when the packaging changes.
For a small operation that is five to fifteen hours a week once you are past the first month. It comes in fragments across the day rather than one manageable block, mostly landing on whoever is nearest the laptop, which tends to be the owner or the single marketing person. It comes out of whatever they were doing before.
Ask yourself who that person is by name, and check what they would drop to make room. If nobody comes to mind, what you are buying is a website that takes money and then hands you a problem every morning. We have talked a few clients out of ecommerce on that point alone.
03 When a shop does earn its keep
Some businesses should be selling online, and the pattern is fairly consistent: products that post well, meaning light, sturdy, not perishable and not easily broken. People buy again, so the cost of winning a customer is spread over four or five orders. Margin sits above about fifty percent, and the basket can be pushed up with bundles, multipacks or a subscription. Candles and coffee sit on that list, wedding cakes do not.
Food and drink brands sit awkwardly across that list, which is why they are worth thinking about slowly. We worked with Clonakilty Food Co, an Irish food company whose products are mainly bought in supermarkets, and for a business like that the website carries the brand, supports the trade relationship and answers the shopper holding the pack. A direct shop is a separate decision with its own economics.
If you are in that position, run a test before you commission anything: take pre-orders through a form and a payment link for six weeks, pack them yourself and time it. You will learn more about whether you want this business than any amount of planning will. Our notes on how we approach retail and ecommerce projects cover the platform side.
A hybrid suits a lot of businesses: a strong destination or enquiry site, plus a small shop selling six or eight things that post easily. Gift vouchers, a hamper, a signature product, tickets. Keep it small, because a shop with three hundred SKUs and nobody assigned to it will decay inside a year. Most of the ecommerce work we do in year two is expansion for clients who started that way.
04 When the site's job is to fill a room
Plenty of retail businesses make all their money in person, and a shop on the site would be a distraction. Crescent Shopping Centre is one of those. The centre sells you nothing online, and its site exists to tell you what is open, what is on, where to park and which brands are in, and to make going there on a wet Tuesday feel like a reasonable plan.
That is a different design problem and a demanding one: opening hours have to be right on Christmas Eve, events need to be easy for a marketing person to publish at four on a Friday, and everything has to work on a phone in a car park with one bar of signal. The build cost is usually lower, $15,000 to $28,000, with no fulfilment tail afterwards.
Judge these sites on visits, bookings, direction requests and calls rather than transactions. If you run a destination business, a venue, a food hall, a guesthouse, a festival, the honest measure is whether more of the right people showed up, and that is harder to report than a sales figure but closer to what you care about.
05 Selling through a conversation instead
For anything sold through a conversation, a considered enquiry route usually beats a cart. That covers most B2B work, most services and any product where the price depends on quantity, spec or delivery date. Forcing those into a checkout gives you a shop that quotes the wrong price to the wrong person, and somebody has to ring them back and unpick it.
A good enquiry route is a set of pages that answer the questions a buyer has before they will pick up the phone: what you make, who you have made it for, what it roughly costs, what the lead time is and what happens after they get in touch. Then a short form and a commitment to reply inside a working day.
Those pages are usually the ones nobody wants to write, so we run the writing as a real part of the project rather than leaving it with the client. Brainstorm, our Irish practice, established 2010, has written up how that content work runs alongside a build . Expect it to add two or three weeks, and to change your enquiry quality more than the design will.
06 How to decide this week
Work out your average order value and gross margin, then your all in cost to fulfil one order including labour. If the gap is thin, the shop needs volume you may not have. Look at your last hundred enquiries and count how many could have been a purchase with nobody speaking to anybody, and if that is fewer than half, your money is in the enquiry route.
If you want a sanity check before you commit, Dave Curtin will go through the numbers with you on a call and tell you if the answer is no. That happens often enough to be worth saying. Projects with us run 10 to 16 weeks and the first two are largely spent on this question, so it is cheaper answered before anyone signs.
For a good number of established businesses the right answer is a site that makes going to see you the easy option, with the shop left for later. That is a duller recommendation than a full ecommerce build, and it is the one we would give you across a table. Year two is a fine time to start selling, once there is somebody to do the packing.
Continue the decision Move from guidance to relevant evidence.
Related case Clonakilty Food Co. Related engagement Website Design
What decision is your team trying to make?
Use the planner to share the condition, not a polished brief.
Start the three-minute planner Explore all guides