Standard advice on where to sell online is written for people who hand a box to a carrier. Spread across marketplaces, let the platform bring the traffic, pay the commission, and treat fulfillment as a fixed cost someone else absorbs. It is reasonable advice, and almost none of it holds once you are the one driving.
The reason is simple arithmetic. A marketplace commission of 15% is painful but survivable when fulfillment is outsourced and the channel delivers customers you could not otherwise reach. The same 15% sits on top of a delivery cost you are personally absorbing in fuel, time and vehicle wear, and the combination can take an order past break-even without anything looking obviously wrong.
So the question of how to sell online when you deliver your own orders is really a question about net contribution per channel, not reach. This guide works through the channels one at a time (your own store, marketplaces, social, and direct ordering) with the fee structure and the operational fit for each. If you are earlier than that and still setting the shape of the operation, start with how to start an ecommerce business with local delivery, and if you need the channel mix written into a forecast, what belongs in an ecommerce business plan covers how to model it.
Key Takeaways
- On a $50 order, all-in fees run about 3.2% through your own store with native payments, against 15–20% on Amazon, 9.5–10% on Etsy and 13.25% on eBay (Webgility).
- When you deliver the order yourself, the commission is additive to a fulfillment cost you already carry, so channel choice is a margin decision before it is a marketing one.
- Marketplace fulfillment rules assume parcel carriers. Many cannot represent a delivery window, a service radius, or a same-day cutoff.
- Treat third-party channels as discovery and your own store as the destination; the goal of a marketplace order is the second order, placed directly.
- Your checkout needs three things a standard one lacks: address validation against your service area, a delivery slot, and an order cutoff time.
Lower your delivery costs by 23%
How we reduce costs:
- No delivery vehicle expenses
- Optimized local routes
- Pay-per-delivery model
- Average 23% delivery cost reduction
Why delivering your own orders changes the channel maths
Three things shift when fulfillment is yours.
- Commission compounds with a cost you already bear. Outsourced fulfillment means a marketplace fee buys you access to demand while someone else handles the physical work. Self-delivery means you pay the fee and do the work, so the fee has to be justified by customer acquisition alone.
- Geography becomes a hard constraint. You can only serve addresses inside your radius. Channels that cannot express that constraint will happily take orders you cannot fulfill, and the cancellation falls on your account rating, not theirs.
- Timing becomes part of the product. Roughly 80% of consumers expect same-day delivery to at least be an option, and 70% say they are more likely to buy online when it is available (Local Express). That is an advantage you hold over national sellers, but only on a channel that can display a delivery window and enforce a cutoff.
The practical consequence is that the channel which brings the most traffic is often not the channel worth building on.
Comparing online selling channels for self-delivered orders
| Channel | Typical fees | Controls your service area? | Best used for |
|---|---|---|---|
| Your own store | ~2.9% + $0.30 processing, plus $29–$39/mo | Yes, by radius or postcode | Repeat customers and all local delivery |
| Amazon | 8–20% referral, most categories ~15% | Poorly | Shelf-stable products only |
| Etsy | 6.5% plus listing and processing fees | Poorly | Giftable, non-perishable, nationally shippable |
| eBay | ~13.25% all-in | Poorly | Clearance and surplus inventory |
| Instagram and Facebook | Low or none for discovery | No, handled off-platform | Discovery, then redirect to your store |
| Local ordering apps | 15–30% | Yes, built around it | Trial demand in a new area |
| Direct phone, text, email | None | Yes, manually | Wholesale and standing orders |
Fee figures are indicative and vary by category and plan. The column that matters most for a self-delivered business is the middle one.
Your own online store is where self-delivery belongs
For anything you deliver yourself, your own storefront should be the primary channel, and the reasons go well beyond the fee saving.
It is the only channel where you control the service area. You can refuse out-of-range addresses at checkout rather than discovering the problem after payment. You can set a delivery fee that reflects your actual cost per stop, show a delivery window, and enforce a cutoff time that protects the route.
It is also the only channel where you own the customer relationship. The email address, the order history, the ability to tell them you are running a Valentine’s pre-order: all of it is yours. Marketplaces hold that relationship deliberately, and no amount of volume converts into it.
The trade-off is honest: nobody arrives at a new storefront on their own. Your own store solves margin and operations; it does not solve discovery. That is what the other channels are for.
Selling on marketplaces when you deliver your own orders
The fee is the obvious objection and the smaller one. The larger problem is that marketplace fulfillment systems were designed around parcel carriers and tracking numbers.
Concretely, most major marketplaces expect you to ship with a trackable carrier service and mark the order shipped with a tracking number. A self-delivered order has no tracking number. Workarounds exist, but they tend to involve marking orders shipped with placeholder data, which risks account standing, or absorbing late-shipment metrics you cannot actually control.
They also struggle to represent a service radius. A listing visible nationally will attract orders from addresses you cannot reach, and cancelling them damages your seller metrics.
Where marketplaces do earn a place:
- Shelf-stable products that survive the mail. Roasted coffee, preserves, dry mixes, candles. Run these as a separate parcel-shipped line, fully distinct from your local delivery catalog.
- Clearing surplus. Inventory that would otherwise be written off is worth selling at a 13% commission.
- Brand discovery in categories where buyers browse the marketplace first. This does happen, but measure it rather than assume it.
What rarely works is listing your perishable, locally delivered catalog on a national marketplace and hoping the geography sorts itself out. It does not.
Social commerce and local ordering platforms
Social is the strongest discovery channel most local sellers have, and it is strongest when you do not try to make it a checkout.
Instagram and Facebook are where people see the product and decide they want it. The conversion can happen on your own store, and for a locally delivered business it generally should, because that is the only place the delivery area and cutoff time are enforceable. Treat the social post as the top of the funnel and link to the product page.
Assume that journey happens on a phone. Current smartphone usage statistics put the overwhelming majority of social browsing on mobile, which means your product page and checkout are being judged on a small screen with one thumb. A delivery slot picker that only works on desktop is, in practice, a checkout that does not work.
Local ordering apps and delivery platforms are a different proposition. They are built around exactly your constraints: service radius, delivery windows, same-day. That is why they work operationally. The cost is a 15–30% commission and a customer who belongs to the platform rather than to you.
They are worth using for a specific job: testing demand in a neighbourhood you have not served before, or filling capacity on slow days. They are a poor permanent home for your core business, because the commission never falls and the customer list never becomes yours.
Direct ordering still carries wholesale
For restaurant, retail and corporate accounts, the channel that works is often not a storefront at all. Standing orders by email, a text message the night before, a shared spreadsheet, a recurring invoice.
This looks unsophisticated and it is extremely profitable: no commission, no payment processing on invoiced terms, predictable volume, and route density that retail orders cannot match. A single wholesale stop can be worth a dozen residential ones to your cost per delivery.
The upgrade worth making is not to force these customers onto a public storefront but to give them a private one: a login with their negotiated pricing and a standing order they can adjust. Most ecommerce platforms support this through a wholesale channel or a password-protected catalog.
What your checkout needs that a standard one does not
Three features, and most default configurations have none of them.
- Address validation against your service area. The checkout should check the delivery address against your zones and either quote a fee, offer pickup, or decline. Taking payment for an address you cannot serve creates a refund and an unhappy customer out of what should have been a clean no.
- A delivery date and time slot. This is what lets you batch orders into routes instead of reacting to them individually, and it sets the customer’s expectation precisely rather than leaving it to hope.
- An order cutoff time. Without one, a 4pm order silently commits you to a same-day run. With one, the customer sees “order by 2pm for delivery today” and self-selects into tomorrow’s route.
A fourth thing matters just as much, and it is about presentation rather than function: show the delivery fee early. Extra costs appearing at checkout drive 39% of cart abandonment (Baymard Institute), and the same fee shown on the product page reads as information rather than a surprise.
Order notifications do the customer service for you
Self-delivered orders generate a specific kind of anxiety, because there is no carrier tracking page to check. The substitute is a short sequence of messages:
- Order confirmed, with the delivery date and window restated.
- Out for delivery, sent when the route leaves, ideally with a rough window.
- Delivered, with a photo of where it was left.
Most platforms send the first automatically and the second with a basic app. The third is the one that matters most, because for perishable goods a delivery left in the wrong place is a total loss, and a photo turns a dispute into a conversation.
Operators who put this sequence in place generally find inbound “where is my order” contact falls to almost nothing, which saves real time on a busy delivery day.
How to move marketplace customers to your own store
If you use third-party channels for discovery, the point is the second order. A few things that work within platform rules:
- Package inserts. A card in the box with your store URL and a first-order discount. Permitted on most platforms as long as it does not disparage the marketplace or solicit reviews.
- Branded packaging. The customer needs to remember who sold it to them, which is harder than it sounds when the marketplace brand dominates the experience.
- A reason to come direct. Local delivery is itself the reason, and it is one the marketplace cannot match. Say so plainly: faster, fresher, and available only through you.
- Follow-up that reflects what they bought. Once an order is in your own system you know the address, the products and the delivery window, and personalization statistics consistently show tailored follow-up outperforming generic campaigns. A marketplace will never hand you that data.
- Local search presence. People who discovered you on a marketplace often search your name later. Make sure what they find is your store.
The metric to watch is not marketplace revenue. It is how many marketplace customers placed a second order directly, and what that cost you per customer acquired.
Frequently asked questions
Should I sell on marketplaces if I deliver my own orders?
Only for products that survive the mail, or to clear surplus stock. Marketplace fulfillment systems expect a carrier tracking number and cannot express a delivery radius, so your locally delivered catalog belongs on your own store.
What does it cost to sell on my own store versus a marketplace?
On a $50 order, roughly 3.2% all-in through your own store with native payments, against 15–20% on Amazon, 9.5–10% on Etsy and 13.25% on eBay (Webgility). Your own store also carries a monthly subscription, which is fixed rather than per-order.
Can I sell on Instagram and deliver locally?
Use Instagram for discovery and send the order to your own checkout, where the service area and cutoff time can be enforced. In-app checkout does not handle a delivery radius well.
How do I stop customers ordering from outside my delivery area?
Set delivery zones by postcode or radius in your platform so the checkout validates the address before payment, and offer local pickup as the fallback for near-miss addresses.
Do I need a tracking number for self-delivered orders?
Not on your own store, where a delivered notification with a photo serves the same purpose. On marketplaces you usually do, which is one of the main reasons self-delivered orders fit poorly there.
Choosing where to sell
Run every channel through the same test: what is left per order after commission, payment processing, and the cost of getting it to the door. The figure to hold onto is net contribution per order, not revenue and not order count, because it is the only one that tells you whether a channel is building the business or subsidising a platform.
For most operators delivering their own goods, the answer settles into the same shape. Own store as the home, social as the way people find it, marketplaces for the shelf-stable subset, and direct relationships carrying the wholesale volume that makes the routes worth running.
