Failed Deliveries: The Cost You Never See on an Invoice

A failed delivery is the most expensive ordinary event in e-commerce operations, and it never appears as a line item called "failed delivery". It shows up as a return-leg charge here, a second outbound charge there, twenty minutes of WhatsApp, a discount to save the relationship, and sometimes a refund plus a lost customer. The parcel is fine. The product is fine. The data was wrong, or the customer did not know it was coming. Both are fixed at checkout and at dispatch — never at the door, where it is already too late.

The four causes, in the order they cost you money

Notice that only one of these is about logistics. Three are about information — which is why "switch courier" is almost never the fix and "improve your checkout" almost always is.

Fix 1 — make a good address the easy path at checkout

The default Malaysian address field is a single free-text box, and it produces exactly the addresses you would expect from a single free-text box.

• Separate the fields. Address line 1, address line 2 (unit/floor/block), postcode, city, state — as distinct inputs. A rider reads the second line first; if it does not exist as a field, most customers will not think to type it.
• Derive city and state from the postcode. Malaysian postcodes map cleanly to locality, so autofill both and let the customer correct rather than compose. This removes the single most common data error in Malaysian shipping: a postcode that does not belong to the stated state.
• Make the unit field explicitly required for high-rise addresses. "Level / unit number" with a note that couriers cannot deliver without it is far more effective than a generic "Address line 2".
• Add a short, purposeful instructions field. Not "Notes" — something like "Landmark or delivery instructions (e.g. guardhouse name, blue gate, leave with reception)". The example text is doing the work; a blank "Notes" box gets left blank.
• Validate the phone number's shape and state plainly that the courier will contact this number. A number that fails a basic Malaysian mobile format check should not be accepted silently.

Our checkout and payments guide covers doing all this without adding friction that costs you the sale.

Fix 2 — tell the customer the parcel is coming

This is the cheapest intervention available and the one most sellers skip. One message before dispatch, one after.

Before dispatch: "Hi [name], your order of [item] is packing today and will be sent to [address, last line shown]. Reply YES to confirm, or tell us if the address needs changing." This catches wrong addresses while it is still free to fix them, and it is close to mandatory for cash-on-delivery orders — see the real economics of COD.

After dispatch: tracking number, courier name, and a realistic window. "Realistic" matters more than "fast": a customer told 2–4 working days who receives it on day three is satisfied; a customer told "next day" who receives it on day three opens a complaint.

Automating this is the single best argument for connecting your store to a shipping platform rather than booking parcels by hand — the message goes out because the system sent it, not because someone remembered. Our tutorial on connecting a store to a delivery aggregator walks through the setup.

Fix 3 — plan for the addresses that are genuinely hard

Some Malaysian delivery contexts fail for structural reasons, and the fix is operational, not motivational.

• Guarded condominiums and gated communities. Many require resident authorisation before a rider is let in. Ask for the guardhouse name and a contact number in the instructions field, and set the expectation that the customer should inform security.
• Office addresses. Deliveries after 5pm and on weekends fail. If the address looks commercial, offer a weekday-morning preference.
• Rural and kampung addresses. Where there is no formal street numbering, a landmark is more useful than a line of text. Prompt for it explicitly.
• East Malaysia. Longer transit and thinner coverage in some areas mean the delivery window itself has to be different. Quote the real window from live rates rather than reusing your Peninsular promise — the shipping cost calculator returns times as well as prices by postcode.
• Festive and holiday peaks. Volumes spike, riders are stretched, and recipients travel. Say so on the product page during those periods instead of absorbing the complaints afterwards.

Fix 4 — decide your redelivery and RTS policy before you need it

Write these three rules down now, and put them in your published terms:

1. How many redelivery attempts you will pay for before the parcel returns to you. Most couriers attempt a set number automatically; know yours.
2. Who pays for a re-dispatch after a return to sender caused by a wrong address the customer supplied. Charging for it is legitimate; springing it on someone is not. It must be in the terms the customer accepted — the Consumer Protection (Electronic Trade Transactions) Regulations 2024 require online sellers to disclose their terms and conditions up front.
3. When you switch to a pickup point. For a customer who has failed a delivery once, offering collection from a nearby drop-off point is often better than a third attempt.

Measure it, or it will not improve

Track two numbers monthly. Both are simple and neither needs new software:

• Failed delivery rate = parcels not delivered on the first attempt ÷ parcels dispatched.
• Cost per failure = (return leg + re-dispatch + any refund) ÷ number of failures.

Multiply them by monthly volume and you have the annual cost of doing nothing, which is the number that gets the checkout changes prioritised. Then segment the failures by cause using the table above — most sellers find that one cause accounts for over half, and that fixing it is a single afternoon's work on the checkout form.

Frequently asked questions

Is a failed delivery the courier's fault or mine?
Occasionally the courier's, usually neither — the dominant causes are data and expectation problems that originate on your side of the transaction. Before escalating to a courier, sample ten failures and read the actual delivery remarks. If they say "address incomplete" or "phone unreachable", no courier change will help.

Should I charge customers for a redelivery?
You can, provided the term was disclosed before purchase and you apply it consistently. In practice most sellers absorb the first failure and charge from the second — it costs little, avoids an argument over a small sum, and the second failure is where a genuine pattern shows.

Do pickup points actually reduce failures?
They remove the two biggest causes at once: nobody home, and access restrictions at the delivery address. The trade-off is that the customer has to travel. Offer it as an option rather than a default, and specifically to customers who have failed a delivery before.

How much detail should the delivery instructions field allow?
Enough for a landmark and an access note, and no more — riders read it on a phone screen while working. One or two short lines beats a paragraph. Prompt with an example so people write the useful thing rather than "please deliver fast".

What about deliveries that fail because the customer refuses the parcel?
That is a different problem with a different fix. Refusals are concentrated in cash-on-delivery orders and are driven by buyer's remorse rather than logistics — the pre-dispatch confirmation message above is the main defence, and value caps are the backstop. The COD guide covers the economics in full.

Sources: disclosure of terms, conditions and the full price payable is required under the Consumer Protection (Electronic Trade Transactions) Regulations 2024 (in force 25 December 2024, revoking the 2012 Regulations), administered by the Ministry of Domestic Trade and Cost of Living (KPDN). Courier redelivery-attempt counts, return-to-sender charges and coverage differ by courier and contract — confirm yours against your current rate card and service agreement rather than assuming an industry norm.