Most delivery failures are born at checkout rather than at the carrier. The five recurring causes, the control that prevents each, and what to measure.
Key takeaways
- Address and phone are the number one source of failures, and both are entirely within your control.
- A customer who knows when you are coming is home; one who does not, is not.
- A pickup point eliminates most failure causes at once.
- Every cause has a control that costs hours rather than money.
A failed delivery costs twice: once in the redelivery and once in the service conversation. The figure that surprises businesses is that most failures are not the carrier's fault but are born at checkout - in a missing address, a wrong phone number, and an expectation nobody set.
The five recurring causes
| Cause | Where it originates | The control that prevents it |
|---|---|---|
| Partial or wrong address | A free-text field at checkout | City and street as separate fields plus validation |
| Wrong or landline number | A form that does not check | Israeli mobile format validation |
| Customer not home | No coordination | Advance notice with a time window |
| Building with no access | Missing detail | A delivery notes field: floor, gate code, parking |
| Parcel never collected from a point | No reminder | A reminder on day three before it returns |
Row two is critical in Israel: a mobile number is the channel carrying notifications, courier calls and locker codes. A landline or wrong number disables all three at once.
Why does a free-text address fail?
Because it lets the customer write "next to the corner shop, building C". The courier needs a city, street and number - and when those are not separated, they also do not reach the carrier's system correctly.
The control is structural: city from a list, street, house number, and then a free-text notes field. That notes field matters and should not be removed - it is where the customer writes "third floor, no lift" or "gate code 4417", which is exactly the information that prevents a second trip.
Also allow a delivery address different from the billing address. A customer ordering from work who wants it delivered at home - or the reverse - will enter the wrong address when the two cannot be separated, and will discover it only when the parcel reaches the wrong place.
What to send the customer before delivery
Three messages cover nearly every case: confirmation the parcel has shipped with a time range, a message on the delivery day, and a message when the courier is on the way. Each one reduces the chance the customer is out, and together they also reduce the number of enquiries reaching you.
What matters is that the message allows an action - change the date, redirect to a neighbour, or switch to a pickup point. A message that only says "your parcel is on its way" passes information without giving the customer a way to prevent a failure.
The controls that cost least
- Mobile number format validation at checkout.
- Separate address fields for city, street and number.
- A delivery notes field with examples inside it.
- A proactive message on the dispatch day.
- A reminder before a parcel at a point is returned.
- Flagging repeat failures - a customer who failed before is marked.
Point six is the least common and the most effective: the list of previously failed addresses is very short, and once flagged you can call those specific customers in advance rather than everyone. All six together are a day or two of work in the checkout and the process, and they move the number more than changing carrier does.
What to measure
First-attempt delivery rate is the single metric that summarises everything, and it is worth measuring by area and by service type. After it: failure reasons from a short list, and how many redeliveries per month.
Segmenting by area usually reveals the problem is concentrated somewhere specific - one region, an office building, or particular hours. That is actionable: change carrier for that area, offer a pickup point, or change the delivery window.
What to do when a parcel comes back
The flow that works: identify the cause, contact the customer the same day, and offer two options - redelivery at an agreed time, or a pickup point. What does not work is waiting for the customer to get in touch, because some of them simply ask to cancel.
Decide in advance who pays for the redelivery when the failure came from a wrong address the customer supplied. There is no single right answer, but a consistent policy prevents an argument in every case - and above all prevents reps deciding differently from one another.
A pickup point as a fix, not as a saving
Most discussion of pickup points is about cost, but the larger benefit is that they eliminate three of the five failure causes at once: no dependence on when the customer is home, no building access problem, and no failed delivery attempt. What remains is one new cause - a parcel never collected - and that has a simple fix in the form of a reminder.
Which makes it worth surfacing a pickup point at checkout specifically for customers who have already failed, rather than only as a general option. How to present it without eroding margin is covered in shipping rules at checkout.
Where the line runs between your fault and theirs
This is worth answering internally before it comes up with the carrier. A failure from a partial address, a wrong phone number or no coordination is yours - and as noted, that is most cases. A failure from missing a promised time, a lost parcel, or a delivery attempt that never actually happened is theirs.
The distinction matters because it determines what is worth discussing with the carrier at all. A business arriving with separated data - "40 failures this month, 28 address and 12 non-arrival" - gets a completely different response from one complaining about poor service. More importantly: it knows the first 28 are its own, and that they are also the easier half to fix.
Sources
Frequently asked questions
What failure rate is considered reasonable?
There is no published benchmark, so the useful measure is your own trend: where you started, and where you are after adding controls. A consistent decline over three months says the controls work, regardless of how it compares to anyone else.
Should we call every customer before dispatch?
Not everyone - that is expensive and unnecessary. Call only the risky shipments: a new address, a high-value order, a customer who failed before, or an area with a failure history. That removes most failures at a cost of a few calls a week.
What about customers who repeatedly fail?
Flag them and offer a pickup point as the default on their next order. It is a solution rather than a punishment: a customer who was out twice probably genuinely is out, and a pickup point solves it for both sides.
Can you know in advance that an address is problematic?
Partly. Structural validation catches invalid addresses, but it cannot catch a missing floor or a gate code. That is why the notes field matters so much: it is the only place the customer supplies information no automated validation knows to ask for.
Keep reading
Related service
Business Automation
I build custom automations that remove repetitive work end to end.
About the author
Yehonatan Saadia
Freelance automation, web & MVP developer
I'm Yehonatan Saadia, a senior developer who builds business automation, custom websites, and MVPs for small and mid-sized companies across the US, Europe, and Israel. These guides come from real client work, not theory.
Work with meHave a project like this?
Tell me what you're trying to automate or build and I'll tell you the fastest reliable way to ship it.