The distributor who doesn't reorder. Nobody will come and tell you.
The first shipment is a conquest, the second is the business. Between the two lies the most neglected moment in all of export, because it is the only one that produces no email to read.
A customer who stops buying doesn't write to say so. There is no complaint, no cancellation, nothing to put in an inbox. There is only an order that doesn't arrive, and an order that doesn't arrive makes no noise.
You need a single number, and you already have it: for each customer, the average interval between one order and the next. Whoever has gone past one and a half times their own interval should be called today. It is calculated with a spreadsheet, in an afternoon, and needs no software.
The threshold goes per customer, never in general. Whoever ordered every two months should be called on day ninety. For whoever ordered once a year, day ninety means nothing.
This piece is part of the guide to export operations for the small food producer and covers the last of its eight steps: the one that comes after everything else has worked.
Why this loss is invisible by construction
All the other export problems announce themselves. An incomplete sheet produces a stalled enquiry, a wrong document produces goods held up, a price out of line produces a no. They are all events: they happen, and someone notices.
The reorder that doesn't come is the only problem that shows up as an absence. And absences appear in no list, because no system in the world generates a notification for something that didn't happen.
Then there is a second, more insidious reason, which is that in export silence is normal. Weeks pass between one contact and the next with a foreign distributor even when the relationship is excellent. So three months without hearing from each other set off no mental alarm: they look like ordinary functioning, until someone looks at the calendar and discovers the last order dates back to November.
A customer lost through inattention is indistinguishable, from the inside, from a customer who is simply taking their time.
The number, and how to calculate it
You need one thing only, and you already have it in your invoices: order date and customer name, for the last two years. From there, for each customer:
- Line up the dates of their orders, from oldest to most recent.
- Calculate how many days pass between one order and the next.
- Take the average: it is their typical interval.
- Count how many days have passed from the last order to today.
- Divide the second number by the first. Above 1.5 the customer is late. Above 2 the relationship has probably already ended, and you don't know it.
An example, built to explain the calculation and not taken from a real customer: a distributor ordered on 10 January, 15 March, 20 May. The intervals are 64 and 66 days, so their average is 65. If today is 20 August, 92 days have passed since the last order: 92 divided by 65 gives 1.4. It isn't an alarm yet, but that customer goes on this week's list, not next quarter's.
With twenty or thirty customers all this is an afternoon's work on a spreadsheet. And doing it by hand the first time has an effect no dashboard produces: you notice things about your customers you didn't know, such as two of those you consider solid not having ordered for eight months.
The three causes, and how to tell them apart
When the number flags someone, the question becomes why. The real causes are almost always three, they call for different responses, and they are distinguished with different questions.
| Cause | How to recognise it | What needs doing |
|---|---|---|
| They didn't sellthey still have stock | They reply willingly but put off the order. If you ask, they can tell you how much they have left. |
The problem isn't the relationship but shelf rotation. You need to help them sell: materials, arguments, formats, an action at the point of sale. |
| They found an alternativeprice or service | Short, vague replies, or silence after a discount request that went unanswered. |
You need to understand what you lost, and ask directly. An uncomfortable answer is worth more than a customer who fades away without explanation. |
| Something changed on their sideperson or strategy | The person you dealt with no longer replies, or someone else replies who doesn't know you. |
You haven't lost anything: you dropped out of the loop without anyone deciding it. You need to reintroduce yourself, not win them back. |
The third is the most frequent and the most underrated, and it is also the one recovered with the least effort. Nobody chose against you: a purchasing manager simply changed, and your product dropped off the mental list of someone who didn't even know they had one.
The message that reopens the conversation
Two sure ways to waste the attempt: the reminder («I haven't received an order from you for a while»), which puts the other person in your debt, and the empty message («how are things?»), which gives no reason to reply.
What works brings three things: a concrete piece of news, a question useful to them, and an easy way out.
Subject: [product] - new format available / quick question Dear [name], we have just released [new format / new harvest / a batch of X available from week Y], and I thought of you because [specific reason connected to their market or channel]. Before sending details: how has [product] been rotating on your side? If it has been slower than expected, I would rather help you move the current stock than push a new order. If this is not the right moment, just let me know and I will get back to you in [month]. Kind regards, [name] - [company]
The third paragraph is the one that makes the difference, and it is the most counter-intuitive: making it easy to say no increases replies. Whoever hasn't ordered for months often doesn't reply out of embarrassment, not lack of interest. A message that explicitly allows for «not now» removes that embarrassment, and together with the embarrassment it removes the silence.
The 1.5 times interval threshold doesn't come from a study: it is a rule of thumb, and we state it as such. It serves to turn a feeling into a list, and it should be calibrated to your seasonality: whoever sells a product tied to holidays has intervals that don't average well, and for those customers what counts is the comparison with the same period of the previous year.
We don't cite a statistic here on the reorder rates of small Italian food exporters because we found no reliable public source on this. The number that counts, after all, is yours, and you calculate it in an afternoon.
When the spreadsheet stops being enough
The calculation is easy. What is hard is redoing it every week, for ever, while producing. That is what breaks, not the maths.
With twenty customers it holds. At forty, with orders arriving from three different channels and different seasonality per product, the sums get done once in January, something useful is discovered, and then they aren't redone until the following year. Meanwhile customers fade out one at a time, each in silence.
That is the point where the task should be taken out of someone's hands and put in a place that watches it on its own: the list of who is idle, for how long, and whose turn it is to act. With a constraint that matters more than the technology, written in Ethics: the system prepares and flags, but no message reaches a distributor without a person having read and approved it. A reorder is won back with a relationship, and relationships aren't automated.
Questions and answers
How can I tell that a distributor has stopped buying?
By looking at the dates, because nobody will tell you. For each customer you calculate the average interval between one order and the next, then compare it with how many days have passed since the last one. Whoever has gone past one and a half times their own interval should be called.
All you need is the date and customer of every order over the last two years. It is an afternoon's work with a spreadsheet, and it needs no software.
After how long should a foreign customer be considered lost?
There is no number valid for everyone, and using a generic one is the most common mistake. The reference is the historical interval of that customer: past one and a half times their average it is a case to open, at double the relationship has probably already ended.
In practice: whoever ordered every sixty days should be called on day ninety. For whoever ordered once a year, day ninety means nothing.
Why does a foreign distributor stop reordering?
Three causes cover almost every case. They didn't sell and still have stock: the problem is shelf rotation, and you need to help them sell. They found an alternative, on price or service: you need to understand what you lost, by asking. Something changed on their side, a person or a strategy: you dropped out of the loop without anyone deciding it, and you need to reintroduce yourself.
The third is the most frequent and the easiest to recover.
What do I write to a distributor who hasn't ordered for months?
Not a reminder and not a «how are things». You need a message with three things: a concrete piece of news (new format, new vintage, an available batch), a question useful to them (how the product is rotating), and an easy way out if it isn't the moment.
The last is the most counter-intuitive and the most effective: whoever hasn't ordered for months often stays silent out of embarrassment, not lack of interest. Making the no easy removes the embarrassment, and with it the silence.
Do you need software to monitor reorders?
No, and the first time it is better without. With twenty or thirty customers the sums are done with a spreadsheet, and doing them by hand teaches you things about your customers that a dashboard wouldn't.
Software is needed when the calculation has to be redone every week on dozens of customers, that is, when the problem is no longer calculating but remembering to calculate. Whoever buys the tool before having done the sums once usually buys a dashboard nobody then looks at.
Notes on sources
- This page doesn't cite sector statistics on the reorder rates of small Italian food exporters because we didn't find a reliable public source on that specific figure. When the source doesn't hold up, the number isn't published.
- The 1.5 times historical interval threshold is a rule of thumb stated as such, not a research finding. It should be calibrated to the product's seasonality.
- The numerical example in the text is built to explain the calculation and doesn't describe a real customer.
The dates nobody looks at can be made to surface on their own.
Do the sums by hand the first time: it teaches more than any tool. With Itria we start from the outside to build custom digital systems for exporters. For you that translates into more requests, fewer losses and less manual work. Write us a line about what weighs on you. We take the first step: what a buyer sees when they look you up, and what we found there. Even if we don't end up working together.