The customs-record habit: how reading shipment data changed how I prospect

I want to write about a habit I've fallen into and can't seem to break. Most Sunday mornings, before I'm fully awake and before the rest of my household is, I sit down with coffee and read customs records. Not all of them. The ones for the markets I sell into, the HS codes I care about, the months I haven't covered yet. I scroll, I note, I keep a small text file open. It takes me an hour. Maybe ninety minutes if I find something interesting.

I started doing this in late 2022, almost by accident, and it has changed how I prospect more than any tool, framework, or course I've encountered. I want to explain why, because the practice itself is unglamorous and most of the people I describe it to dismiss it as overkill until they try it.

What's actually in a customs record

For the importing countries that publish or sell the data, a customs record is a row that includes some combination of: the importer of record, the exporter, the HS code of the goods, the country of origin, the country of consignment, a description string (often unhelpful but sometimes very specific), a declared value, a date, sometimes a quantity, sometimes a port. It's the bureaucratic exhaust of a single shipment crossing a single border.

Different countries publish different fields, and at different granularities. US import records have been a public-records gold mine for decades. EU member states are mixed, with some data flowing through Eurostat aggregates and some through commercial providers who clean and resell it. Latin American customs data tends to be detailed and accessible if you know where to look. African customs data is patchier but improving. The reference tables for HS codes themselves come from the World Customs Organization and they're maintained pretty rigorously.

A row from any of these sources, on its own, is a single transaction. A few thousand rows, sorted and filtered, becomes a picture of a market.

What I do with them

I'm looking for three things, almost always. First, who's importing what I sell. Second, who's importing what's adjacent to what I sell, in volumes that suggest they could be a buyer if their current arrangement broke. Third, what the rhythm of those imports looks like over time. Steady monthly draws look different from quarterly bulk landings, which look different from a single test shipment, which looks different from a buyer who imported heavily for two years and then went quiet.

Each of those patterns tells me something. A buyer with a steady monthly rhythm has a working supplier relationship and probably isn't shopping. A buyer with quarterly bulk landings might be price-sensitive and consolidating orders. A buyer who tried once and stopped probably had a bad experience worth understanding. A buyer who went quiet after two strong years is, in my experience, the single most underrated prospect in cross-border B2B, because something changed and nobody is calling them yet.

The shift it produced in how I work

Before I started reading customs records, I prospected the way most people do. I'd build a list from a directory, score it against intuition, draft outreach to the top of the list, and react to whoever replied. The list was made of plausible names. The outreach was made of generic value propositions. The reactions were rare.

After about six months of Sunday-morning reading, my lists stopped being plausible and started being specific. I knew which Czech wholesaler had just lost their Turkish supplier and was sourcing from a single new exporter in Romania, probably as a stopgap. I knew which Belgian industrial buyer had quietly added a second HS code to their import profile, signaling they were building a new product line. I knew which French distributor's volumes had collapsed in Q4 and recovered in Q1, which usually means a one-off inventory problem that's about to repeat.

I started reaching out with sentences I couldn't have written before. "I noticed your imports of [category] from [country] dropped off in October and resumed in February, which usually means a logistics gap. I wanted to introduce myself in case it happens again." Those sentences get replies. They're not clever, they're just true, and the recipient knows immediately that they're talking to someone who's done their homework.

The first three signals I now run before pursuing any buyer come largely out of this work. I wrote up the full breakdown in three signals I look at before I pursue a foreign buyer, and the customs-record habit is where every one of those signals lives.

Why it isn't already standard practice

Two reasons, I think. The first is that the data has historically been hard to access cleanly. The commercial providers who package it well are expensive, and the public sources require patience and a tolerance for messy formats. The second is that reading customs records is not a fast activity. You cannot blast through a country's import data in twenty minutes. You have to sit with it, sort it, notice patterns, follow threads. This does not fit the velocity-obsessed mode that most sales operations encourage.

Both of those reasons are getting weaker. Customs data is cheaper and more accessible every year. And a small but growing number of export practitioners are figuring out that the time spent on this kind of slow reading converts at multiples of the time spent on fast outreach. I made my case for slowing down in the cold-email post and I won't relitigate it here, but the customs-record habit is, in practical terms, what the slow-down looks like in my week.

A note on what customs records can't tell you

They cannot tell you about a buyer's intent. They tell you about their behavior, which is different. A buyer who has imported steadily for five years may be perfectly happy with their supplier and have no reason to talk to you. A buyer who has never imported in your category may be about to start, and you'd never know from the data alone.

They also cannot tell you about price, terms, or the quality of the underlying relationship. A €200,000 declared value tells me a buyer can move volume. It tells me nothing about whether they pay on time, whether they value reliability over price, or whether their procurement manager is a pleasure to work with.

What customs records can do is filter. They turn a list of 500 plausible companies into a list of 40 verified buyers in your category, sorted by recency of activity. That filtering, alone, is worth the Sunday mornings. Everything I do on Monday through Friday is built on top of it, and I don't see myself going back.