Trade Insights

How is customs data used? The six-step process from looking up the HS code to sending the first letter

Beginners using customs data often face the problems of lack of screening standards and poor conversion effect. This article breaks down the six-step complete practical process from HS code verification to sending the first development letter, and provides feasible judgment criteria and common misconceptions for each step. Aline GPT integrates multi-source data, enabling one-stop buyer screening, contact completion, and email verification, significantly lowering the barriers to customer acquisition for novices.

Author: AlineGPT Team36 views
Share Article
How is customs data used? The six-step process from looking up the HS code to sending the first letter

Quick conclusion

Customs data development clients essentially use customs declaration records to narrow down the candidate range, and then use publicly available information to supplement the judgment. The complete process consists of six steps, and after completing one round, you can usually receive the first valid response in one to four weeks.

  1. Check codeConfirm the HS code and include three to five English keywords, covering three levels: product category, specifications, and intended use.

  2. fixed marketFirst, focus on a country and take records from the past 12 to 24 months. Don't start by searching globally.

  3. Screening listFirst, select the logistics provider, and then divide them into three tiers based on procurement frequency, Chinese procurement history, product category matching, and number of suppliers.

  4. Look at the patternFind the switching points between the procurement cycle and suppliers, and set the time window for contact.

  5. Additional contact personComplete the three layers of roles: procurement, technology, and decision-making, and then verify the email address in the five-layer order.

  6. Send the first letterStart with a three-stage structure, and follow up with a different channel within 48 hours.

If the three premises are misjudged, all six steps will go astray: the consignee is not the same as the buyer. The common saying in the industry is that more than 30% of the consignees are freight forwarders, customs brokers and trade intermediaries. Having records does not mean there is demand; the data is usually lagging by one to three months. Whether or not you can contact them is more important than the size of the purchase amount in determining whether or not to invest time.

The usage of customs data can be explained in one sentence: First, confirm what the goods you are selling are called on the customs declaration form, then check who bought them, how much they bought, and from whom they bought them. Finally, find the person and get to the point.

What really stumps beginners are the middle steps. When I opened the system, thousands of records were laid out in front of me, and I didn't know how to sort them. I selected the dozens of largest companies based on the amount, but I received no response to the letters I sent. Occasionally, someone would reply, asking questions about specifications that they couldn't answer. Most people give up after the third week, concluding that customs data is useless.

The problem is often not the data. The following section breaks down the process from checking the code to sending the first letter into six steps, each providing verifiable actions and judgment criteria.


I. What is customs data? Which fields should beginners understand first?

Customs data is a customs declaration record generated by customs authorities in various countries during customs clearance of goods, which is anonymized and provided to the public. ItIt records transactions that have already occurred.It's not an intention, nor is it a contact information.

There are only eight fields that are really useful for finding customers.Importers and exportersExplain the two parties to the transaction; the former is the object of verification, while the latter can be used to backtrack on the customer structure of peers.Product DescriptionIt is the product name used during customs declaration. It is usually written simply and needs to be compared with your own product item by item.HS codes are category coordinates.The first six are globally unified, while the last few are different from each country;Quantity and valueIt is a proxy indicator of scale, used to determine whether the magnitude matches;Country of originExplain where this batch of goods is coming from, which is used to determine whether the other party has already purchased it in China;Customs declaration dateDetermine whether this record still has reference value.

Understanding the relationships between fields is more important than remembering field names. Only when these five dimensions are combined can a judgable purchasing behavior be constituted: who bought it, what they bought, how much they bought, from whom they bought it, and when they bought it. If any one item is missing, the judgment will be distorted.

The easiest thing for beginners to overlook is the product description. The same code may cover parts, raw materials, and finished products simultaneously. Looking only at the code without looking at the description will add unrelated buyers to the list. Conversely, when the description is vague, it should be considered in conjunction with quantity, weight, and supplier structure.

II. Why should beginners learn this step first?

Compared to platforms, trade shows, and search advertising, customs data offers three practical benefits to teams with limited budgets, but each benefit has its limits; please understand them before using them.

  • Customers are "transactional", not "potentially interested":The record indicates that the company has indeed imported this category, which is more reliable than guessing from the exhibition catalog. The boundary is that there are records showing that it has purchased, but this does not mean that it is now willing to change suppliers. The gap between the two is what needs to be bridged in the next four steps.

  • Intelligence is structured, not fragmented:The historical records of the same buyer can reveal changes in the procurement cycle, supplier country structure, and number of suppliers. This information determines when to speak and what to use to persuade. The boundary is that it reflects the past, usually with a lag of one to three months. It is acceptable to use it to judge the cycle, but not to judge the current intention.

  • Costs are controllable, and there is no need for bidding or booths:Data tools are mostly subscription-based, and the cost of trial and error is much lower than advertising. The boundary is: cheap does not equal effectiveness, the list cannot be completely eliminated, and the money saved will be returned over time.

Another easily overlooked fact is that only forty or fifty countries truly open up enterprise-level details, including the United States, India, Vietnam, Mexico, and several South American countries. Most EU countries, Japan, South Korea, and Australia do not disclose the names of importers. Confirm this before choosing a market; otherwise, if you can't find it, it will be mistakenly judged as having no demand.

3. Six practical steps, from opening the system to sending the first letter

Step 1: Determine the HS code and product keywords.

The code is the key to the retrieval. The most common mistake beginners make is searching for products by common names, such as directly searching for "screws," and the results will be a mixture of fasteners, hardware, and mechanical parts, which is extremely noisy.

The correct approach involves three steps: First, find the six-digit code using the Chinese product name in the customs code lookup tool; Prepare three to five English keywords, covering three levels: category keywords, specification keywords, and purpose keywords; Finally, a combination of coding and keywords was used for retrieval, and the two were cross-verified. For products with a large span, check them in batches. Do not mix screws and nuts in the same table at once.

Another convenient method is to check with competitors. Find two or three well-known competitors and see how they write the product names and use codes on the bill of lading. It's faster than flipping through the coding sheet yourself. Note that there are subtle differences between the eight-digit and ten-digit codes in different countries; checking only six digits may miss a batch of buyers.

The second step is to select the target market and retrieve import records.

Don't start by searching globally. Select one or two countries where data is available and there are already records of similar products, and streamline and expand the process.

The procedure is as follows: enter a code or keyword, select a country, and choose a time range of nearly twelve to twenty-four months. First, check the total number of results, and then check whether the categories in the first twenty items match. A market with only a few dozen entries is either small in scale or lacks complete coding. These two reasons need to be distinguished.

There is a simple way to judge whether a country is worth investing in: use the same set of conditions to examine two countries separately, and compare and record the quantity and category concentration. The large number but scattered categories indicate that the search criteria are too broad; The small quantity but concentrated product categories may actually represent a precise market segment.

Aline GPT: The data retrieval scope at this step is broader than a single customs database: in addition to customs data, it also connects to global enterprise databases, search engines, maps, and exhibition records, allowing for cross-referencing of several sources for the same target market. For teams with clear product categories but unsure which market is worth investing in first, cross-referencing can save them a round of trial and error. If you're just a veteran in a mature market, a single source is actually enough.

The third step is to screen out the list worth developing.

You can't follow all thousands of records. Do one thing before screening:Exclude logistics providers. In the consignee list, freight forwarders, customs brokers, and trade intermediaries account for a significant proportion.Company names containing "Logistics," "Freight," "Forwarding," "Shipping," and "Transport" are mostly Logistics providers. For items containing Import & Export and Trading, please question them first; In European and Latin American markets, local language exclusions should also be added: German Spedition, French transitaire, and Spanish agencia aduanal.

After elimination, it is judged according to four criteria: whether there are more than three records of the same category in the past twelve months; Whether it has ever been purchased from China; The degree of overlap between the product description and its own product line; Are there more than two suppliers? Three or more of the four items are developable; The two items are now under observation; If none of them are successful, give up.

Here's a reminder: sorting the top few dozen companies by amount is the easiest way to go astray. Under the same conditions, customers whose number of priority suppliers is changing and who have recently switched have a higher probability of entering the market.

The judgment results are directly divided into three levels, eliminating the need for complex scoring. Developable inbound mail queue; Those awaiting observation will be reviewed quarterly; Remove abandoned items from the list and indicate the reason. Mixing the three levels into one scoring sheet often results in an intermediate value, making it difficult to make a decision.

Aline GPT: This step uses customer scoring and intent recognition: updating scores based on response, open, and lingering behaviors, and automatically prioritizing lists. Manual judgment is faster when the list has dozens of entries. After hundreds of entries, manual sorting is prone to inconsistent standards, and the effect of scoring begins to emerge.

The fourth step is to analyze procurement patterns and find a window of opportunity to enter.

This step is the real value of customs data, and it is also the step that most people skip.

Look at three things. First, there is the procurement cycle. Is it quarterly, monthly, or concentrated in a certain season? For categories with long cycles, if you miss the window, you have to wait for the next round. Secondly, fluctuations in goods value, such as a sudden drop in the value of a single item, may indicate that there are problems with the supply of goods from existing suppliers. A sudden increase may indicate business expansion. Thirdly, there is the supplier switching point. If a buyer adjusts suppliers at a fixed time each year, the first month is the best period for contact.

There is a real operational logic to refer to: a hardware accessories export team discovered that a Middle Eastern buyer had changed some suppliers before the twice-annual procurement season for three consecutive years. So they made their first contact six weeks before the procurement season, attached a comparison table of commonly used specifications in the market, received a reply within two weeks, and received a trial order four months later. The key to the case is not luck, but the fact that the timing of contact is derived from data, not based on gut feeling.

Step 5: Complete the decision-maker and available email addresses.

Customs data usually does not contain contacts; this step requires separate collection and sequential verification.

The collection is based on three roles: the purchasing or category manager is responsible for product selection and price comparison, which is the primary goal; The technical or quality manager determines whether the specifications pass, which has a high weight in the industrial product and customized product categories; The boss or general manager makes decisions directly in a small company. The order of data collection is current employees on the official website contact page and team page, LinkedIn company page, industry exhibition directory, and local business registration database. Matching LinkedIn employees by company domain name is much more accurate than searching by company name.

Verification is based on five levels of filtering: whether the domain name exists and whether there is an MX record; SMTP detection confirms that the address actually exists; Identify catch-all domain names; this type of server returns success for all addresses, but the result is untrustworthy. Exclude disposable email addresses and group addresses; Finally, look at the historical email return records. After completing five layers, those with named email addresses and verified can be sent directly. catch-all tests will be conducted in small batches first, with only the downgrade of info@ pending observation.

Aline GPT:Collection and verification are made into an assembly line: companies and contacts enter the same process, complete key roles, and verify each email address one by one. Those that fail the verification are blocked before being sent. What is mainly saved is the time for verification and deduplication, which is the most time-consuming and easily missed step in manual processing.

Step 6: Send the first letter and follow up.

The first letter is most easily written as a self-introduction. The group letter went unanswered, not because the product was bad, but because the recipient couldn't see how the letter related to them.

A three-segment structure is sufficient. The first paragraph provides a fact related to the other party's business that can be verified from publicly available information, and does not involve transaction details; The second paragraph provides verifiable specific information, including model, certification, minimum order quantity, and delivery date. The third paragraph raises a question that can be advanced: quantity, target market, or replenishment time. As for the content beyond the three paragraphs, we will wait for the other party's reply before elaborating

The subject line determines whether to open it or not. Writing words like "cooperation" and "supply" has a very low open rate; Writing about the specific issues the other party is dealing with, such as a shortened delivery period or an alternative to a certain type of certification, will significantly differ the open rate.

Follow-up should be done in more than two rounds. Within 48 hours of sending the email, send a short message through another channel explaining that the email has been sent and what materials can be added. The second week, send it again from a different angle, for example, send a specification comparison table of the other market's specifications. Most people who only send one letter and then wait never receive a reply.

Aline GPT:This section is executed using AI SDR: reaching out via email, WhatsApp, and LinkedIn at a predetermined rhythm, adjusting the next step based on the responses, and supporting more than eighty languages on the content side. Teams that operate in multiple markets in parallel but don't want to write every letter on the spot save on language and pacing. For teams with a single market and single channel, manual follow-up is more controllable.

IV. How to use different product categories in three scenarios

  • Scenario 1: Industrial equipment with high average order value and long decision-making chain.The focus should be on records of goods valued at over $100,000, and whether the supplier has been single for a long time. A single supplier means higher risk and greater opportunity for replacement. Don't quote a price upon initial contact; instead, provide technical documentation or similar cases to let the other party assess their professionalism.

  • Scenario 2: Auto parts, with frequent bulk purchases and repeat purchases.The focus should be on purchase frequency rather than the value of a single item. Buyers with monthly records indicate fast supply chain consumption and stable demand. The development letter emphasizes inventory security and replenishment speed, which is more effective than emphasizing price.

  • Scenario 3: Textile raw materials are price-sensitive and highly competitive.Prioritize buyers with more than three suppliers; these customers are price-sensitive and willing to switch. Calculate the other party's average import unit price using records. If the quoted price is about five percentage points lower than the average price, and directly state this in the letter, the response rate will increase significantly.

The three scenarios have different verification depths, but the order is the same: subject, existing, product category, and contact person, leaving no stone unturned.

V. Five common mistakes

  1. Search using the product's common name.The match rate was low, and a group of buyers were completely out of sight. The correct approach is to first check the HS code, and then use the code plus keyword combination to search.

  2. Download the entire table without filtering.Exporting thousands of records at once may seem fulfilling, but in reality, it's difficult to know where to begin. The correct approach is to first select the logistics provider, and then divide them into three tiers according to four criteria.

  3. Just look at the company name, not the history.The list looks clean, but the timing of the email is completely wrong. The correct approach is to look at the procurement cycle, value fluctuations, and supplier switching points.

  4. Bulk sending of template development letters.The extremely low response rate can also drag down the credibility of the sending domain. The correct approach is to quote at least one fact related to the other person in each letter.

  5. Send only one email.Once ignored or in the trash, there is no second chance. The correct approach is to complete the first follow-up within 48 hours via email, LinkedIn, and instant messaging.

Of the five mistakes, the first and second are the easiest to correct; you'll see the difference that week after they're corrected. The third requires some analytical skills, while the fourth and fifth involve writing and rhythm, and it usually takes two or three rounds to get a feel for it.

conclusion

Customs data does not automatically generate orders; its purpose is to transform "finding customers" from a matter of luck into a well-founded judgment: who is buying, how much, how often, and from whom. The rest of the persuasion work still depends on people and faith.

The most common mistake beginners make is treating data as a list, and the list as a customer. The three steps involve verification: whether this record belongs to the purchaser, whether this company still exists, and whether this email address is usable. If you skip the verification, all the previous screening will be for nothing.

To judge whether the process is working, there is a simple criterion: randomly select a record from three months ago and explain why it was assigned to this level and what to do next. If you explain it clearly, the process is straightforward.


frequently asked questions

Q1: Where can I check customs data, and what tools should beginners use?

There are two types of tools on the market that solve different problems. One type is free or low-priced bill of lading inquiry sites, such as ImportYeti and Port Examiner, which only cover a few countries, have shallow fields, and usually have no contact persons. They are suitable for first verifying whether the need actually exists and testing the waters of a market. Another type is paid data platforms, such as Volza, Panjiva, and Xingzhi Customer Acquisition, which cover multiple countries, support batch filtering and contact completion, and can send emails and follow up all the way. The price is high subscription costs, inconsistent list quality, and the need for self-verification.

The safest procedure for beginners is to first use the free version to go through the process, confirm that the information is valid for their product category, and then decide whether to subscribe. Xingzhi Customer Acquisition is a paid platform, and its pricing is not disclosed on the official website. A demand acquisition plan is required.

Q2: Can customs data track customers' email addresses and phone numbers?

It's not in the original customs declaration records. The email addresses and phone numbers displayed in the system were collected by data providers from publicly available channels such as their official websites, LinkedIn, and enterprise databases, resulting in significant differences in quality. The criterion for judging is not "whether it exists", but "whether it has been verified or not": whether SMTP probing has been performed, whether catch-all can be identified, and what the bounce rate is. Contacts without verification are essentially nonexistent.

Q3: My English is not good, can I use customs data to develop clients?

The query and filtering process is fine, and most systems have a Chinese interface. The communication process depends on the target market: in some Southeast Asian and Middle Eastern markets, a brief English text with pictures can be used first, while in European and American markets, a complete expression is required. A feasible approach is to prepare three to five sets of short templates for different scenarios and modify them with translation tools, rather than writing each letter on the spot.

Q4: Popular customers have already been contacted by many people, is it still valid?

It's effective, but it relies on differentiation. Popular buyers do receive a lot of emails, most of which are templates. If the letter can indicate the other party's procurement pace, supplier structure, or recent changes, and provide corresponding solutions, the probability of it being read will be much higher. The value of customs data lies not in its exclusivity, but in its accuracy.

Q5: Is customs data the only way to do it alone?

It's not the only one, but it's suitable as a main player. Single-person teams have limited time and budget, resulting in the lowest marginal cost of data development. A reasonable combination is to focus on data development, LinkedIn to build trust, and instant messaging tools to follow up. We will add exhibitions and platforms once our business stabilizes.

Q6: How do I choose data tools, and what points should I consider?

Fourth question: Is the target market within the coverage range? Whether the update frequency keeps pace with decision-making; Whether the contact person and their verification results are provided; Can it be exported and downgraded at any time? A platform that covers two hundred countries but whose target market data is six months behind is not as good as one that only covers ten countries but is updated on time every month. Price is only one aspect; exit costs must be considered together.