How to judge the quality of foreign trade customer leads? What do companies, contacts, and email addresses verify?

The development letter went unanswered, with the common explanation being poor copywriting. But before changing the copy, there is a more fundamental thing that needs to be ruled out: whether the clue itself is true. The business may have long since ceased operations, the contact person may be an assistant, and the email address may not even exist. All three scenarios look exactly the same in the inbox: no reply.

The problem with the quality of clues rarely lies in quantity; most of the time, they are empty at a certain level. The fact that the company entity is genuine does not mean that the contact person is correct; The fact that the contact person is correct does not mean that the email address will be delivered; Just because the email address was delivered doesn't mean the person is still employed. Each of the three levels makes mistakes, and they don't guarantee each other. This paper breaks down the verification into three layers: what the business entity checks, what the contact person checks, and what the email address checks. Each layer provides operable judgment criteria and public inquiry channels, and finally explains the order and waiver conditions of the three layers.

Quick conclusion

1. A single lead carries three independent risks: the subject is not real, the contact person is incorrect, and the email address cannot be delivered.If no layer is verified, the previously invested list costs and letter writing time will be wasted.

2. The company checks four things: whether it is in operation, whether the name and address are consistent, whether there are genuine import and export traces, and whether the official website and social media can verify each other.If one of the four items doesn't match, there's no need to spend any more time moving to the next level.

3. The contact person checks three things: whether the position affects procurement, whether they are still employed, and whether the name and email naming rules match.For people whose positions are unrelated to procurement, no matter how high the response rate is, it does not constitute a valid clue.

4. The email layer first answers a question: will this letter be delivered to a real person?Domain ownership, whether the mailbox exists, and historical bounce history determine whether it is available, pending confirmation, or directly rejected.

I. Why must verification be divided into three layers?

The three levels of risk are unrelated, which is the reason for layered verification. The company is genuine but the contact person has left; the email address is valid but the address belongs to the sales department's public email address; the subject and email address are correct but the other party is only an end user and does not have purchasing rights—any combination will make a well-written development letter fall through.

The verification costs of the three layers vary greatly, which determines the order. Most company information can be verified free of charge, contacts need to cross-check publicly available information, and email verification often requires batch verification using tools. Starting with low costs allows you to weed out substandard leads early on, avoiding paying three times for a single false lead.

More practical than scoring is elimination. The lead scoring model requires historical transaction data to be calibrated, and most teams do not have this foundation; A one-vote veto only requires a few hard rules, allowing for quick judgment and minimal controversy. It is more reliable to eliminate obvious substandards first, and then sort the rest according to relevance, than to give each clue an unfounded score.

II. Business Entity: First, confirm that the other party is doing this business.

Four basic verifications

• The entity exists and is in operation: registration information can be found, and the registration status is not deregistration, revocation, or liquidation.

• The name and address match: the header, official website domain name, registered address, and shipping address point to the same entity.

• There are genuine business traces: the product page has specific specifications and models, not just a few pictures and a "Welcome to inquire" message.

• Mutual verification between the official website and social media: The address, phone number, and year of establishment disclosed on the official website match the LinkedIn company page.

Exporters check customs announcements, importers check local registrations.

• Domestic suppliers: The China Customs Enterprise Import and Export Credit Information Disclosure Platform can access customs registration and filing information, credit ratings (including AEO certification results), administrative licensing and administrative penalty records.

• EU buyers: The European Commission's VIES system can verify whether the VAT number is valid, that is, whether the other party is a registered entity for cross-border trade; Please note that some member countries do not return names and addresses; they can only confirm that the number itself is valid.

• Other markets: Based on publicly available information from your local business or Companies registry; Secretary of State Business Database is available in North America; Companies House is available in the UK.

One change is important to note: after the abolition of foreign trade operator registration at the end of 2022, "whether or not there is import and export operation right registration" is no longer suitable as a judgment criterion. Instead, customs registration information and corporate credit disclosure are more accurate.

Three easily misjudged signals

• The domain name has been registered for a very short time, but the official website claims to have been operating for twenty years.

• The address is a virtual office or agency registered address, and there is no verifiable information about warehouses, factories, or stores.

• There are only social media accounts, no official website; Or the official website may only have one page, without product details or company information.

These three categories do not equate to scammers, but they mean that verification must be more stringent: at least one phone call must be made, or the other party must be asked to provide official documents with the company's name on them.

An easily overlooked layer: the contracting entity

The brand owner and the contracting party are often not the same company. The group has multiple legal entities, and procurement may be centrally executed by headquarters or decided by regional branches. During the verification process, the signing party on the contract should be considered, and then its relationship with the target brand should be examined: if the signing party is inconsistent with the brand owner and the authorization relationship cannot be explained, such clues should be clarified before the quotation is made.

III. Contact Person: Looking for someone who can influence procurement.

Three judgment points

• Whether positions and responsibilities fall within the procurement chain: Procurement, supply chain, products, R&D, and operations usually have a say; Administrators, receptionists, and interns usually do not have

• Whether they are still employed: LinkedIn updates, the company's official website team page, and recent publicly available activity records. If any of these three can prove their employment, their credibility will be significantly increased.

• Does the name and email address match? Corporate email addresses often use first.last or f.last naming conventions. If the name and address structure do not match, this clue needs to be checked.

The decision chain is usually not one person.

In the same purchase, there are at least three types of roles: those who decide on budgets and suppliers, those who oversee specifications and certifications, and those who use and provide feedback on daily use. The three types of people care about different content. Who to write the development letter to determines whether the letter should talk about price, certification, or delivery and after-sales service. Finding only one of these types is only half the clue.

The contact layer can also be handled by tools.Xingzhi Customer Acquisition Pass (Aline)It uses AI to complete buyer profile analysis and 360° background testing, outputting multi-dimensional contact information for key decision-makers, including more than just one email address.

How to find the right person

• Starting from the company page: On the official website's team page, investor relations page, and recruitment page, you can often see the names and reporting relationships of purchasing and supply chain positions.

• Search by job: Combining industry keywords with job keywords makes it easier to find specific positions than just searching for company names.

• Check back against the industry list: exhibitor list, industry association member list, usually with company and contact information, then return to the official website for confirmation.

• A reverse examination of publicly available procurement information reveals that the names of contact persons and departments appear in tender announcements and supplier change announcements, which is more credible than a list pieced together by data providers.

Four common misjudgments

• Treat assistants or receptionists as decision-makers: Most job titles that include "assistant" or "coordinator" do not have the authority to make purchasing decisions.

• Treat those who have left their jobs as if they were currently employed: LinkedIn pages are not automatically updated, and old pages often remain for months after changing jobs.

• Different people with the same name: In group companies, positions with the same name may be distributed across multiple subsidiaries, requiring company domain name and department confirmation.

• Treat the end user as the purchaser: the other party does use the product, but the purchase is determined by their customer or headquarters.

IV. Email Address: First, confirm that it can be delivered, then see if you are interested.

First look at the domain name

• Enterprise-owned domain names: the most credible and can also be used to infer company information.

• Free email (Gmail, Outlook, Yahoo, etc.): In B2B scenarios, appearing with small customers or individual traders is an abnormal signal for large purchases.

• Disposable email addresses and temporary domains: Remove them directly; these types of addresses will not lead to real conversations.

Then check if the mailbox exists.

• Check the domain's MX record to confirm that the domain can receive emails; For domains without MX records, emails will inevitably be bounced.

• Perform SMTP handshake verification to determine if the specific address exists; However, when the enemy server enables catch-all, any address will be accepted; at this point, it can only be marked as "pending confirmation" and cannot be considered valid.

• Role emails (info@, sales@) are usually deliverable, but the response rate is lower than that of named emails; It can be used during batch development; for formal follow-up, it should be replaced with a person's name and address.

The five types of email addresses are handled differently, and mixing them together can mask the problem:

Finally, let's look at history

• Whether it appears in the bounce record: If the same address bounces twice in a row, it will be removed from the list.

• Whether on the suppression list: Addresses where complaints have been filed and where explicit refusals have been made should not be accessed again.

• Does it match the corporate domain name? If the domain names don't match, even if it can be delivered, it indicates that there is a problem with the contact ownership of this lead.

What verification tools can and cannot confirm

Verification tools are good at answering questions such as "Does this address exist?" and "Can this domain name receive emails?", but they are not good at answering questions such as "Is this person currently employed?" and "Is this position responsible for procurement?" It is a common misuse to treat the tool's results as a complete conclusion about lead quality: a verified email address may belong to an employee who left six months ago, the address is still valid, but the recipient has changed jobs. The tool results are only used to judge delivery, not to judge the person.

Do small-batch testing first.

Before sending on a large scale, first select twenty or thirty items from the list and send them in a round according to the normal process, observing the three items of bouncing, opening, and replying. Bounce emails are concentrated in a few domains, indicating that there are problems with the quality of emails from these sources; If the message opens normally and the reply is zero, the problem is more likely in the content of the message. Small-batch testing is inexpensive, and it avoids depleting the credibility of an entire list at once.

Email verification only answers "can it be delivered?" and cannot answer "is the recipient interested?" What can be collected after delivery are behavioral signals: whether it was opened, how many times it was opened, whether the link was clicked, and whether the letter was forwarded to a colleague.Xingzhi Customer Acquisition Pass (Aline)The email probe can identify customer intentions, automatically screen high-intention customers, complete value assessments, and clarify follow-up priorities. The order of the two layers cannot be reversed: first, verify that the message will be delivered, and then use behavioral signals to decide who to follow first. Skip the verification and directly look at the open rate. The data will be diluted by bounce emails and invalid addresses, resulting in a distorted judgment.

V. How to Implement Verification: Order, Stop-Loss, and Tools

Ranked by cost from lowest to highest

1. Free public search: company registration information, customs announcements, VIES, official website and social media comparison

2. Batch validation of tools: email syntax, MX records, SMTP handshake, one-time domain name recognition

3. Manual confirmation: phone call, LinkedIn interaction, requesting official documents from the other party

The first two steps can cover more than 80% of non-compliant leads, and the remaining 20% should be invested in manual labor to keep unit costs down.

Four stop-loss lines

• The company entity cannot be found, or the registration status is abnormal: it will be removed directly, and the contact person and email address will no longer be verified.

• The email address domain is unrelated to the company and cannot be explained: it will be removed unless the recipient proactively uses that address to send a letter.

• The contact person is not found, or the position is not related to procurement: return the list pool, find the person again, and do not force the letter.

• Verification time exceeds the threshold: If a lead has been investigated for more than ten minutes and remains unconfirmed, it should be classified as "pending confirmation" instead of being investigated indefinitely.

Sampling review

A percentage of each batch of leads is selected for manual review, and the type of error is recorded. An error rate exceeding 20% indicates a problem with the data source or search conditions. In this case, changing the source is more effective than continuing to clean. The purpose of this review is to determine the credibility of the source, not to correct errors one by one.

In three levels, the results must be traceable.

Split the company name, registration number, country of origin, contact person's name, position, email address, and verification status into independent fields, with each field indicating the source and judgment rules. There are only three levels of verification results: available, pending confirmation, and rejected.

• "Pending confirmation" will be stored separately and will not enter the sending queue. It will be transferred after manual confirmation.

• Each time the reasons for recording are removed, monthly statistics are compiled, and the areas with the highest concentration of errors are those that need to be corrected.

• When a clue is eliminated, it's important to be able to identify which item fails to meet the standards, rather than relying on intuition.

Good tools can improve work efficiency.Xingzhi Customer Acquisition Pass (Aline)It is an AI customer acquisition intelligent agency for foreign trade B2B scenarios. Its accurate potential customer discovery capabilities across the entire domain cover three layers in this article: The enterprise layer uses more than ten resources such as customs data, global enterprise databases, search engines, yellow pages, exhibitions and trademarks to collect data, and combines HS code analysis with target country risk warnings to first eliminate mismatched product categories and high-risk market targets. The retrieval, backtesting, and verification at the contact layer and email layer are completed in the same process, eliminating the need to import data back and forth between several tools.

Sum up

The value of three layers of verification lies not in the detail, but in the sequence. First, confirm that the company actually exists and is operating, then confirm that the contact person is in the procurement chain, and finally confirm that the email address can be delivered to a living person. If you don't pass any level, you don't need to start any subsequent actions.

The verification process ultimately involves two actions: rejection and review. To eliminate the need for hard rules, a few veto standards are sufficient; The verification relies on sampling, using 10% of the samples to determine whether the source of the entire batch is credible. Once both are done, the list size will usually shrink, but the probability of each letter sent ending up in the hands of a real buyer will increase significantly.

Frequently asked questions

Q1: To what extent can verification be achieved using only free channels?

The enterprise level is basically sufficient. Registration information, customs credit information, and EU VAT numbers can all be checked for free, and the official website and social media can also be compared for free. The email layer can only check the syntax and MX records to determine whether the domain name can receive emails; it cannot confirm whether the specific address exists. This step requires a verification tool or manual emailing test.

Q2: Do people on LinkedIn have to be added as friends to be verified?

No need. The public information page already shows the position, current company, tenure, and recent developments, which is enough to determine whether it has fallen into the procurement chain. The value of adding friends lies in the subsequent interaction, which is two separate things from the verification itself. There is no need to send invitations in batches for verification.

Q3: Do I need to send the email address for the catch-all domain?

You can post them, but they need to be categorized separately. The existence of these types of addresses cannot be confirmed by a handshake; it can only be determined based on the results of the first sending: if the email is returned, it will be rejected; if the email is not returned, it will be followed up according to low priority. Mixing them into the normal list will lower the overall delivery rate and response rate statistics.

Q4: The quality of the leads is poor. Should we change the data source or the development letter?

First, look at the email return rate. The high bounce rate lies in the email address and data source. The bounce rate is normal but the response rate is low; the problem lies in the relevance of the list to the content of the letter. Only by looking at the two indicators separately can we avoid misjudging data issues as copywriting issues.

Q5: How many leads can be verified per day?

It depends on the degree of automation. For a purely manual search of companies and contacts, processing 50 to 100 entries per day is the upper limit. Batch email verification can achieve thousands of requests per hour, but the verification results still require manual sampling and verification. Once the scale increases, the bottleneck is usually not in the query, but in the manual judgment in the "to be confirmed" part.