Customer acquisition data combination FAQ: How to combine the sieve number system, base materials and testing to go overseas efficiently?

#Customer Acquisition Data Combination FAQ: How to combine the sieve number system, base materials and testing to efficiently go overseas?

In overseas marketing, the quality of customer acquisition data directly determines promotion costs and conversion efficiency. Although the original number pool (commonly known as “base material”) obtained by many teams is huge, it contains a large number of invalid numbers, inactive users and even wrong numbers, which are directly used for Telegram group sending, WhatsApp marketing or Line private messages. The result is often low message sending rate, high risk of account suspension, and bleak ROI. This article uses FAQ form to sort out the matching logic of the sieve number system, base materials and detection types, helping overseas teams avoid common pitfalls and establish a reusable and efficient sieve number process.

What is customer acquisition data? What do screen number system, base material and detection mean respectively?

Before talking about customer acquisition data, let’s clarify three core concepts to avoid confusion:

Base material: refers to the original number pool without any filtering, which may come from global random generation, batch purchase of number segments, CSV import or API crawling. The base material only contains the number itself, without any status information (whether it is activated, active, user gender, etc.).
Screening system: refers to a tool for batch detection of number status, such as a platform like KK-DATA. After receiving the base material, the screening system calls each platform interface (Telegram, WhatsApp, Line, etc.) to return the test results corresponding to each number.
Detection type: refers to the specific detection items performed on a certain number, such as “Telegram activation detection” (to determine whether the number is registered with Telegram), “Telegram activity detection” (to determine whether there has been activity in the last X days), “Gender detection” (to identify the user’s gender, age and other attributes).

The relationship between the three is like: the bottom material is unprocessed ore, the screen number system is the mineral processing equipment, and the detection type is the screening net with different accuracy. Only with a reasonable combination of the three can the ore be turned into usable “valuable metals” - that is, high-quality customer acquisition leads.

How to match the screen number system with the base material? Why can’t we just use the base material for marketing?

Directly using base ingredients for marketing is equivalent to casting a net into the sea, and only a few fish will be caught. The reason is that there is a lot of invalid data in the base material:

The number has not been registered on any social platform (empty or not activated)
The number is activated but has been inactive for a long time (for example, it has not been online for half a year)
The number belongs to a non-target user group (mismatch in gender and region)

The correct sequence for matching the sieve number system should be:

Remove duplicates → 2. Activate detection → 3. Activity filter → 4. Targeting (gender/age) → 5. Export

Every step saves subsequent marketing costs. For example, among 1 million profiles, only 500,000 may have opened Telegram, of which less than 200,000 have been active in the past 30 days, and 100,000 are left after gender filtering is added. The final 100,000 messages sent are high-converting leads, far better than sending directly to 1 million invalid numbers resulting in a large number of failures or account bans.

Base material sources and common formats

There are three main sources of base ingredients:

Global random generation: Select the country and number segment within the KK-DATA platform and randomly generate any number of numbers. This feature is completely free and provides zero-cost access to a pool of numbers from 240+ countries around the world.
Global number segment generation: Generate consecutive numbers in batches for specific operators or number segments, suitable for regional targeted marketing.
Custom number segment CSV import: If you already have your own number list (such as numbers collected from exhibitions and website registration forms), you can directly upload the CSV file as the base material.

Note: There is no charge for generating the base material itself. Fees will be deducted based on the number of test items only after the screening task is submitted and completed. Therefore, it is recommended to generate a large amount of base materials for free first, and then gradually screen them to find the most cost-effective batches.

Key checkpoints in the screening process

The following is the recommended standard sieve number assembly line, each link corresponds to different detection types:

Steps	Detection content	Function	Sample configuration (Telegram scenario)
1. Deduplication	Data deduplication warehouse	Avoid duplicate detection of the same number and waste of balance	Automatic cross-task deduplication
2. Activation test	Whether the number is registered with the platform	Exclude empty numbers and unregistered numbers	tg activation test
3. Activity filtering	Whether the number has been active recently	Lock online users to increase the reach rate	tg active (last 30 days)
4. Targeted screening	Gender, age, region, etc.	Match target customer portrait	tg Gender (male) + age about 30 years old
5. Export	Export activated/active/targeted list	Get high-quality lead list	CSV or TXT export

It is recommended to use the “Estimated Cost” function to confirm the required balance before submitting a batch task to ensure that it can be completed in one go.

There are so many detection types (Telegram activation/active/gender, etc.), which ones should I choose?

The more tests the better, but the combination according to the marketing scenario. Common scenarios and recommended detection types are as follows:

Typical detection configuration of Telegram customer acquisition scenario

Group promotion (group promotion, channel promotion): “tg activated + tg active (last 7 days or last 30 days)” is required. Activation ensures that messages can be received, and activeness ensures that users will open and view it to avoid wasting resources caused by sending to zombie accounts.
Private message advertising (one-to-one marketing): It is recommended to add “gender detection (such as tg gender-male)” on the basis of activation + activity to screen out users of the target gender and increase the click-through rate.
Coin Circle/Community Newcomers: You can add “tgid export” for further processing in the Telegram official API (such as adding friends in batches). The age field comes from the accessory field of the gender detection result, which can assist in determining whether a person is close to 30 years old, but it cannot be used as an independent age detection product.

Age field description

The age field in the gender detection results is based on group trend data inferred by the model. It can be used to screen people who are “about 30 years old” and does not have ID card-level accuracy. It should not be understood that the specific date of birth of each user can be precisely known.

Applicable scenarios for WhatsApp / Line / Zalo

WhatsApp (Global Market): It is recommended to “activate + active (last 30 days)”. After exporting the wsid, you can use automated tools to create private messages. WhatsApp has strict restrictions on marketing and it is recommended to only push it to active users.
Line (Japan, Taiwan, and Thailand markets): Line has a concentrated user base. You can add “gender detection (male/female)” and export the uid for subsequent Line official API access.
Zalo (Vietnam market): The first choice in the Vietnamese market. Zalo screening usually includes activation, activity, and gender, and is targeted at localized promotion in Vietnam.

The specific export fields supported by each platform detection type are subject to the export list after the console task is completed, and may be updated at different times.

How to implement a “Generate → Filter → Export” pipeline?

The complete operation closed loop is completed in three steps within the KK-DATA platform:

Generate base material: Enter the console, select “Global Number Generation”, specify the country, number segment, quantity, and click Generate. The base is automatically saved in the task list.
Submit screen number task: Import base material into a new task, select the detection platform and specific detection type (multiple platforms can be combined for simultaneous detection, and the maximum number of items in a single task is about 1 million). The system displays the estimated fee, and submit it after confirming that the balance is sufficient.
Task Completion Notification: After the screening is completed, the system can send a notification through the Telegram robot (which needs to be bound in advance) to avoid staring at the progress bar all the time.
Export results: Download the CSV or TXT file on the task details page, including each detection field (such as tgid, activation status, active time, gender, etc.).

Free generation + on-demand detection

The number generation function is completely free, and random numbers and global number segments from all 240+ countries can be obtained at zero cost. Only after the screening task is submitted and completed, the fee will be deducted based on the number of tests. See the real-time price on the console for details.

This pipeline is very suitable for batch processing: for example, use global number segments to generate 500,000 Vietnamese numbers, first detect Zalo activation, then detect activation of activated users, and finally export a list of 100,000 active users for marketing.

Why is data deduplication important? How to avoid wasting balance due to repeated testing?

Data deduplication is the most direct way to save budget. Suppose you have tested a batch of numbers before, but 20% of duplicate numbers are mixed in the newly imported base material. If you do not remove duplicates, the 20% will be deducted again and wasted.

KK-DATA’s data deduplication warehouse can be intelligently merged across tasks: all detected numbers automatically enter the deduplication pool. When a new task is submitted, the system automatically excludes already detected numbers and only charges for new numbers.

Best practices:

Before importing base materials each time, use the “Duplicate Warehouse” to upload the number list on the console, and the system will automatically mark which numbers have been detected.
It is recommended to manage different batches of base materials in a unified manner and regularly clean up expired data (for example, the number tested 3 months ago may no longer be active and needs to be retested).
Don’t submit the entire big list at once as one big task. You can split it by market or platform to facilitate adjustments midway.

Insufficient balance cannot submit new tasks

When the balance is lower than the estimated cost of the task, the system will prevent submission. It is recommended to recharge regularly, the minimum USDT (TRC20) is about 50 USDT, and the balance will be automatically updated upon arrival. Pay what you use, no pressure to subscribe to a package.

How to set the filter number parameters in different customer acquisition scenarios?

Listed below are recommended screen size parameters for three typical scenarios for team reference:

Scenario 1: Zalo marketing for Southeast Asia (Vietnam market)

Base material generation: Select all number segments in Vietnam and generate 100,000-500,000 numbers.
Detection Platform: Zalo
Detection type: activated + active (last 7 days) + gender (male, if the product is targeted at males)
Export fields: uid, activation status, active time, gender
Expected results: Obtain 30,000-50,000 active male users, and can cooperate with local language promotion in Vietnam.

Scenario 2: WhatsApp group messaging for Europe and the United States (cross-border e-commerce)

Base material generation: Select country number segments such as the United States, the United Kingdom, and Canada to generate 300,000 items.
Detection Platform: WhatsApp
Detection Type: Open + Active (last 30 days)
Export fields: wsid, activation status, active time
Expected results: Acquire 100,000–150,000 active WhatsApp users and avoid the risk of account suspension.

Scenario 3: Telegram community attracting new users for cryptocurrency users (global)

Base material generation: Import existing lists from known currency circles or CSV.
Detection Platform: Telegram
Detection type: activated + active (last 7 days) + gender (male, if the product is for males) + tgid export
Export fields: tgid, activation status, active time, gender, age
Expected results: Obtain a list of highly active TG users for adding friends or inviting to the group.

FAQ

**Q: Is the more base the better? ** Answer: Not necessarily. Unscreened feeds can have a high proportion of invalid numbers, and direct marketing wastes time and delivery costs. It is recommended to use the screening system to check activation/activation before use.

**Q: Can the screening task be canceled midway after it is submitted? ** Answer: Once a task is submitted, it enters the processing queue and cannot be canceled. It is recommended to carefully check the number and detection type before submitting, and use the “Estimated Fee” function of the console to confirm that the balance is sufficient.

**Q: Can the age field be accurate to the ID card level? ** Answer: No. The age field comes from the inferred value in the gender detection results and can be used for group trend analysis (such as screening people around 30 years old), but does not have identification-level accuracy. Please do not understand it as an independent age detection product.

**Q: Where can I check the testing unit prices of different platforms? ** Answer: The real-time unit prices of each platform and detection type are displayed on the console before submitting the task, and there is also an overview on the [Billing Instructions] (https://kkdata.cc/billing/) page of the official website. The specific figures are subject to the actual quotation.

**Q: What fields does the exported data contain? ** Answer: Different detection types export different fields. For example, Telegram may include tgid, activation status, active time, gender, age, etc.; WhatsApp may include wsid, activation status, etc. The details are based on the export list after the console task is completed.

The above is the FAQ about the combination of customer acquisition data. If you still have questions about the specific configuration of the sieve number system, base material generation or detection type, you are welcome to log in to the console to experience the entire process, or contact customer service directly for one-on-one support.

👉Log in to the console to start screening numbers Two-way contact customer service: https://t.me/kkdata_robot

For more details, please refer to Usage Document or Official Blog.