Practical practice on batch screening of 100,000-level US WS numbers: batch splitting, deduplication, notification and export strategies

#100,000-level US WS number batch screening practice: batch splitting, deduplication, notification and export strategy

The team that is engaged in customer acquisition in the US market has accumulated 100,000 or more US ws numbers (i.e. US WhatsApp numbers) and needs to quickly verify which target users are open, active and reachable. If you cram all the numbers into one task, it will be easy to waste all your efforts due to unexpected interruptions, and you will not be able to flexibly adjust the detection strategy. This practical guide will teach you step by step from batch splitting, data deduplication, task notification to result export, through the large batch US WA number screening process.

Why do large batches of US WS number screens need to be refined and split?

The upper limit of a single KK-DATA screening task is about 1 million, and theoretically 100,000 levels can be submitted at one time. However, in actual operations, submit in batches is a more secure approach for three main reasons:

Reduce the risk of failure: The larger the task size, the longer the single running time. Once the network fluctuates or the detection type needs to be adjusted temporarily, the entire task has to be restarted. Split into small batches, even if there is a problem with one batch, other batches will not be affected.
Easy to monitor progress: After the batch processing is completed, you can view the results in batches to judge the quality of the current number segment. For example, the first batch of 50,000 accounts has an activation rate of 70%, and the second batch of 50,000 accounts has an activation rate of 50%. The difference in account sources can be discovered immediately and the strategy can be adjusted in a timely manner.
Flexible adjustment of test type: You may only want to test “activation” at first, and then decide to add “active” or “gender” test after reading the data. By submitting in batches, you can assemble the detection plan like building blocks without waiting for a giant task to be completed.

Batch splitting suggestions

The upper limit for a single mission is about 1 million, but it is recommended to split the 100,000-level US WS numbers into 50,000-100,000 per batch. This not only ensures efficiency, but also leaves room for trial and error.

How to scientifically split the batches of 100,000-level US WS number tasks?

Splitting the batches is not a brainless split. Depending on your customer acquisition goals, you can choose one of the following three splitting strategies or use them in combination.

Split by number quantity: How many numbers are recommended per batch?

This is the most basic method of disassembly. If you have 300,000 US WhatsApp numbers, you can split them into 3 batches × 100,000 numbers, or 6 batches × 50,000 numbers. It is recommended that novices start with 30,000 to 50,000 items per batch, and then gradually increase to 80,000 to 100,000 items after running through the process. When the console submits a task, the estimated cost of the batch will be displayed to facilitate budget control.

Split by detection type: filter activation first, then filter activity/gender

This is a cost-effective approach, especially for teams with a limited budget.

First round: Only do “activation detection” (that is, verify whether the number is registered with WhatsApp), and quickly eliminate invalid numbers.
Second round: Conduct “activity detection” on the activated US WS activation numbers and screen out users who have been online in the past 7 days or 30 days.
Third round (optional): Conduct “gender/age detection” on active numbers to further refine the population.

The advantage of this is that the first round costs the lowest, and after eliminating most of the invalid numbers, the investment amount in the next two rounds is significantly reduced. And you can pause at each step and analyze the data before deciding whether to continue.

Split by number segment or source file

If your numbers come from multiple channels (such as advertising forms, offline collection, purchased data packages), the quality of numbers from different sources is often different. It is recommended to directly split it into different batches according to the source files. After running, compare the activation rate and activity rate of each batch, so that you will know which channel source to focus on purchasing in the future.

Cross-batch data deduplication: avoid duplicate detection wasting balances

Submitting in batches is prone to a hidden danger: duplicate numbers between different batches. For example, there may be 20,000 duplicates among the 100,000 items imported for the first time and the 50,000 items imported for the second time. If not processed, these 20,000 numbers will be tested twice, and charges will be deducted twice in vain.

KK-DATA’s data deduplication warehouse solves this problem. You can import all numbers into the warehouse first, and the system will automatically identify duplicates across batches and only deduct one inspection fee.

Operating steps:

Enter “Data Deduplication Warehouse” in the console.
Upload your number file (supports CSV/TXT), the system will automatically remove duplicates and generate a deduplicated number pool.
Create a filter number task from the deduplication pool, and no duplicate numbers will appear when submitting.

Saving balance key

Developing the habit of “removing duplicates first, then screening numbers” can help you save 10%–30% of your balance, especially for long-term operating teams. For details, please refer to the platform documentation Data Deduplication Instructions.

Task notification settings: Understand the progress of US WS number screening in real time

High-volume tasks often take several hours or more. You can’t keep your eyes on the console. KK-DATA supports pushing task completion notifications through Telegram robots, which is very practical.

Bind Telegram robot to receive notifications

On the “Personal Settings” or “Notification Settings” page of the console, find the Telegram binding entrance.
Click and jump to Two-way contact customer service robot @kkdata_robot, and follow the prompts to obtain a binding code.
Enter the binding code into the console to complete the binding.

Automatically push the result summary after the batch task is completed

After binding, you will receive a push on Telegram when each screening task is completed, containing:

Task number and name
Detection platform (such as US WhatsApp)
Total number of submitted numbers and number of successful tests
Number of hits for each detection type (such as number of activations, number of actives, number of gender identification)

In this way, even if you are running multiple tasks at the same time, you can still see them clearly without having to log in to the console repeatedly.

Export and data utilization: How to use filter results efficiently?

After the screening is completed, the most important thing is to export the data and use it in real marketing scenarios.

Export format selection and field description

You can choose to export to CSV or TXT in the console. CSV is recommended because the field structure is more complete.

Typical export fields include (subject to actual console export): -Mobile phone number (country code + number)

Whether to activate WhatsApp
Last online time (based on activity)
Gender (male/female/unknown)
Age (approximation, can be used to interpret people around 30 years old)
wsid (WhatsApp Business ID)

Connecting result data with overseas marketing channels

Exported CSV can be imported directly to:

CRM system: Tag according to activity and gender to achieve differentiated contact.
Marketing Automation Tools (such as HubSpot, SendGrid): Set trigger conditions, for example, users who are “active in the past 7 days + male” will receive copy A, and users who are “female + active in the past 30 days” will receive copy B.
Third-party dialing/messaging tools: carry export fields to initiate marketing in batches to reduce invalid outbound calls.

Best practices and pitfall avoidance guide for screening large batches of US WS numbers

Based on actual operating experience, the following five points deserve special attention:

Don’t submit too many numbers at once: Although the upper limit is 1 million, it is recommended to divide it into batches of 50,000-100,000 to leave room for trial and error.
Be sure to turn on data deduplication: If the same number is detected twice, it is equivalent to spending twice as much money. The deduplication warehouse was born for this purpose.
The active window should be chosen reasonably: If it is a short-cycle marketing (such as a promotion activity), choose to be active in the past 7 days; if it is a member’s continuous operation, choose the last 30 days. An active window that is too narrow will result in a sharp decrease in the reachable number.
Pay attention to the real-time price of the console: The unit prices of different platforms and different detection types are different. Before submitting the task, the console will display the estimated cost, so be sure to check it.
Use the anti-fraud query function: The platform provides official Telegram account verification, and beware of scams pretending to be customer service.

FAQ

**Q: Can I submit the 100,000-level US WS number at once? ** Answer: The upper limit of a single task on the platform is about 1 million, but in order to avoid long running times and unexpected interruptions, it is recommended to submit in batches, for example, 50,000-100,000 items per batch. See console prompts for details.

**Q: How to choose the activity window for US WhatsApp number detection? ** A: When submitting a task, you can select a time window (such as the last 7 days, the last 30 days, etc.) in “Activity Detection”. For marketing scenarios, it is recommended to select numbers that have been active in the past 30 days to ensure reach.

**Q: Does the data deduplication warehouse support cross-task deduplication? ** Answer: Supported. You can upload numbers from different batches and sources to the warehouse, and the system will automatically identify and eliminate duplicate numbers to avoid repeated deductions.

**Q: After the detection is completed, what fields will be included in the exported US WA number? ** Answer: The export result includes the number, whether WhatsApp is activated, activity level (such as the last online time), gender (if any), wsid, etc. The actual export from the console shall prevail.

**Q: How long does it take for USDT to arrive after deposit? ** Answer: Usually the balance is automatically updated after USDT (TRC20) arrives in the account, usually within a few minutes. If your account has not been credited for a long time, please contact customer service.

Are you ready? Now is a great time to batch screen numbers. 👉Log in to the console to start screening numbers, or click Two-way Contact Customer Service to get real-time help. For more detailed operations, please refer to Usage Document and Official Home Page.