KK-DATA avatar KK-DATA

Number Generation and Deduplication Warehouse Workflow: A Complete Guide from Batch Generation to Cross-Task Cleaning

去重 生成 kkdata 数据仓库

Number Generation and Deduplication Warehouse Workflow: A Complete Guide from Batch Generation to Cross-Task Cleaning

In overseas customer acquisition and community operations, data quality directly determines conversion efficiency. Many teams repeatedly encounter the same issue when batch-verifying numbers: the same set of numbers gets checked multiple times across different batches, resulting in wasted balance. This article focuses on the generation and deduplication workflow, starting from number generation to cross-task cleaning in the data deduplication warehouse, explaining step by step how to build a reusable high-quality number pool, avoid duplicate charges, and improve operational efficiency.

Why Combine Number Generation with a Deduplication Warehouse?

Common Pain Points in Data-Driven Customer Acquisition Scenarios

In practice, operations staff often need to collect numbers from different channels in multiple batches and then submit them to platforms like Telegram or WhatsApp for validity checks. The problem arises:

  • Duplicate numbers: The same number may come from multiple sources (e.g., list A and list B overlap). If submitted separately, the system charges for the overlapping numbers twice.
  • Balance waste: Duplicate checks not only waste balance but also extend task processing time.
  • Data chaos: Results from different tasks are scattered across multiple files, making it difficult to manage the verified number pool centrally.

How a Data Warehouse Solves Cross-Task Data Redundancy

The data deduplication warehouse acts as a “blacklist/whitelist” pool. It automatically records all numbers processed in each verification task and excludes existing numbers when submitting subsequent tasks. This way, the same number is only checked once, no matter how many tasks it appears in, and subsequent tasks skip it directly. Combined with the number generation module, you can build a number pool from scratch, filter it, store it in the warehouse, and form a closed-loop pipeline of “Generate → Filter → Deduplicate → Export.”

Step 1: Start with Global Number Generation

The number generation module is free and helps you quickly create an initial number pool without worrying about balance consumption. Generated numbers can be immediately submitted for screening or temporarily stored in your task list.

Tip

Generation is free; screening charges per number: The number generation module does not consume balance; charges only apply when using the screening function.

Batch Generate Numbers by Country and Number Segment

After logging into the KK-DATA Console, navigate to the number generation function. You can:

  • Select a country: Supports 240+ countries and regions, covering major overseas markets (USA, Southeast Asia, Europe, Middle East, etc.).
  • Set number segments: Enter a specific prefix (e.g., US +1 number segment), and the system generates random numbers within that segment.
  • Set quantity: Generate as needed, for example, 100,000 numbers.

Generated numbers are previewed in CSV or TXT format, ready for download or direct submission to a screening task.

Import Custom Number Segments (CSV Upload)

If you already have an existing number list (e.g., purchased data, business card collection), you can upload a CSV file directly. The system identifies the number column and incorporates it into the generation and deduplication workflow. This approach is especially useful for teams that need to clean old data.

Tip: When importing, it is recommended to first remove obvious invalid numbers (e.g., numbers with incorrect digit lengths) to reduce subsequent screening costs.

Step 2: Set Screening Goals and Platforms

After generating numbers, the next step is to define your screening goals. Before submitting a task, you need to decide:

  1. Detection platform: Telegram, WhatsApp, iMessage, RCS, etc. (multiple selections allowed).
  2. Detection metrics: Registration check (whether registered), activity (active in last 7/15/30 days), gender recognition (based on avatar AI recognition), export tgid/wsid, etc.

Choose the Right Detection Type

  • Registration Check: The most basic check, confirming whether a number is registered on a platform. Suitable for initial cleaning to quickly remove unregistered numbers.
  • Activity Check: Suitable for scenarios where you need to reach active users (e.g., community invitations, DM promotions). Activity windows differ by platform; choose the appropriate time period (e.g., for Telegram, select “active within 7 days”).
  • Gender Recognition: Used for targeted marketing (e.g., promoting women’s skincare products). Gender is determined via avatar AI; note that accuracy is not 100%.
  • tgid/wsid Export: Obtain the user’s unique ID for direct API communication, avoiding reliance on phone numbers.

Single Task Quantity Limit and Batching Strategy (~1 Million Entries)

The system supports a maximum of approximately 1 million numbers per single screening task. If your data exceeds 1 million, it is recommended to process in batches:

  • Batch 1: First 1 million → Screen → Write results to the deduplication warehouse.
  • Batch 2: Remaining numbers → Bind the same deduplication warehouse when submitting → System automatically excludes already checked numbers.
  • Subsequent batches follow similarly, seamlessly connected.

Batching ensures stable task execution while leveraging the deduplication warehouse to avoid duplication.

Step 3: Cross-Platform Multi-Verification Strategy

A single number may be registered on Telegram, WhatsApp, and iMessage simultaneously. In one task, you can select multiple platforms at once, and the system will check each number’s status on each platform and return combined results.

For example, if you need “active female users on Telegram who are also valid on WhatsApp.” Configure the task by selecting:

  • Detection platforms: Telegram + WhatsApp
  • Detection metrics: TG active (select 7 days) + TG gender (female) + WA valid

This completes cross-screening in one task, saving time and balance compared to two separate submissions. Cross-platform detection results are merged into a single export record for easy follow-up.

Step 4: Enable the Deduplication Warehouse – Best Practice for Cross-Task Deduplication

The deduplication warehouse is the core component of the generation and deduplication workflow. Without it, every new batch would re-check all numbers, causing massive waste.

Important

Always enable the deduplication warehouse: Before submitting a new task, ensure you have bound the corresponding deduplication warehouse; otherwise, the new task will not automatically deduplicate.

Create and Manage Data Warehouses

In the left navigation of the console, go to the “Data Warehouse” module. You can:

  • Create a warehouse: Enter a name (e.g., “US TG-2025Q1”), add notes for identification.
  • Set tags: Use tags to differentiate projects (e.g., “Project A”, “Project B”) for multi-project isolation.
  • View warehouse contents: The warehouse records all checked numbers along with their check time and result. You can export or clear them.

Complete Explanation of the Cross-Task Deduplication Mechanism

  1. Submit a new task: When selecting the generated numbers, choose the previously created warehouse in the “Deduplication Warehouse” option.
  2. System automatic comparison: After submission, the system matches the numbers to be checked against the existing numbers in the warehouse and removes duplicates.
  3. Write results: After the task completes, the numbers checked in this task (both successful and failed) are appended to the warehouse.
  4. Subsequent tasks: On next submission, the newly written numbers are also excluded.

This way, no matter how much overlap exists among your data sources, you only pay for the first check of each number. Effective cross-task deduplication.

Common Notes

  • Balance and Deduplication Relationship: The deduplication warehouse only excludes duplicate numbers from being checked; it does not affect already charged records. For example, if the first check cost 100, and the second check only processes 50 new numbers after exclusion, the second charge is50.
  • Duplicate Marking: Numbers in the warehouse are not automatically deleted. If you need to recheck (e.g., activity has expired), you must manually clear the warehouse or create a new one. Clearing is irreversible, please proceed with caution.
  • Multi-Warehouse Isolation: Different projects should use different warehouses to avoid interference.

Step 5: Export Screening Results and Subsequent Applications

After the screening task completes, results can be exported. Supported formats: CSV, TXT. Export fields include: number, platform status (registered/active/unregistered), gender (if detected), tgid/wsid, etc. Based on the export results, you can:

  • Import into a CRM system to start targeted outreach.
  • Use for community invitations or DM promotions.
  • Re-check inactive numbers at a later time (by generating a new task and skipping the warehouse).

Complete Checklist and Best Practices for the Generation and Deduplication Workflow

Below is a reusable operations checklist for each screening task:

  1. Generate or import numbers: Use global number generation or CSV import, ensuring quantity does not exceed 1 million.
  2. Create/select a deduplication warehouse: For first use, create a warehouse; for subsequent tasks, reuse the existing one.
  3. Configure the screening task: Select detection platform, metrics (activity, gender, etc.), and consider cross-platform combinations.
  4. Bind the deduplication warehouse: Tick or select the corresponding warehouse in task settings.
  5. Submit and wait for completion: Task progress can be monitored in real-time in the console.
  6. Export results: Choose the desired format and use fields as needed.
  7. Maintain the warehouse: Regularly check warehouse capacity; clear or archive if necessary.

Best Practice: Do I need to recheck “activity” for numbers in the warehouse weekly? Not necessarily. The warehouse stores historical results; activity changes over time. To update, create a new task and do not bind the old warehouse for rechecking. Alternatively, create a new warehouse specifically for rechecking.

Summary and Next Steps

The generation and deduplication workflow integrates number generation, multi-platform screening, and cross-task deduplication into an efficient data pipeline. By automatically deduplicating through the data warehouse, teams can avoid unnecessary charges of 30%–50% (exact savings depend on data overlap), while keeping the number pool clean for long-term management.

Take action now:


Frequently Asked Questions

Q: Does the generation and deduplication workflow support generating numbers from multiple countries simultaneously?

A: Yes. You can select multiple countries/regions in the number generation module and generate them together or separately. Generated numbers are automatically saved to your task list and can be submitted uniformly to the deduplication warehouse for subsequent screening.

Q: Does the deduplication warehouse automatically exclude all previously checked numbers?

A: Yes. As long as you select the correct deduplication warehouse before submitting a new task, the system automatically compares the numbers to be checked against existing warehouse data and excludes exact duplicates. This prevents the same number from being checked and charged multiple times.

Q: What if I want to recheck numbers that have already been checked?

A: You can create a new deduplication warehouse or manually clear the warehouse (i.e., erase all detection records) to perform rechecking. Please note that clearing is irreversible, so proceed with caution.

Q: Can the generation and deduplication workflow handle numbers exceeding one million?

A: Yes. A single screening task supports up to ~1 million numbers. For data exceeding one million, submit in batches. After each batch completes, the system writes results to the deduplication warehouse, and the next batch automatically excludes already checked numbers.

Q: Does using the deduplication warehouse consume my balance?

A: No. The deduplication warehouse itself is a free auxiliary function. Charges only occur when screening tasks are actually executed based on the number of checked entries. By enabling the deduplication warehouse, you avoid duplicate checks, thus saving balance.


Pricing information in this article is subject to the official billing page or real-time display in the console.