A must-read for base material screening: Generating number segments vs. external import, how to judge the quality of the number source?

In the entire chain of overseas customer acquisition, “base material screening” is often the first hurdle that determines cost and conversion efficiency. The so-called “base material” is the original number pool to be tested - it may be a random number generated through the platform, or it may be an inventory list purchased or crawled from the outside. The quality of the base material directly determines whether every penny you spend on screening numbers will result in a large number of effective users, or whether it will be wasted on invalid numbers.

Many teams import tens of thousands of numbers right from the start. As a result, the activation rate is less than 10%, most of the balance is burned, and there is very little valid data. This is precisely because the core of base material screening is ignored: judging the quality of the source. This article will start from the two mainstream sources of generated number segments and external import, and give 4 practical checkpoints to help you establish a reusable quality assessment process.

What is base material screening? Why does source quality determine customer acquisition cost?

“Basic screening” refers to detecting, classifying and purifying the original number pool to extract effective users that can actually be reached (such as Telegram/WhatsApp registered and active numbers). The quality of the base material source directly affects three key indicators:

• Opening rate: A high opening rate means that there are more registered users per 100 numbers, and the cost of screening numbers is lower.
• Activity: Even if it is activated, numbers that have been inactive for a long time will not be able to interact, and the conversion rate will be extremely low.
• Cost Efficiency: Each time a number is screened, fees will be deducted on a per-item basis. The higher the proportion of invalid numbers, the higher the single cost of valid numbers.

Therefore, it is far more cost-effective to spend 10 minutes to evaluate the quality of the base material before starting large-scale screening than blindly testing the entire quantity.

Generating number segments vs. external import: Comparison of the pros and cons of two base material sources

Currently, there are two main ways for overseas teams to obtain information: The platform’s built-in global number generation function (random generation/number segment generation/custom number segment) and external import (CSV/TXT file). The following is a comparison from five dimensions:

Dimensions	Generate number segments	External import
Number source	Random combination based on country code + number segment	Crawl, purchase, historical inventory, etc.
Randomness and coverage	Evenly covering all number segments, and you can specify any country/province/city	Coverage is limited by source and may be concentrated in specific number segments
Initial validity	Most of them are not registered and need to be tested to know whether they are valid	Some may have expired or duplicated, but may contain real numbers
Import difficulty	Simply select parameters to generate, no file processing	Need to organize the format, handle duplicate numbers, blank lines, etc.
Suitable for scenarios	Market understanding, first screening of new countries, coverage of large number segments	Accurate reaching of known groups, reuse of old data

Advantages and application scenarios of generating number segments

When you need to quickly cover a new market (such as Vietnam, Indonesia), or want to test the overall activity level of all number segments in a certain country, generating number segments is the most efficient choice. For example:

• In the KK-DATA console, you can select the target country (such as Brazil), specify the number range, and generate 100,000 numbers with one click.
• Then submit it to the Telegram or WhatsApp screening task to check the activation rate and activity.
• After the results come out, you can quickly determine which number segment in the country is more valuable, and then conduct a second in-depth screening of high-quality number segments.

The advantages of this method are no inventory pressure, immediate availability, wide coverage, and is suitable for the exploratory customer acquisition stage.

Externally imported data sources and quality challenges

Externally imported base materials usually come from:

• Own historical user data (such as old CRM list)
• Public numbers for web crawling
• “Accurate list” provided by agents

These data seem accurate, but in fact there are many hidden dangers:

• Duplication issue: Multiple teams may use the same source, resulting in an extremely high duplication rate.
• Expired: The number may have been canceled, changed platforms, or marked.
• Confusing format: Missing international area code, containing spaces, etc.

It is recommended to use the platform’s data deduplication warehouse (such as KK-DATA’s deduplication function) to clean up duplicate items before importing to avoid wasting balances.

How to judge the quality of your base material source? 4 key checkpoints

After getting the base material, don’t rush to screen all the sizes. First do a small-scale test (500–1000 items), focusing on the following four dimensions.

Checkpoint 1: First round opening rate

The opening rate is the “demon mirror” of the quality of the base material. The activation rate of generated number segments is generally stable within a certain range (varies by country); if the external import is lower than expected, there may be a problem with the source. For example:

• Detect 500 numbers. If the activation rate is less than 20%, it means that most of the base materials are invalid numbers. Continuing to test all numbers will only waste the balance.
• It is recommended to test with a small number of samples first and ensure the activation rate is ≥ 30% before continuing.

Checkpoint 2: Whether the activity field is rich

Opening up is not enough. Telegram and WhatsApp’s activity detection can return fields such as “recent active window” and “online frequency”, which are key to determining whether a number is worthy of a second contact. The activity distribution detected by the generated number segment is often more even; if a large number of externally imported numbers show “inactive for more than 30 days”, it means that the timeliness of this batch of base materials is poor.

Checkpoint 3: Gender/age field enrichment

If you are doing targeted marketing (for example, targeting men around 30 years old), it is important whether the base material contains gender and age information. The generated number segment can be accompanied by these fields through detection (some platforms such as KK-DATA support gender recognition, including age, etc.). If the externally imported numbers do not have supporting attributes, you need to supplement the data yourself or accept low enrichment.

Checkpoint 4: Deduplication results and duplication rate

Put the externally imported files into the repository and run it again to observe the repetition rate. If the duplication rate exceeds 50%, it means that the source has been used by a large number of teams, and the penetration value of the remaining numbers will be greatly reduced. A high repetition rate also suggests that you may have bought a “batch processing budget.”

How to use the screen number tool to quickly verify the quality of the base material?

Talking about it on paper is worse than actually running it. Taking a certain screening platform (such as KK-DATA) as an example, the verification operation only requires three steps:

1. Prepare sample: Randomly select 500-1000 numbers from the base material (or directly use the number generation function to generate the corresponding country number segment).
2. Submit screen number: Select “Telegram activation detection” or “WhatsApp activity detection” in the console to submit a sample task. The estimated cost will be displayed before submitting the task to ensure sufficient balance.
3. View results: After the task is completed (usually within a few minutes to an hour), download the results report. Focus on: opening rate, activity tag, and whether the gender/age field exists.

If the quality of the sample meets the standard (for example, Telegram opening rate > 40%, active ratio > 20%), the entire base material can be screened on a large scale. Otherwise, change the source of base material immediately.

This “Generation → Preliminary Screening → Analysis → Full Quantity” process can be seamlessly connected in KK-DATA: first generate a test number segment through the global number generation module, submit a sample task, check the dashboard results, and then decide whether to expand full detection. All operations are done within the same console.

Common misunderstandings in base material selection and pitfall prevention guide

Overseas teams often encounter pitfalls in base material selection. The following are the three most typical problems and how to deal with them.

Misunderstanding 1: Import and use without pre-judgment

• Performance: Obtain tens of thousands of numbers and directly submit all the screening numbers.
• Consequences: If the activation rate is extremely low, balances will be wasted and valid data will be scarce.
• Anti-pits: Always adhere to the principle of “sample testing” and run 500 tests first to see the effect.

Misunderstanding 2: Ignoring the risk of number expiration/invalidation

• Performance: Using the old base material from a few months ago, I consider it to be of “passable quality”.
• Consequences: Telegram users may have logged out or changed their numbers; WhatsApp accounts may be revoked.
• Anti-Pit: Pay attention to the “Recently Active” field in the detection results. If the active window of more than 60% of the numbers is “more than 30 days”, it means that the base material is too old.

Misunderstanding 3: Submit the entire amount without deduplication.

• Performance: After merging numbers collected from different channels, duplicates are not removed and the filter numbers are imported directly.
• Consequences: Repeated detection of the same number will result in repeated deductions.
• Anti-pit: Use the platform’s deduplication warehouse to run the deduplication once before submitting the screening task.

Summary: Establish your best practice process for “base material screening”

Base material screening is not a one-time task, but a process that requires long-term optimization. The following is a reusable closed loop of operations:

1. Select source: Based on the target market, decide whether to use generated number segments or external import.
2. Duplication Cleaning: Regardless of the source, first use the deduplication warehouse to eliminate duplication.
3. Small sample verification: Take 500–1000 items for activation/activity detection, and calculate key indicators (activation rate ≥ 30%, activity rate ≥ 20% as passing).
4. Quality analysis: If the sample is qualified, perform full screening; otherwise, return to step 1 to replace the base material.
5. Export and reuse: Export the filtered valid numbers and store them in your own database for direct use next time.

Remember, base material screening is the first hurdle to reduce customer acquisition costs. Spending 10 minutes analyzing the quality of your base can save you money and time several times over.

FAQ

**Q: Is the number detected by the generated number segment really valid? Will my account be blocked or empty? ** Answer: The generated number segment is a random combination based on global real number rules, not the actual user number. You need to confirm which ones have been registered through activation detection (such as Telegram registration detection). The detected valid number can be used normally and will not be blocked - because the detection itself only determines whether the number is registered on the platform and does not send a message to the user. Empty accounts (not registered) will be automatically filtered and no fees will be deducted.

**Q: How to quickly screen out high-quality numbers from externally imported base materials? ** Answer: Take it in two steps. First, use the global number generation function of the platform to generate a small range of number segments in the target country as a “control sample” to test its activation rate; then, submit the externally imported sample for testing to compare the activation rate and activity of the two. If the externally imported indicators are significantly lower than the generated number range, it indicates that the source quality is poor. Also, be sure to do the deduplication first.

**Q: What is the effect of mixing generated number segments and external import? ** Answer: This is an advanced strategy. You can first generate number segments to screen out high-quality numbers as a “basic pool”, then merge them with externally imported low-repetition files and submit the screening numbers together. Mixing can achieve both coverage and accuracy, but attention needs to be paid to deduplication and task splitting.

**Q: After the number screening is completed, how to evaluate the renewal value of the valid number? ** Answer: Record the “number of activated numbers” and “spent balance” of each number screening task, and calculate the cost of a single valid number. If costs continue to rise, it means that the quality of the base material is declining and the source needs to be re-evaluated. The KK-DATA console will display the details of each task for easy tracking.

**Q: How long does it usually take to screen base materials? ** Answer: Sample detection usually takes a few minutes to dozens of minutes to complete (depending on the task size and platform load). A full batch of millions of numbers may take several hours. It is recommended to set aside sufficient time, turn on task notifications (such as notifications through Telegram), and process the results in a timely manner after receiving completion reminders.

---

Want to start optimizing your base screening process right away? 👉 Log in to the console to start screening numbers or contact customer service through two-way https://t.me/kkdata_robot to get one-on-one guidance. For more documentation and usage tutorials, please visit https://docs.kkdata.cc/.