Agent operation and studio tg filter delivery SOP: full process guide from number screening to acceptance and re-screening

When agency operations and studios undertake tasks such as Telegram community operations, private message promotion, and TG followers, tg filtering (Telegram number filtering) is often the most frequent and error-prone link in the delivery chain. The customer wants “100,000 active users”, but the team may hand over half of the unactivated numbers after one operation; and the customer requires re-screening after the second quality inspection, and the cost of rework increases sharply. A set of standardized tg filtering SOP can unify team actions, reduce customer complaints, improve data reuse rate, and make every test result stand up to acceptance.

This article is based on the number screening practice of the KK-DATA platform, from number preparation, task configuration, execution export, acceptance re-screening to platform selection, and provides a direct operation guide that can be implemented.

---

Why does the agency operation studio need standardized tg filtering SOP?

Agent operation teams usually face three big problems: lack of consistent actions in multi-person collaboration (employee A checked active detection, employee B only checked activation), ambiguous customer acceptance criteria (what kind of number counts as “active”? 7 days or 30 days?), repeated testing wastes balance (the same batch of numbers are repeatedly submitted by different projects). The SOP is like an operation manual. It clearly states “what detection type to choose”, “how to set the active window” and “how to read the result field” so that new members can get started quickly.

Direct benefits after normalization:

• Reduce customer disputes caused by missing fields;
• Reduce inspection costs by 30%-50% through deduplication warehouse;
• Problems can be quickly located during acceptance (poor quality of number source/wrong detection configuration).

---

What is tg filtering? Core detection types and applicable scenarios

tg filtering (Telegram filtering) refers to detecting Telegram registration status, active period, gender/age and other characteristics through an automated platform to filter out a collection of numbers that meet the purpose of promotion. KK-DATA provides three types of core detection, corresponding to different business scenarios:

Detection type	Return fields	Typical uses
tg activation	Whether Telegram has been registered (activation status)	Verify whether the number is reachable in batches to avoid sending it to unregistered users
tg active	Last active time (such as within 7 days/30 days)	Private message marketing requires “recently online” users, and the activity rate is ≥70% to achieve conversion
tg Gender	Gender, age range (including fields such as about 30 years old)	Accurate stratification of vertical products (such as men’s skin care products, maternal and infant products)

Note: Gender/age data is based on Telegram public information and model inference, suitable for group stratification, and does not have personal identity verification level accuracy**. The age field reads “approximately 30 years old” but not the exact date of birth.

For agent operation delivery, it is usually required to choose two or all three of “tg activation + tg active (7 days) + tg gender”, depending on the customer’s budget.

---

Preparation for tg filtering: number source and task configuration

Number generation and import

Agency operation teams often purchase number lists from third parties, but many of the numbers themselves are invalid (unallocated, empty numbers). In KK-DATA you can:

• Global number generation: Covering 240+ countries/regions, generation is completely free, no need to worry about invalid numbers;
• Number segment generation: Batch generation based on number segment (such as +86 138xxxx);
• CSV Import: Upload your own number list, and the system will automatically recognize the format.

Best Practice: Prioritize the use of numbers generated by the platform, because the generation rules exclude invalid areas of number segments, which can improve subsequent activation rates. After generation, you can directly enter the screening task without additional deduplication (the platform has a built-in data deduplication warehouse).

Detection type selection and estimation

When creating a new screening task in the console, you can check multiple tests at the same time. For example:

• Check tg activated + tg active (7 days) + tg gender, the system will detect and return the three fields in sequence.
• Note: Active detection requires a specified time window (7 days / 30 days); additional charges will be deducted for gender detection. Please see the real-time price of the console for unit price details.

The system will display the estimated fee before submission, and the studio can decide whether to check gender based on the client’s budget. If the client has a limited budget, it is recommended to give priority to staying active, and gender can be discarded as appropriate.

Balance recharge check

KK-DATA uses USDT (TRC20) for anonymous recharge, with a minimum of about 50 USDT and no subscription package. Be sure to check whether the balance is sufficient before submitting the task, otherwise the task cannot be submitted. You can check the real-time deductible balance on the “Account Balance” page.

---

tg filtering execution process: from task submission to result export

Task submission and message notification

1. Log in to the console → select “Screen Number Task” → click “New Task”;
2. Import number source (generate or upload);
3. Check the detection type (tg activated / active / gender) and set the active window;
4. Optional: Fill in the task notes (such as “Project A - European and American male 35 years old+”) to facilitate subsequent tracing;
5. Submit the task.

Suggestion: Bind a Telegram account in “Notification Settings”. After the task is completed, the results will be automatically pushed through https://t.me/kkdata_robot to avoid frequently refreshing the console.

Result export and field interpretation

After the task is completed (usually tens of seconds to minutes), enter the task details page and click “Export” to download the CSV or TXT file. Typical export fields:

Field	Meaning
phone	Original Number
tg_open	Whether to activate Telegram (1=activated, 0=not activated)
tg_active	Active duration (7d / 30d / 90d, etc.)
tg_sex	Gender (male / female / unknown)
tg_age	Age range (e.g. 25-34)
tgid	Telegram internal ID (used for subsequent API operations)

Important: The age field reflects group characteristics and should not be used as an exact value for an individual’s age.

Use of data deduplication warehouse

KK-DATA has built-in data deduplication warehouse: when importing new numbers, the system automatically compares the numbers that have been detected in history and skips the ones that have been processed. The studio can mark each delivered number as “processed”, and the next time a similar customer reuses the original number, the fee will not be deducted again. It is recommended to manually mark or set automatic deduplication rules after each export.

---

tg filtering data acceptance criteria and re-screening rules

Acceptance field description

When delivering to the client, it is recommended to provide the following summary of metrics, attached to the data file:

• Activation rate: tg number of activated numbers ÷ total number of submitted numbers. The normal value is ≥85% (number segment generation can reach 90%+); if it is lower than 85%, it means that the number source may contain too many unregistered number segments.
• Activity rate: Number of active numbers in 7 days ÷ Number of activated numbers. Generally, ≥50% is required to be considered “high quality”; if customers are conducting private message marketing, it is recommended that the activity rate be ≥70%.
• Gender Clarity Rate: The number of numbers whose gender is successfully identified ÷ the number of activated numbers. Typically 60%−80% identifiable.
• tgid integrity: Whether the exported tgid has 100% value. If too many are missing, it may affect subsequent API calls.

Rescreening trigger conditions

When the data quality inspection finds one of the following situations, delivery should be stopped immediately and the re-screening process should be entered:

1. The activation rate is less than 60% (after excluding the number segment problem);
2. The activity rate suddenly drops below 10% (possibly due to the detection window being too narrow or the number source invalid);
3. More than 90% of the gender field is unknown (platform model or configuration problem);
4. Customers reported that a certain field was seriously inconsistent with expectations (for example, male results were requested but mostly female).

Suggestions for re-screening operations

1. Change number source: Regenerate a new number (240+ random countries), or use number segments to generate more detailed country number segments;
2. Reduce the activity window: Change the activity period from 30 days to 7 days to increase the activity rate;
3. Additional Gender Detection: If only activation was tested in the first round but gender was not tested, you can only check “tg gender” during re-screening and cooperate with the deduplication warehouse to avoid repeated billing;
4. Segmented delivery: Small batch (1000 pieces) results will be sent to customers for confirmation first, and then mass production will be carried out.

---

Things to note when choosing the tg filtering platform

When choosing a platform for an agency operation studio, it is recommended to focus on:

• Data Security: Whether the platform registers anonymously (such as USDT recharge, no identity information is required); whether the results can be exported by themselves and the platform does not retain long-term copies.
• Billing Transparency: Whether billing is by item and there is no subscription package; whether there is a cost estimate before submitting the task (beware of hidden consumption).
• Function Stability: Whether it supports multiple detection items at the same time; whether the active window is flexible (7 days / 30 days / customized); whether the exported fields are complete.
• Customer service response: Is there any instant messaging tool (such as Telegram robot), and can problems be solved quickly?

KK-DATA has clear designs in the above four points: users recharge anonymously through USDT, the console displays the unit price in real time, supports 3 detections to be checked at the same time, and provides https://t.me/kkdata_robot 7×24 hours automatic customer service and manual customer service. You can log in to the console to see the full list of features.

---

FAQ

**Q: Can tg filter detect activity and gender at the same time? ** Answer: Yes. When submitting a task in KK-DATA, you can check both the “tg active” and “tg gender” detections, and the system will return the corresponding fields respectively. Note that active testing requires a specified time window (such as within 7 days), and the two tests will be charged separately.

**Q: Is the gender and age data filtered by tg accurate? ** Answer: Gender and age data are based on Telegram public information and machine learning model inference. They are suitable for group stratification (such as screening users around 30 years old) and do not have ID card-level accuracy. Do not use it in personal identity verification scenarios.

**Q: How does the studio avoid detecting the same number repeatedly? ** Answer: Using KK-DATA’s “Data Deduplication Warehouse” function, the system automatically removes numbers that have been detected in historical tasks when importing new numbers to avoid wasting balances. It is recommended to mark processed numbers after each export.

**Q: After the tg filtering task is completed, can the results be encrypted and transmitted to the customer? ** Answer: Currently, the platform does not support encrypted packaging, but it is recommended to export CSV and use third-party tools (such as Zip encryption) before sending. The data exported by the console is only available for 7 days, please download it in time.

**Q: What is the minimum task size for tg filtering? ** Answer: A single task supports up to about 1 million numbers. There is no minimum quantity limit, even a few hundred numbers can be submitted, and fees will be deducted based on the actual number of tests.

---

If you are looking for a stable, pay-as-you-go tg filtering solution for your team, welcome to try KK-DATA for free. No need to subscribe to the package, just use it directly after recharging.
👉Log in to the console to start screening numbers Two-way customer service contact: https://t.me/kkdata_robot Detailed documentation: https://docs.kkdata.cc/