How to Optimize Number Screening with Activity Window A/B Testing (7-Day/15-Day/30-Day Comparison Guide)

In Telegram customer acquisition scenarios, do you often face this dilemma: should you use a 7-day activity window to filter “recently online active users,” or a 30-day window to reach more potential contacts? Choose the wrong window, and you either waste budget on a bunch of “dead” numbers or miss a huge number of real users due to insufficient quantity. Activity window A/B testing helps you answer this question with data. This article will guide you step by step on how to use KK-DATA’s number filtering feature to design a complete comparative experiment using 7-day and 30-day windows, and find the optimal activity window setting for your business scenario.

Why Do You Need Activity Window A/B Testing?

Many teams choose an activity window based on experience or simply go with the maximum “31-day” option. While this seems convenient, it can significantly reduce your marketing efficiency.

How Activity Windows Affect Your Marketing Results

The activity window refers to whether a number has any Telegram online or operational behavior (such as sending messages, receiving verification codes, etc.) within a specified number of days. Different windows directly determine what kind of users you get:

• 7-day activity window: The filtered numbers have the highest activity level, most likely real users who have been actively using Telegram recently. However, the pass rate is the lowest—many phone numbers may have registered Telegram but not logged in for over a week. If you’re promoting time-sensitive marketing content that requires quick responses (e.g., limited-time event invitations, flash sale notifications), the 7-day window typically yields the highest reply and conversion rates.
• 15-day activity window: A middle-ground option. Activity levels fall between 7-day and 30-day windows, but the number of qualifying numbers is significantly higher than the 7-day window. Suitable for general community invitations or product promotions, balancing activity assurance with broader reach.
• 30-day activity window: Yields the highest number of qualifying numbers, but may include many “low-frequency users”—people who only occasionally open Telegram to receive verification codes and are less responsive to marketing messages. If your goal is large-scale volume and you don’t mind some user churn, this window is worth trying.

What A/B Testing Can Help You Answer

• Which activity window yields the highest DM read rate or group join rate?
• Which window gives you the lowest total cost to acquire an effective user (filtering cost ÷ number of reachable active users)?
• Can you find a window that balances reach quantity with sufficient activity quality?

These answers cannot be guessed; they can only be verified through parallel comparative testing.

What Is an Activity Window? Meaning and Suitable Scenarios for 7-Day/15-Day/30-Day

The Telegram activity window is a feature in KK-DATA that, based on tg valid detection, further filters numbers that have shown “active behavior within a certain time period.” Detection logic: The platform attempts to detect whether the number has generated any visible online or operational markers at the Telegram protocol level within the last N days.

Window	Activity Criteria	Pass Rate (Relative)	Typical Use Cases
7 days	Any Telegram activity within the last 7 days	Low (highest quality)	High-time-sensitivity marketing, flash sales, urgent notifications, high-ticket product DMs
15 days	Any activity within the last 15 days	Medium	Regular community invitations, product promotions, daily outreach
30 days	Any activity within the last 30 days	High (largest quantity)	Mass list warm-up, brand exposure, large-scale viral campaigns

Note: The activity window applies only to numbers that have passed tg valid (i.e., registration check). If a number is not registered on Telegram, it will not enter the activity detection phase.

Complete Steps to Design Activity Window A/B Testing in KK-DATA

Below, we use “testing 7-day vs. 30-day activity windows on the same number source” as an example to demonstrate the full experimental process.

Step 1: Prepare the Same Original Number Pool

The core principle of A/B testing is single variable control. Therefore, both test groups must use an identical source number pool. You can prepare it in either of the following ways:

• Global Number Generation: In the console, use the “Global Number Generation” feature, select target countries (e.g., USA, Indonesia, Brazil), and generate 20,000–50,000 numbers. You can choose to generate only active number segments or all segments.
• Own CSV Import: If you already have a list of phone numbers, upload a CSV or TXT file directly. Ensure the number format is correct (with international code, e.g., +1xxxxxxxxxx).

Step 2: Create Two Number Filtering Tasks (7-Day vs. 30-Day)

Go to the KK-DATA App Console, click “Create Task” → “TELEGRAM Number Filtering”.

1. Task A (7-day window):

   • Detection types: check tg activated + tg valid + tg active (select 7 days as the activity window)
   • Optional addition: tgid export (recommended for subsequent tracking)
   • Import the number pool prepared in Step 1 (e.g., 5000 numbers)
   • Confirm the estimated task cost and submit

2. Task B (30-day window):

   • Create another task with detection types exactly the same as Task A (tg activated + tg valid + tg active)
   • Change only the “activity window” to 30 days
   • Import another subset from the same source as Task A (e.g., also 5000 numbers, ensuring similar number segment distribution)
   • Confirm the cost and submit

Key reminder: The two tasks should be submitted almost simultaneously, with no more than a 5-minute gap. If a day passes, the status of the original numbers may change (e.g., some users log in/out), leading to unfair testing. Also, ensure the number counts for both tasks are similar to avoid quantity differences affecting comparison.

Step 3: Export Results and Compare Metrics

After tasks complete, download the filtered results (CSV/TXT format) for each task from the console. Focus on comparing the following dimensions:

1. Active pass count: Total numbers marked as active.
2. Active pass rate: Active count / Total original numbers × 100%. A higher percentage means more active users for the same budget (but actual conversion depends on subsequent data).
3. tgid export count (if checked): This is key for tracking conversions. You can import tgids into other tools or combine with your CRM data to analyze which window yields higher user reply rates.

Example: Task A (7-day) filters 800 active numbers out of 5000 (16%); Task B (30-day) filters 2000 active numbers (40%). However, in actual follow-up DMs, Task A’s user reply rate might be 2.5 times higher than Task B’s. For high-conversion-focused businesses, the 7-day window has higher “unit cost efficiency.”

How to Analyze A/B Test Results and Find the Optimal Activity Window

After obtaining data, don’t just look at surface quantities. Perform a systematic analysis in three steps:

1. Calculate “Effective Reach Cost”:

   • Cost = total task fee (charged per number; see real-time price in the console)
   • Effective reach cost = total fee ÷ number of actually reachable active numbers
   • If Task A produces only 800 active numbers but costs 0.01 per number (hypothetical), total cost8; Task B produces 2000 numbers, total cost 20. Then Task A’s cost efficiency is actually higher (0.01/number vs. $0.01/number), but Task B wins on quantity. The final decision depends on your budget and business goals.

2. Combine with Subsequent Conversion Data:

   • Use the exported tgids or phone numbers from both tasks for DMs or community invitations. Track reply rates, join rates, or click-through rates within 24 hours.
   • If Task A’s user reply rate is 3 times higher than Task B’s, then even if Task A has a smaller pass count, it may be the better option.

3. Decision Matrix:

   • High-conversion, high-ticket businesses (e.g., online courses, consulting, premium products): Choose the 7-day window, focusing on reply rates.
   • Mass acquisition, brand exposure: Choose the 30-day window, sacrificing some quality for reach volume.
   • Balanced businesses: First test the 15-day window, then compare with 7-day or 30-day.

Best Practices and Precautions for Activity Window A/B Testing

• Change only one variable at a time: Keep all other options (detection types, number source, export format) consistent. If you change the country or number segment simultaneously, you cannot attribute the results.
• Run multiple tests and average the results: A single test may involve randomness (especially with small number sets). Perform 2–3 rounds of testing with the same number pool and average the activity rates.
• Allocate budget carefully for testing: Small-scale tests consume very little balance. Start with free-generated numbers or upload a few hundred for low-cost testing, then commit full budget after confirming the best window.
• Avoid cross-contamination: Ensure the numbers imported into the two tasks do not overlap. If duplicates exist in the pool, the deduplication warehouse may block them, causing biased results. It is recommended to deduplicate the original number pool first.
• Record Task IDs: In the console’s “Task History,” you can find a unique ID for each task. Save them for future review.

Summary: Double Your Filtering Efficiency with A/B Testing

Choosing an activity window is not a gut feeling—it can be optimized through A/B testing. From 7-day to 15-day to 30-day, each window has unique value and cost characteristics. Using KK-DATA’s multi-task parallel capability, you can create two comparative tasks in 5 minutes today, replace experience with data, and spend your filtering budget wisely.

Log in to the KK-DATA App Console now to start your first activity window A/B test, or check the documentation for more filtering details. If you have any questions, contact customer service @kkdata_robot, and we will help you design more precise acquisition strategies.

Frequently Asked Questions

Q: How much difference is there between 7-day and 30-day activity window filtering results?
A: Typically, the 7-day window has higher activity rates (20%–30%) but fewer qualifying numbers; the 30-day window can yield 2–5 times more qualifying numbers, but some users may be long-term offline. The exact difference varies by target country and number pool, so A/B testing with your own data is recommended.

Q: How many numbers are needed for activity window A/B testing?
A: At least 5000 valid numbers per group is recommended; 10,000 or more is better for stable results. The console supports up to about 1 million numbers per task. For small-scale tests, use the random generation feature to obtain numbers.

Q: Can I test 7-day, 15-day, and 30-day windows simultaneously?
A: Yes, create three tasks. Ensure each task uses a subset from the same number source and submit them close in time. This allows direct comparison of activity rates and pass quantities across the three windows.

Q: After testing, how do I determine which window is best for my business?
A: Combine two indicators: active pass rate + subsequent actual conversion rate (e.g., DM reply rate, group join rate). If you prioritize quick responses, choose the window with slightly lower pass rate but highest activity; if you prioritize reach quantity, choose the window with larger pass count. Make the decision based on final ROI.

Q: Does KK-DATA support custom activity windows (e.g., 10 days)?
A: Currently, the console provides preset options like 7/15/30 days. Specific options depend on the real-time interface. If you need a custom range, contact customer service @kkdata_robot to confirm availability.