A Must-Read for Number Screening Teams: Funnel Metrics from Number Quality to ROI Tracking Operations Dashboard Guide

If you are responsible for overseas customer acquisition through Telegram or WhatsApp number screening, you must have encountered scenarios like these: the same batch of numbers shows fluctuating connectivity rates when tested at different times; after spending a significant budget to screen out tens of thousands of “valid numbers,” the conversion rate from direct message campaigns is very low; your team relies on “gut feeling” to judge which channel’s numbers are of high quality, lacking reusable measurement standards.

Behind these issues, the core problem is the absence of a number screening team metrics system. Without a funnel, you cannot identify bottlenecks; without a funnel, you cannot quantify the real output of each screening investment. This article starts from the four key levels of the number quality funnel, provides recommendations for building an operations dashboard and ROI tracking formulas, helping you shift from experience-driven to data-driven operations.

The Four Key Levels of the Number Quality Funnel

Numbers go through multiple layers of filtering from “raw numbers” to “marketable leads.” Each layer has attrition, and each attrition point corresponds to an optimizable step. The funnel model is as follows:

Funnel Level	Business Meaning	Common Causes of Attrition
Coverage Volume	Total number of numbers entering the screening system	Uncontrollable data sources, duplicate numbers
Connectivity Rate	Proportion of numbers that have activated accounts on the target platform	Expired numbers, platform account bans, invalid number ranges
Activity Rate	Proportion of numbers with recent online behavior	Abandoned numbers, unreasonable activity window settings
Gender Identification Rate	Proportion of numbers whose gender can be identified	Unidentifiable avatars, no public information

These four levels form a complete number quality funnel and are the fundamental dimensions for building an operations dashboard.

Coverage Volume: The “Entry Traffic” of a Screening Task

Coverage volume is the total number of numbers entering the screening system. It can come from manually uploaded CSVs, random generators, or external purchases. Teams should track coverage volume by source channel, because the quality of numbers from different channels varies greatly. For example, numbers generated by a specific number range generator may have a higher connectivity rate than random generation but a lower activity rate.

Recommendation: When exporting numbers, tag each number with a source label (e.g., source=generator_US, source=CSV_october). This allows you to compare the conversion efficiency of each channel in subsequent funnels, providing a basis for budget allocation.

Connectivity Rate and Activity Rate: The “Core Quality” of Qualifying Numbers

Connectivity rate is the proportion of numbers that have active accounts on the target platform (Telegram / WhatsApp / iMessage, etc.). Activity rate further filters out numbers with recent online behavior. Both together determine the marketing value of a number.

It is recommended to break down the “social platform effective detection results” into four categories and track their proportions on the dashboard:

• Registered: The number has an account on the platform, but activity status is unknown.
• Active: Registered and recently active (e.g., 7 days / 15 days / 30 days).
• Invalid: Not registered or has been deactivated.
• Unknown: Detection failed or result uncertain.

Practical Advice: If your goal is bulk direct messaging, prioritize numbers in the “Active” tier; if you only need to join groups or verify identity, “Registered” is sufficient. The “Activity” filter parameters can be flexibly adjusted, but each change should be recorded because the activity window (e.g., 7 days vs. 30 days) significantly affects the activity rate.

Gender Identification Rate: “Added Value” for Targeted Scenarios

For scenarios that require targeted push (e.g., women’s skincare, men’s games), the gender identification rate can supplement funnel information. However, note that gender identification relies on avatar information, has limited coverage, and accuracy is not 100%. It is recommended to treat it as a secondary metric, not the sole criterion for filtering. For example, you can first filter high-potential numbers by activity rate, then count the gender ratio in the results as a reference for content optimization.

Pre-Screening: Why You Cannot Skip “Number Cleaning” Before Direct Promotion

Some teams, in a rush to acquire customers, skip number screening and push directly, facing three major risks:

1. Charged for Invalid Numbers: Direct messaging platforms typically charge based on volume sent (or per message). Invalid numbers also consume budget.
2. Increased Complaint Rate: A large number of non-existent numbers trigger error return codes, and the platform may limit your sending capacity.
3. Account Ban Risk: A high proportion of invalid numbers can trigger anti-spam policies, leading to the banning of promotional accounts.

Pre-screening is equivalent to performing a quality check at the “traffic entry.” Using KK-DATA as an example, you can import raw numbers, run a multi-platform screening task (Telegram, WhatsApp, iMessage, RCS, etc.), and the system automatically returns connectivity, activity, and gender identification results for each number on each platform. After screening, export the valid numbers before investing in promotion.

Operations Dashboard Setup Recommendations and ROI Tracking Formula

Now that you understand the funnel, it’s time to implement an actionable dashboard. For a number screening team, a minimum viable dashboard should include the following indicators.

5 Essential Dashboard Metrics

Metric	Description	Why It’s Necessary
Total Tasks	Number of screening tasks submitted this period	Reflects team workload, can be compared with output
Number Coverage	Total numbers entering the screening system	Funnel entry, determines conversion ceiling
Average Connectivity Rate	Weighted average of valid number proportions across all tasks	Quickly assess overall number quality trend
High Activity Number Count	Valid numbers within the active window (e.g., 7 days)	Actually reachable marketing leads
Cost Per Lead	Per-person cost of valid leads (see formula below)	Directly related to ROI, determines budget allocation

Use UTM Parameters to Track the Full Screened-to-Conversion Chain

Many teams neglect downstream attribution. It is recommended to preset UTM tracking parameters for each number before exporting. For example, embed batch_id, source, and campaign into the remark field of the export column. Later, when sending promotional messages, links carry UTM parameters, allowing you to see which screening batches generated actual registrations or purchases via analytics tools (e.g., Google Analytics / custom statistics).

Example UTM: utm_source=kkdata&utm_medium=telegram&utm_campaign=wechat_oct&utm_content=batch_20241001

ROI Tracking Formula

Cost Per Lead = (Total Screening Fees This Period + Total Recharge Fees This Period) ÷ Total Valid Leads This Period

Where “valid lead” is defined as a number that ultimately completes the expected conversion action (e.g., clicking a link, registering, joining a group). If you cannot track to final conversion, use “valid numbers” instead to get “cost per valid number.”

Full ROI: (Final Revenue – Total Cost) ÷ Total Cost × 100%

Total Cost = Screening Fees + Recharge Fees + Message Sending Fees + Labor Costs

Reducing Duplicate Waste: Statistical Significance of Cross-Task Deduplication

One of the most overlooked wastes during screening is repeated detection. The same batch of numbers may be submitted multiple times in different tasks, leading to multiple charges for the same number. This not only inflates costs but also contaminates funnel metrics (inflated coverage, distorted connectivity rates).

Solution: Enable a cross-task deduplication repository. For example, KK-DATA provides a built-in deduplication repository that automatically removes numbers already detected when you upload. For your team dashboard, consider adding a “Deduplication Savings” metric:

Deduplication Savings = (Number of Deduplicated Entries × Unit Price of Corresponding Platform)

Publish the deduplication count monthly so team members can visually see the value of this action. If deduplication accounts for more than 10%, it indicates significant duplication in number sourcing or generation, requiring investigation at the source.

Batch Processing Efficiency Metrics: Task Completion Time and Concurrency

Team size affects focus on batch processing efficiency. For small studios (1–2 people), the most critical metric is “single task completion time”—from task submission to receiving a notification. If detecting 50,000 numbers takes 3 hours, you can complete multiple batches within a day without waiting overnight.

For medium-to-large teams, you also need to focus on “concurrency”: how many tasks can be submitted simultaneously. This affects overall throughput. It is recommended to record the weekly average single-task duration and compare it with previous weeks. If the duration gradually increases, it may indicate increased system pressure or lower data quality leading to more retries.

Note: Don’t just look at price. Some low-cost platforms may have queue delays due to insufficient processing capacity, making actual completion speeds much slower than advertised. When choosing, prioritize “speed” and “stability.”

From “Gut Feeling” to “Quantifiable”: 3 Actionable Suggestions for Data-Driven Screening

1. Weekly Core Funnel Data Statistics
   Use Google Sheets or Excel to create a simple weekly report template. Record these fields: number of tasks, total coverage, average connectivity rate, high-activity number count, cost per lead, deduplication savings. After each update, note the reasons for significant changes (e.g., “switched to a new number generator,” “recharge channel adjusted”).

2. Monthly Comparison of Channel Conversion Rates
   Use number source (generator_A, CSV_purchase, customer service import, etc.) as dimensions, and calculate the conversion rate from coverage to high-activity numbers for each channel. Channels with low conversion rates should be considered for replacement or optimization.

3. Reallocate Budget by ROI Ranking
   Calculate the cost per lead for each channel. Prioritize investing more budget in channels with the lowest cost. If a channel has a high cost but the customer lifetime value after conversion is high, keep it separately. Regularly review the situation and channel budget to the most efficient spots.

Frequently Asked Questions

Q: What is the minimum number of people a screening team needs to effectively track these funnel metrics?
A: One person can do it, but it is recommended to set a fixed weekly routine of “export dashboard screenshot + compare with the same period last week.” If you have two or more people, you can divide the work: one person responsible for task submission and data export, the other for statistics and dashboard maintenance.

Q: Which dashboard metric is most easily overlooked?
A: Deduplication savings from cross-task deduplication. Most teams only focus on new number coverage, not calculating the hidden losses from repeated charges. After enabling the deduplication repository, it is recommended to publish the number of deduplicated entries and savings monthly.

Q: If the activity rate of the same batch of numbers drops for two consecutive weeks, what should be done?
A: First, check the activity window parameters used during screening (e.g., changing from 30 days to 7 days will lower activity rate). If the parameters are consistent, the quality of that batch has deteriorated. Suspend its use and replace the source channel; also check if old numbers are being retested due to data duplication.

Q: How is “cost per lead” calculated in the funnel?
A: Cost Per Lead = (Total Screening Fees This Period + Total Recharge Fees This Period) ÷ Total Valid Leads This Period. Valid leads refer to numbers that ultimately successfully join a group/complete an interaction/open a link, not just screened active numbers. If you don’t track final conversion, use “cost per valid number” as an alternative.

Q: Do I have to do these statistics within the KK-DATA platform?
A: Not necessarily. The KK-DATA console provides task lists and balance change records, but for a complete funnel dashboard, it is recommended to build your own (Google Sheets / Airtable / Excel) and manually update core metrics weekly. For automated integration, refer to the documentation (based on the console’s available capabilities).

---

Next Steps:

• Log in to the KK-DATA Console to experience the number screening pipeline
• View the complete tutorial and operation templates
• For dashboard setup questions, contact customer support via Telegram: @kkdata_robot