---
title: "Behind the Scenes: Our 4-Stage Verification Pipeline That Cuts Indian Lead Bounce Rates Under 3%"
url: https://leadkosh.com/blog/behind-the-scenes-verifying-indian-phone-numbers-emails
canonical: https://leadkosh.com/blog/behind-the-scenes-verifying-indian-phone-numbers-emails
category: Verification
tags: [Data Verification, Email Deliverability, WhatsApp API, Data Engineering, Lead Quality]
author: Leadkosh Data Engineering Team
published: 2026-02-11
updated: 2026-02-11
publisher: Leadkosh
---

# Behind the Scenes: Our 4-Stage Verification Pipeline That Cuts Indian Lead Bounce Rates Under 3%

> Verifying contact data in India is 5x harder than in the US or Europe. Learn how our engineering team built a resilient 4-stage pipeline combining deep SMTP handshakes, MX cluster latency testing, WhatsApp presence telemetry, and GSTIN matching.

**Key takeaways**

- Traditional email verifiers fail in India because over 40% of Indian business domains use local hosting or misconfigured catch-all DNS records.
- Leadkosh uses a 4-stage validation pipeline: deep SMTP socket handshakes, dynamic catch-all heuristics, WhatsApp business protocol checks, and statutory registry validation.
- Contacts with unverified status or high bounce probability are quarantined and never sold to users.
- If a user encounters a hard bounce or disconnected phone number, Leadkosh automatically replaces the credit without bureaucratic hassle.

In the world of sales intelligence, there is dirty data, and then there is *raw Indian scraped data*.

If you download a raw list of businesses from typical online directories in India, you are looking at an entropy rate of over 45% per year:

- Phone numbers churn rapidly due to dual-SIM usage, employee departures, and prepaid balance expirations.
- Corporate emails are frequently hosted on shared cPanel servers with non-standard mail transfer agents (MTAs) that lie about mailbox existence.
- Catch-all mailboxes accept all emails on the surface, only to silently drop them into a black hole 10 seconds later.

Here is the technical breakdown of how Leadkosh’s verification engine works under the hood to deliver sub-3% bounce rates.

## 1. The Indian Contact Data Entropy Problem

Most verification APIs (ZeroBounce, NeverBounce) were calibrated against Google Workspace and Microsoft 365 servers. In India, while top startups use Workspace, hundreds of thousands of established MSMEs, distributors, and manufacturers run their email on:

- Rediffmail Pro
- Netcore / Netmagic enterprise mail
- Hostinger / GoDaddy shared cPanel Exim servers
- Zoho Mail India clusters (`smtppro.zoho.in`)

Standard Western tools misinterpret the bounce codes returned by these servers, resulting in massive false-positives (marking good leads as invalid) and false-negatives (marking dead mailboxes as deliverable).

## 2. Stage 1: Deep SMTP VRFY & RCPT TO Handshake

Before any contact enters the Leadkosh searchable index, our verification workers perform a direct socket-level handshake with the recipient’s authoritative MX server:

 > CONNECT mx1.zoho.in:25

 < 220 mx.zohomail.com Mail Server ESMTP

 > EHLO verify.leadkosh.com

 < 250-mx.zohomail.com Hello verify.leadkosh.com

 > MAIL FROM:<probe@verify.leadkosh.com>

 < 250 Sender <probe@verify.leadkosh.com> OK

 > RCPT TO:<rajesh@apexindustrial.in>

 < 250 Recipient <rajesh@apexindustrial.in> OK

 > QUIT

Notice that we terminate the connection with `QUIT` immediately after receiving the `250 OK` status code. We never actually send an unsolicited test email to the user's inbox.

## 3. Stage 2: Dynamic Catch-All Detection

Many mail servers return `250 OK` for *every* email address, even `randomnonexistentstring987@company.com`. This is called a Catch-All configuration.

If a server is catch-all, standard SMTP verification is useless. Leadkosh generates a high-entropy pseudo-random hash probe. If the server accepts the fake probe, the domain is flagged as `catch_all`. We then run secondary heuristic checks (domain age, employee MX traffic patterns, and SPF/DKIM alignment) before deciding whether to verify the contact.

## 4. Stage 3: Telecom & WhatsApp Telemetry

Phone verification in India requires an entirely different playbook:

1. **National Numbering Plan (NNP) Validation:** We ensure the 10-digit number conforms to the Department of Telecommunications (DoT) allocation blocks for mobile carriers (Airtel, Jio, Vi, BSNL).
2. **WhatsApp Business Protocol Handshake:** We check whether the phone number has an active WhatsApp cryptographic public key registered on Meta’s infrastructure. If it doesn't have WhatsApp enabled, it is flagged so agencies don't waste time on WhatsApp outreach.

## 5. Stage 4: GSTIN & Public Registry Triangulation

Finally, we corroborate the contact with statutory data:

- Does the GSTIN have a status of **Active**? (Cancelled or suspended GSTINs are heavily penalized in our search ranking).
- Does the registered trade name on the GST portal match the company name on MCA?
- Is the city and pin code geographically aligned with the phone circle?

## 6. Our Honest Bounce-Replacement Policy

No verification pipeline in the world can be 100.00% perfect. People shut down companies overnight, telecom towers drop calls, and mailboxes fill up to their quota limit.

That’s why our policy is straightforward: **If you unlock a contact on Leadkosh and the email hard-bounces or the phone number is dead, you get an automatic replacement credit.** Simply click "Report Bounce" in your unlocks dashboard. We investigate the record, suppress it for all other users, and refund your credit immediately.

---
Source: https://leadkosh.com/blog/behind-the-scenes-verifying-indian-phone-numbers-emails · By Leadkosh Data Engineering Team (Data Pipelines & Compliance) · Updated 2026-02-11
