# Bivas Kumar Roy

Bivas Kumar Roy helps businesses and agencies make their website, advertising, and analytics data more reliable. Their work focuses on conversion tracking and audits across GA4, Google Ads, Meta Pixel and CAPI, Google Tag Manager, and Consent Mode.

- Profile: https://b2binfluencer.marketing/creators/bivas-kumar-roy/
- LinkedIn: https://www.linkedin.com/in/bivasroy-conversiontracking-meta-ads-gads
- LinkedIn followers: 3,186
- Topics: ecommerce, paid-ads, marketing-attribution, conversion-tracking
- Based on 21 dated posts, latest 2026-10-09

## Summary

Bivas Kumar Roy is most useful for diagnosing conversion tracking and attribution discrepancies across GA4, Google Ads, Meta and Consent Mode.

- Best for: A B2B ecommerce SaaS advising merchants on Meta CAPI, GA4 and Consent Mode, plus technical content.
- Not for: A consumer retail brand seeking high reach, broad audience video posts and engagement metrics.
- Check first: Confirm current follower breakdown, comparable post engagement, collaboration terms and outcome evidence.

### Takeaways

- Shopify reported 82 purchases while Meta reported 103 in one example.
- Check duplicate Pixel and CAPI events or repeated purchase triggers as possible faults.
- A GA4 purchase event can fire and still carry the wrong numbers.

## Main ideas

- **When Meta reports more purchases than your store, Meta is not automatically the one that is wrong**: Roy uses the familiar mismatch of a store reporting 82 purchases and Meta Ads reporting 103 to challenge the reflexive conclusion that Meta is overcounting. The gap can have several causes at once: different attribution models and windows; the same purchase arriving through both Meta Pixel and Conversions API without correct Event ID deduplication; Purchase firing more than once through a refresh or repeated checkout action; internal, gift, staff or $0 orders reaching Meta as real purchases; a wrong value or currency in the event; and consent states limiting available data. They first determine whether the issue is a tracking discrepancy, an attribution discrepancy, or both, then work through tracking, deduplication, data quality, consent and attribution in that order. The layers are not interchangeable: fixing duplicate events will not resolve an attribution difference, and changing attribution settings will not repair broken purchase tracking.
- **A healthy tag container tells you almost nothing about whether the data is right**: Roy says "tags are firing" is never enough: a tag can fire correctly while sending the wrong information. Their audit follows the whole path, from the website and its data layer through GTM, consent, GA4 and Google Ads, Meta, and checkout or CRM, then reproduces the customer journey to see what data is actually being sent. The audit asks whether the website sends marketing platforms what they need, not whether the container is tidy.
- **A GA4 purchase event can fire and still carry the wrong numbers**: Roy argues that successfully reporting purchases does not show ecommerce measurement is correct; what looks like a GA4 reporting problem is often a data-quality problem in the ecommerce data behind it. Their checks cover whether GA4 receives the actual order value; whether the right currency is sent and interpreted correctly; whether every order carries a unique transaction ID; whether a refresh or revisit to the thank-you page can send the same purchase twice; whether item data such as product IDs, names, prices and quantities is correct; whether shipping and tax are passed consistently with the store platform; whether the event still fires when a customer pays through PayPal or Klarna; and whether it behaves correctly across consent states.
- **Browser-only tracking leaves campaigns working from partial data**: Roy compares running Meta Pixel without the Conversions API to driving with one headlight. They cite ad-blocker data loss, lower Event Match Quality, incomplete attribution, iOS privacy changes and cookie restrictions that cut visibility. In performance terms, they point to lower match rates and higher CPMs, fewer conversion signals and weaker delivery optimisation, and skewed attribution that leads to bad decisions. Adding CAPI feeds Meta verified server-side data that cannot be blocked or lost, with claimed gains including more accurate reporting, stronger signal quality, better algorithmic optimisation, longer cookie lifetime and a clearer view of the customer journey.
- **Fix the conversion signals before changing bids, audiences or campaigns**: Roy describes a brand with good campaigns, strong copy, the right targeting and a healthy budget, where Google Ads is still learning from the wrong signals. The hidden problems include duplicate conversions counting one lead or purchase several times, optimisation toward a low-value conversion action instead of the real business goal, conversion values that do not reflect actual business worth, broken cross-domain tracking between site and checkout, missing Enhanced Conversions, consent-related measurement gaps, and attribution differences between Google Ads, GA4 and a CRM. Google Ads can optimise perfectly for the wrong data, so the first question is whether it can trust the signals it receives.
- **Attribution is a journey problem, not a URL-parameter problem**: Roy argues that UTM parameters tell you where traffic came from but do not guarantee the information survives to purchase. They trace the path from Google Ad to landing page to product to checkout to purchase, noting that attribution data can be lost at any step. Their checks cover campaign parameters, platform click identifiers, first-party capture that preserves campaign information, cross-domain handling when checkout, booking or payment moves to another domain, and CRM or offline conversion data so a lead's story continues past the form. Their summary: the source of traffic is not the same as the source of conversion, and neither tells you how valuable the customer will be.
- **Consent belongs inside the measurement design, not bolted on at the end**: Roy treats Consent Mode V2 and CMP integration as a standard checklist item, not a legal afterthought, and tests the same journey under consent granted, denied and changed states because each state changes what data is available to measure. Several posts address EU businesses specifically, arguing that consent choices must be respected while still measuring what can legitimately be measured. In the discrepancy work, consent appears both as a cause of differences between platforms and as something to test rather than assume.
- **Predictive and automated features depend on clean event data**: Roy explains GA4 predictive metrics: purchase probability within seven days, churn probability and predicted revenue over 28 days. They require roughly 1,000 or more users and relevant conversion events within 28 days; results are not retroactive, and a consistent event taxonomy is needed before predictive audiences can be synced to Google Ads. Roy also outlines how to align Google Ads, GA4 and GTM through data-driven rather than last-click attribution, import GA4 conversions into Google Ads for bidding, use custom dimensions such as campaign_id and session_id to join click data to events, and apply server-side tagging to reduce discrepancies. In both cases, the features turn data into action only when the underlying event setup is consistent.

## Work to read

- [Meta Reports More Purchases Than Your Store: Is Meta Overcounting?](https://www.linkedin.com/posts/bivasroy-conversiontracking-meta-ads-gads_metaads-metacapi-attribution-activity-7504059410848251904--oX0)
- [What a Real Tracking Audit Found Behind a “Healthy” GTM Container](https://www.linkedin.com/posts/bivasroy-conversiontracking-meta-ads-gads_googletagmanager-ga4-conversiontracking-activity-7505868150262116352-z5tS)
- [A 15-Point Conversion Tracking Checklist Before Trusting Any Data](https://www.linkedin.com/posts/bivasroy-conversiontracking-meta-ads-gads_conversiontracking-gtm-ga4-activity-7513189312587309056-OAOe)
- [Start an Audit With What Should Be Tracked, Not Whether Tags Fire](https://www.linkedin.com/posts/bivasroy-conversiontracking-meta-ads-gads__conversiontracking-gtm-ga4-activity-7505144510117752832-kVy4)
- [Running Pixel Without the Conversions API: What You Are Missing](https://www.linkedin.com/posts/bivasroy-conversiontracking-meta-ads-gads_metaads-facebookads-adtracking-activity-7389923848504336384-Tc4z)
- [Why a GA4 Purchase Event Can Fire and Still Be Wrong](https://www.linkedin.com/posts/bivasroy-conversiontracking-meta-ads-gads_ga4-ecommerceanalytics-digitalanalytics-activity-7513552984828489731-oDSK)
- [UTM Parameters Are Not Enough to Explain Attribution](https://www.linkedin.com/posts/bivasroy-conversiontracking-meta-ads-gads_utm-attribution-conversiontracking-activity-7510291624124440576-Izqd)
- [Google Ads Campaigns Cannot Perform Better Than the Conversion Data Allows](https://www.linkedin.com/posts/bivasroy-conversiontracking-meta-ads-gads_googleads-conversiontracking-gtm-activity-7508120694455427072-w3ve)

## Partnerships

Bivas Kumar Roy's posts explain tracking and measurement problems rather than promote brand collaborations. They include no sponsorship disclosure, partner tag, rate card, or media kit. Several close with an invitation to connect or send a LinkedIn message for tracking help, typically with audits, conversion discrepancies, or Consent Mode issues. That opens a conversation about work; it is not a published advertising offer. Buyers should ask directly about collaboration terms and any sponsorship or review formats. Work examples appear in the profile sections on audits, ecommerce measurement, and ad-platform integrations.

## Alternatives

- Samuel Adeitan: Use this work as a reference for agency operations and automation economics, not a hands-on tracking audit. One post weighs the cost of task-based automation at high volumes and suggests self-hosted automation, Python transformations and queueing; another discusses ecommerce acquisition and Meta measurement.
- Reeto M.: For a campaign about incrementality and evaluating attribution approaches, these examples offer a closer angle. They describe a two-cell test of predictive CAPI against an existing setup and discuss Meta’s incremental-attribution reporting.
- Josh Gregory: Use these examples for a brief about scaling ecommerce paid media: they address funnel gaps, coordination across channels, acquisition costs, offers and campaign structure. They are a stronger fit for media strategy than technical measurement troubleshooting.

---
AI-assisted summary of public posts. Audience demographics are not available. Engagement appears only where we checked recent posts. Method: https://b2binfluencer.marketing/methodology/