On this page
Most Indian marketing teams do not have a data problem — they have a measurement design problem. GA4 is installed, three pixels fire twice, UTMs are inconsistent, and the dashboard never matches the order table. Here is the setup we deploy for clients.
Key Takeaways
- Design an event schema first, then implement — not the other way round.
- Server-side tagging recovers 8–20% of lost conversion signal on Indian mobile traffic.
- UTM discipline is a policy problem: one naming convention, enforced by a builder sheet.
- Reconcile GA4 to your order database monthly; accept a 2–5% variance, investigate anything larger.
- Use data-driven attribution for optimisation, and a simple cohort view for board reporting.
The Minimum Viable Event Schema
For an Indian D2C store, track exactly these, with consistent parameters: view_item, add_to_cart, begin_checkout, add_payment_info, purchase (with value, currency INR, payment_method, shipping_tier, coupon), plus payment_failed — the one event nearly every store omits and the one that explains the biggest revenue gap.
For B2B/SaaS: form_start, form_submit (with form_id and offer type), demo_booked, trial_started, qualified (pushed back from CRM).
Server-Side Tagging Is Now Table Stakes
Browser-side tags on Indian Android traffic lose signal to ad blockers, ITP-style restrictions, flaky networks and slow scripts. A server-side GTM container plus the Meta Conversions API and Google Ads enhanced conversions typically recovers a meaningful share of attributed conversions and stabilises ROAS reporting. Budget one-time setup effort plus a small monthly hosting cost.
UTM Discipline in One Table
| Parameter | Convention | Example |
|---|---|---|
| utm_source | platform, lowercase | google, meta, linkedin |
| utm_medium | cpc, paid_social, email, affiliate | cpc |
| utm_campaign | objective_audience_month | prospecting_d2c_aug26 |
| utm_content | creative or variant id | reel_v3 |
| utm_term | keyword or audience | performance_marketing_agency |
Publish one builder sheet, make it the only sanctioned way to create links, and audit monthly. Inconsistent case alone (Google vs google) splits reporting and quietly misprices every channel.
Consent and Compliance
With India's DPDP framework in force, capture consent state, pass it to tags, and store the basis for processing. Practically: implement a consent banner that actually gates non-essential tags, use consent mode signals, and document retention periods. Treat it as a reporting requirement, not a legal afterthought.
Attribution Without the Theatre
Use three views rather than arguing about one model:
- Platform view — for in-channel optimisation decisions.
- GA4 data-driven attribution — for cross-channel budget shifts.
- Incrementality or cohort view — for the board, tying spend to new-customer revenue by month.
If the three views disagree wildly, the problem is usually tagging hygiene, not the model.
The Monthly Reconciliation Ritual
Once a month, export GA4 purchases and compare against the order table by day. Investigate any day with more than 5% variance. This single habit catches duplicate purchase events, broken thank-you-page redirects and payment-gateway callbacks that skip tracking — issues that otherwise distort optimisation for months.
Start Here
Audit what fires today with the GA4 DebugView and Tag Assistant, write the event schema on one page, then fix tracking before you spend another rupee scaling media. Clean measurement is the cheapest performance improvement available to any Indian marketing team.