Marketing attribution for Indian businesses: a practical model

Add up the conversions Meta, Google and LinkedIn each claim, and you'll usually get more sales than you actually made. Every platform takes credit for the same customer. Attribution is about getting a view that's honest enough to decide where the next rupee should go.

Why the numbers don't add up

• Each platform counts its own touch. A buyer who saw a Meta ad, then searched on Google, gets counted by both.
• Privacy changes. Since Apple's App Tracking Transparency in 2021 and the decline of third-party cookies, platforms see less and model more.
• Offline outcomes are invisible. Site visits, demos and COD deliveries happen outside the ad platforms unless you send them back.
• Long cycles. A real estate booking or SaaS deal can take months. Platform attribution windows are often days.

Perfect attribution isn't possible. Useful attribution is.

A practical model in four layers

Layer 1: Capture the source on every lead

Store UTM parameters, click IDs (such as Google's GCLID and Meta's FBCLID) and the landing page on every lead in your CRM. Portal and offline leads get a source too. This one step fixes most attribution problems.

Layer 2: Track outcomes in the CRM

Define the stages that matter, such as site visit, SQL, booking or delivered order, and record when each lead reaches them. Now you can report cost per outcome by source, using your own data rather than the platforms'.

Layer 3: Send outcomes back to the platforms

Use offline conversion imports and server-side tools such as Meta's Conversions API and Google's enhanced conversions for leads. The algorithms then optimise for real outcomes, and their reporting gets closer to reality.

Layer 4: Check with simple incrementality tests

Every quarter, test one big assumption. Pause a channel in one city, or hold out a region, and watch what happens to total outcomes. Tests like this tell you what a channel actually adds, beyond what it claims.

Which model should you report?

View	Good for	Watch out for
First touch	Understanding what creates demand	Undervalues closing channels
Last touch	Understanding what closes	Undervalues awareness
Blended / multi-touch	Day-to-day budget decisions	Needs clean data to be meaningful
Blended CAC or MER	The total health check	Doesn't tell you which channel to change

Most Indian businesses do well with last-touch from the CRM for weekly decisions, blended CAC for the monthly review, and incrementality tests for big budget shifts.

By industry

• Real estate: track enquiry, visit and booking by source, including portals. Judge campaigns on cost per site visit and booking.
• SaaS: track MQL, SQL, opportunity and closed-won, with pipeline value. Report CAC payback by channel.
• D2C: track placed vs delivered orders. Judge on RTO-adjusted ROAS and contribution margin, and include marketplace sales when you look at total impact.

Common mistakes

• Trusting platform ROAS as revenue.
• Leaving portal, phone and walk-in leads without a source.
• Changing the attribution model every month.
• Building a complex dashboard before the CRM data is clean.

A 30-day attribution set-up plan

You don't need a data team to get useful attribution. Most of the work fits into one month:

Week 1: Audit and define. List every lead source, including ads, portals, website forms, phone, walk-ins and referrals. Agree the outcome stages that matter (visit, SQL, booking, delivered order) and write down exactly what each means.

Week 2: Capture. Add hidden UTM and click-ID fields to every form, set up call tracking numbers for major channels, and make sure portal and offline leads get a source when they're created. Test each source end to end with a dummy lead.

Week 3: Connect. Send outcomes back to the ad platforms through offline conversion imports or server-side integrations. Check that what the CRM reports and what the platforms receive roughly match.

Week 4: Report. Build one simple report: spend, leads, outcomes and cost per outcome by source, for the last 30 days. Review it with marketing and sales together, and agree what you'll change.

After that, add one incrementality test a quarter and refine as the data gets cleaner.

A simple incrementality test you can run

Suppose you want to know whether a channel really drives sales, or just takes credit for buyers who would have found you anyway. One practical approach:

1. Pick two comparable regions or cities with similar historical performance.
2. Keep the channel running in one and pause it, or cut it sharply, in the other for three to four weeks.
3. Compare total outcomes (site visits, SQLs or delivered orders) in both regions against their usual trend.
4. If outcomes in the paused region barely move, the channel was mostly capturing demand that existed anyway. If they drop, it was creating demand.

It isn't a perfect experiment, but it's far better than relying on any one platform's view of its own value.

Reading the numbers without fooling yourself

• Look at trends, not single weeks. Attribution data is noisy; month-on-month trends are more reliable than any one report.
• Expect channels to disagree. Platform-reported and CRM-reported numbers will never match exactly. A gap that stays stable is fine. A gap that suddenly widens points to a tracking problem.
• Give long cycles time. For real estate and B2B SaaS, judge campaigns on cohorts that have had time to convert, not on last week's leads.
• Keep one source of truth. When marketing, sales and finance each have their own numbers, every meeting turns into a debate about data instead of decisions.

Frequently asked questions

Do we need an expensive attribution tool?

Usually not at first. A well-configured CRM with clean source tracking and conversion uploads covers most needs for Indian businesses. Specialist tools make sense once you have many channels, high spend and clean data.

How do we attribute portal and walk-in leads?

Give each portal its own source and ask walk-ins how they heard about you, recorded as a required field in the CRM. Unique phone numbers per channel help too.

What about brand and awareness campaigns?

Judge them on leading indicators such as branded search volume, direct traffic and lower costs in performance channels, plus the occasional incrementality test, rather than last-touch conversions.

Getting it set up

Attribution is mostly plumbing: forms, CRM fields, integrations and conversion uploads. Our CRM, RevOps and attribution team sets this up on HubSpot, Zoho, Salesforce or PeerGrowth CRM, and reviews the numbers with you every month.