How to Attribute Marketing Spend to Revenue (When Your Sales Cycle Is Months Long)

Most attribution defaults to last-click: whatever the buyer touched immediately before converting gets the credit.

For a ₹500 impulse purchase, that's roughly fine. For a deal that takes nine months and involves six people at the buying company, it's close to meaningless — and it fails in a specific, predictable direction.

Last-click systematically over-credits the bottom of the funnel. The branded search, the retargeting ad, the “book a demo” click. These are the touches closest to the conversion, so they collect the credit — even though the buyer only searched your brand name because of something they read four months earlier.

The consequence is a budget death spiral: bottom-funnel activity looks efficient, so it gets more money; top-funnel looks wasteful, so it gets cut; six months later there's nobody left in the funnel to retarget, and the bottom-funnel numbers collapse too.

For long B2B cycles, W-shaped is usually the most honest — it credits the touch that created awareness, the one that produced the lead, and the one present when it became a real opportunity.

But no model is correct. They're different lenses on the same incomplete data. Which brings us to the practical advice.

1. Track source at the lead, and keep it attached. The single highest-value thing, and the most commonly broken. When a lead becomes an opportunity and then a customer, the original source must still be attached. If your CRM handoff loses campaign data — which it does when marketing and sales run separate systems — every downstream question becomes unanswerable. (More on the handoff: What Is Lead Management?)

2. Measure to pipeline, not just to lead. Cost per lead is a vanity metric if lead quality varies by source. Cost per qualified opportunity is the number that means something. A channel producing leads at ₹200 that never qualify is worse than one producing them at ₹2,000 that convert.

3. Use self-reported attribution. Add “How did you hear about us?” as an open text field on your form. It's unfashionable and it's often more accurate than your tracking, because it captures what your analytics structurally cannot: the podcast, the WhatsApp forward, the colleague's recommendation, the event conversation.

Combine it with tracked data. Where they disagree, that gap is itself informative.

4. Accept a measurement window. If your cycle is nine months, campaigns run today won't show revenue for nine months. Reporting monthly on revenue attribution guarantees you're looking at decisions made three quarters ago. Report leading indicators monthly, revenue attribution quarterly.

5. Run holdouts when you can. The most rigorous available method: turn a channel off in one region or segment and observe what happens. Harder than reading a dashboard, considerably more truthful.

Indian B2B has a specific complication: a large share of influential touches aren't digital at all. A trade show conversation, a dealer's recommendation, a WhatsApp message from a colleague, a phone call.

What to do: capture them deliberately. Event leads entered with the event as source. Referrals recorded as referrals. Channel-partner leads attributed to the partner. It's manual and imperfect, and it's still better than attributing an offline-driven deal to whichever ad the buyer happened to click on their way to your pricing page.

This connects to a broader point — in many Indian businesses WhatsApp is where the real conversation happens, and none of it reaches the CRM. (See WhatsApp Marketing for Indian Businesses.)

Four numbers, in order of usefulness:

Cost per qualified opportunity, by source. Your primary allocation metric.

Pipeline created, by source. Value, not count.

Win rate, by source. Some channels produce leads that close far better than others.

Time to close, by source. A channel with a longer cycle isn't worse — but it changes what you forecast.

Show ranges, not decimals. Attribution is directional. A dashboard reporting 34.7% of revenue to one channel implies precision that does not exist, and someone will make a decision on the decimal place.

If you have nothing: add source capture to every form and a “how did you hear about us?” field. Make source a required field in your CRM. That alone puts you ahead of most.

If you have source data: connect it through to closed revenue, not just to lead. Then look at cost per qualified opportunity by source, and expect to be surprised.

If you have that: add a multi-touch model, run a holdout, and start reporting quarterly rather than monthly.

Attribution is directionally useful and precisely wrong. Its job is not to tell you exactly which rupee produced which customer — that's unknowable in a nine-month, multi-person buying process. Its job is to stop you cutting the channel that's quietly generating the demand your bottom-funnel activity harvests.

Treat it as a compass, not a ledger.