How to measure performance marketing beyond clicks and platform ROAS
Connect advertising reports to valid orders and qualified leads so budget decisions reflect the business.
Measure performance marketing by connecting advertising spend to qualified leads, valid orders and customer value. Clicks and platform ROAS help diagnose campaigns, but they do not prove profitability. Agree on conversion definitions, reconcile advertising data with your CRM or order system and include cancellations, fulfilment costs and the sales cycle in decisions.
Define the business outcome before configuring tracking
A click on a phone number, a submitted form and a paid invoice are different events. Track them separately. A service business might choose an accepted enquiry as an early quality signal and a completed sale as the final outcome. An online store should distinguish a placed order from a fulfilled, non-refunded order. If cash-on-delivery orders are part of the business, their cancellation and return patterns can change the interpretation of campaign revenue.
Agree which system is the source of truth for each stage. Advertising platforms explain delivery and attributed actions; the CRM explains lead status; the order or accounting system explains realised revenue. They will not always agree because attribution windows, identity, reporting time zones and consent differ. Reconciliation should explain differences rather than force every report to display the same number.
Use a small set of well-defined metrics
| Metric | Calculation | What to check | |
|---|---|---|---|
| Cost per qualified lead | Ad spend divided by accepted leads | A consistent qualification rule and duplicate removal | |
| Lead-to-sale rate | Customers from a lead cohort divided by qualified leads in it | Enough time for that cohort to complete the sales cycle | |
| Media acquisition cost | Ad spend divided by newly acquired customers | Do not label this fully loaded CAC if other acquisition costs are excluded | |
| Fully loaded CAC | Defined sales and marketing acquisition costs divided by new customers | Consistent treatment of people, tools and agency costs | |
| ROAS | Attributed revenue divided by ad spend | Attribution model, refunds and revenue definition; revenue is not profit |
An illustrative example: a lower lead cost can still lose
These numbers explain the calculation; they are not EmergeScale client results or a forecast.
Suppose campaign A spends ₹30,000 and produces 100 enquiries. Twenty meet the agreed qualification rule and four become customers. The cost per raw enquiry is ₹300, the cost per qualified lead is ₹1,500 and the media acquisition cost is ₹7,500. Calling all 100 enquiries leads without qualification hides the quality problem.
Campaign B also spends ₹30,000 but produces 60 enquiries, of which 30 qualify and six become customers. Its raw enquiry cost is higher at ₹500, yet its qualified lead cost is ₹1,000 and its media acquisition cost is ₹5,000. B looks stronger on those outcomes. You still need customer margins, retention and sales costs before concluding that either campaign is profitable.
Build a report that supports a decision
- 01
Test the event journey
Check that success events fire after successful actions, not when a button is merely clicked. Test duplicate submissions, refreshes and failed payments. Where browser and server events are both used, configure deduplication according to the platform's documentation. Respect consent choices and avoid sending personal information in URLs or analytics event names.
- 02
Capture source and lead status
Use consistent campaign naming and permitted source identifiers. Assign someone to record qualification, sale and rejection reasons in the CRM. Keep the original acquisition source alongside later interactions so follow-up does not erase how the relationship started.
- 03
Review mature cohorts
Compare enquiries acquired in the same period after a suitable sales-cycle delay. Flag small sample sizes and seasonal changes. Separate an observation from its explanation: falling sales could come from targeting, stock, pricing, response delays or tracking loss.
- 04
Record the next experiment
Choose one meaningful change, its expected effect and the evidence needed to assess it. Keep a record of budget, landing-page and offer changes. This prevents a monthly report from becoming a collection of charts with no connection to the next business decision.