A Meta campaign can show a low cost per lead and still lose money. The reason is simple: a form submission or WhatsApp message is an early signal, not a sale. If the sales team cannot reach the person, the person is outside the service area, the requirement is unsuitable or the offer is misunderstood, the campaign may be buying activity rather than customers.
This guide is a diagnostic method for Indian businesses receiving enquiries from Facebook or Instagram ads. It separates delivery problems from offer, audience, response and sales-process problems. The example numbers are hypothetical and are included to demonstrate the calculations; they are not industry benchmarks or Kardengey client results.
Start by defining what counts as a lead
Write down the event your campaign is optimising for. An instant form submission, a WhatsApp conversation started, a phone call connected, a qualified enquiry and a paid customer are different events. If the campaign reports 100 leads but the business counts only 18 people who answered the phone, both numbers can be true. They describe different stages.
Create a simple shared definition. For example: a qualified lead is a real person with a genuine requirement, inside the service area, who meets the minimum project or purchase criteria and can be contacted. Adapt the definition to your business. A local salon, a B2B software company and a high-ticket interior designer should not use the same qualification rules.
Map the funnel before changing the ads
Track these stages separately: ad click or impression; form submission or WhatsApp conversation; contact attempted; contact reached; qualified; appointment or quotation; sale; revenue collected. Record counts for the same date cohort. Do not compare leads generated this week with sales closed from last month's leads.
Worked example: a low CPL can hide a weak funnel
Illustrative example: a business spends ₹15,000 and receives 100 enquiries. The reported cost per enquiry is ₹150. The team reaches 60 people, qualifies 24, sends 12 quotations and closes 3 sales. Cost per reached contact is ₹250; cost per qualified lead is ₹625; cost per quotation is ₹1,250; and advertising cost per customer is ₹5,000. These are four different views of the same campaign.
If the three customers generate ₹45,000 in collected revenue, the revenue-to-ad-spend ratio is 3.0. That is not automatically profit or ROAS on contribution: product costs, fulfilment, discounts, taxes, agency fees and other operating costs still matter. Calculate contribution after variable costs before deciding whether to scale.
Use this diagnosis table
Check the response process before blaming targeting
For WhatsApp and phone leads, record the time of first response, number of attempts, whether the person was reached and the outcome. A lead that waits several hours may have already contacted another provider. This is a process hypothesis to test against your own data, not a universal response-time statistic.
Use a consistent first reply that acknowledges the enquiry and asks one useful next question. Avoid sending a long catalogue or generic price list before understanding the requirement. Give staff a short qualification script and a clear next action: call, appointment, quotation or polite disqualification.
Audit the promise made by the creative
The ad, landing page or WhatsApp prefilled message should describe the same offer. If the creative says 'starting at ₹X' but the actual product most customers need costs much more, expect price mismatch. If the ad implies instant delivery but the business needs a consultation, expect unsuitable expectations. Be specific about who the offer is for, what is included and what happens after the enquiry.
Run controlled tests, not random edits
Change one meaningful variable at a time where practical: offer, creative angle, audience, form questions or destination. Keep a dated experiment log with the hypothesis, change, start date, spend, lead count, qualified count and decision. Avoid declaring a winner from a handful of conversions or changing multiple settings before you can interpret the result.
A weekly lead-quality scorecard
Keep the lead register privacy-conscious. Collect only information needed to respond, restrict access to the people handling enquiries, and follow applicable privacy and platform requirements. Do not upload customer lists or sensitive personal information to advertising tools unless you have a lawful basis and the required permissions.
When should you increase the budget?
Do not scale because the platform says results are cheap. First confirm that tracking is reliable, lead quality is stable, the team can handle additional enquiries and the economics work using actual sales and contribution data. Increase budgets gradually enough to detect whether lead quality and cost change. There is no universal percentage increase that is safe for every account.
A simple decision rule
If the enquiry volume is low, investigate delivery, audience size, creative and offer. If enquiry volume is healthy but qualification is weak, investigate message match and targeting. If qualification is healthy but sales are weak, investigate response and sales process. If sales are healthy but profit is weak, investigate margin and acquisition economics. This prevents the common mistake of treating every business problem as an ad-targeting problem.
The best campaign report is not the one with the lowest cost per lead. It is the one that connects spend to qualified demand, customers and sustainable contribution. Start by measuring the whole journey, then decide which part deserves the next experiment.
Related guide: How to track WhatsApp leads for small businesses in India
Related guide: Marketing ROI calculation and free calculator