Subtotal $0.00
A campaign reporting 5x ROAS looks like a win until someone asks how many of those buyers were already searching for your brand. Return on ad spend shows how much revenue a platform has credited to your ads. It doesn’t show how much the ads caused, what that revenue cost to deliver, or whether the customer ever returns.
That gap is pushing performance marketing teams in Bangalore to look past ROAS. At a MediaBrief roundtable in Bengaluru in July 2026, marketers from quick commerce, gaming, fashion and consumer brands raised problems including overlapping audiences on Google and Meta, harder attribution and results that take longer to judge. Some of the strategies below came up in that discussion. Each comes with the metric that decides whether it works.
- Test incrementality before scaling spend
Branded search and retargeting tend to report strong ROAS because they reach people who were already close to buying. An incrementality test holds back a control group, by audience or by region, and compares results. It answers one question: what would have happened without this campaign?
In November 2025, Google lowered the minimum spend for its own incrementality experiments to US$5,000, putting them within reach of mid-sized advertisers.
What decides it: incremental ROAS (incremental revenue ÷ ad spend).
- Optimise for qualified leads, not form fills
In real estate, education, healthcare and B2B, a cheap lead who never answers the phone is expensive. Send CRM stages (contacted, qualified, site visit or appointment attended, closed) back to the ad platforms through Google’s offline conversion imports or Meta’s Conversions API. Bidding then learns from outcomes instead of submissions.
Agree on what “qualified” means with your sales team before you start.
What decides it: cost per qualified lead and lead-to-visit rate.
- Judge campaigns on margin and payback
Two orders with the same revenue can leave very different profit once discounts, returns, delivery and payment fees come out. D2C brands can test sending margin-based conversion values to ad platforms instead of revenue.
Where repeat purchases matter, track how long a customer takes to repay their acquisition cost. At the Bengaluru roundtable, Gameberry Labs said its focus had moved from installs to ROAS, with lifetime value and CAC payback now gaining weight. Until lifetime value shows, watch 90-day repeat rate.
What decides it: contribution margin per order and CAC payback period.
- Build on consented first-party data
Customer lists and CRM records have several practical uses:
- Exclude existing customers from acquisition campaigns.
- Build audiences from your highest-value buyers.
- Connect offline sales back to the ads that drove them.
Consent is about to matter more. Under India’s DPDP Rules, the obligations covering notice, verifiable consent and security safeguards take effect 18 months after the Rules were published, which lands in May 2027.
What decides it: match rate to first-party records and spend saved on existing customers.
- Treat creative as the main optimisation lever
As Google and Meta automate targeting, marketers have fewer manual controls left. LightFury Games’ Aninditta Datta said at the roundtable: “Creative is the new optimisation lever.”
Test hooks, proof points, offers and formats one variable at a time. If your audience research supports it, compare English, Kannada and bilingual versions.
What decides it: qualified conversions per creative, not click-through rate.
- Fix the landing page before buying more traffic
Paid media rarely makes up for a slow or confusing page. Useful tests include:
- Short forms versus detailed forms
- Click-to-call versus WhatsApp buttons
- Visible price ranges
- Locality-specific pages for a clinic, project or campus
What decides it: conversion rate to a qualified action.
- Make WhatsApp a measured journey
WhatsApp works well for lead qualification, appointment reminders and order updates when each conversation ties back to a CRM outcome. Three core rules apply:
- Users must opt in.
- Opt-outs must be honoured.
- Messages sent outside the 24-hour service window need approved templates.
Since July 1, 2025, Meta has charged per delivered template message, and marketing templates are generally billed. That cost belongs in your CAC.
If you work with a WhatsApp marketing agency in Bangalore, ask how they manage opt-in records, template approvals and handover to a person.
What decides it: cost per qualified conversation and reply-to-booking rate.
- Measure speed-to-lead
Response speed can affect whether a lead is ever reached. Track first-response time by lead source, then test a faster call-back target against your current process and compare contact rates.
What decides it: median first-response time and contact rate.
- Add channels for reach, not as replacements
Nobody at the Bengaluru roundtable argued for dropping Google or Meta; the interest was in additional touchpoints that reach people differently. Candidates include LinkedIn for B2B, YouTube for consideration, or creator partnerships.
Size each test so a short dip in results is affordable. Zivame’s Daman Bali raised this risk: a weak month can take time to recover from.
What decides it: the share of genuinely new customers and incremental conversions.
- Use AI, but check it against real outcomes
AI is useful for creative variants, lead scoring, anomaly alerts and budget suggestions. But a predicted conversion isn’t revenue. Review automated recommendations against CRM results and margins before accepting them.
What decides it: qualified outcomes compared with a period without the automated change.
How to test without wrecking the quarter
Before any test, set four things in writing:
- Hypothesis: what you expect to change, and why.
- Budget cap: an amount you can afford to lose.
- Duration: long enough to cover your sales cycle.
- Decision rules: the result that means scale, the result that means hold, and the result that means cut.
Hold any partner to the same standard. A performance marketing agency in Bangalore should report qualified leads, CAC, margin and payback alongside platform ROAS. It should also be willing to run tests that might show part of that ROAS was never earned.
FAQs
Why isn’t ROAS enough to judge performance marketing?
ROAS shows revenue a platform credits to your ads, not revenue the ads caused. It also ignores product, delivery and discount costs, and says nothing about whether customers come back. Use it to compare campaigns, not to set budgets alone.
What should brands measure alongside ROAS?
Incremental ROAS to see whether ads create demand, cost per qualified lead for lead-generation businesses, contribution margin for e-commerce, and CAC payback wherever customers buy more than once.
How can I tell whether my ads are actually driving sales?
Run an incrementality test: hold back a control group by audience or region and compare its results with the group that saw your ads. Run it for at least two weeks, and longer for considered purchases such as property.
Which metric should I start with for my sector?
- Real estate: cost per site visit.
- Healthcare: cost per attended appointment.
- Education: cost per qualified enquiry or enrolment.
- D2C: contribution margin and repeat rate.
- B2B: cost per sales-qualified lead and pipeline value.