The Starting Point
The industry benchmark CPL for restaurant lead generation on Meta sits between Rs 95 and Rs 125. Most campaigns I inherited were operating at or above that range, with budgets consolidated at the city level and creatives running brand awareness objectives.
The problem was not the spend. It was the architecture. City-level campaigns with link click objectives generate impressions and traffic. They do not generate qualified conversations. The moment I shifted the question from "how much reach did we get?" to "how many people actually expressed intent?", everything changed.
The Structural Shift That Made the Difference
The single biggest lever was moving from city-level to outlet-level campaign structures. Instead of one campaign targeting an entire city, each outlet ran its own campaign with its own audience, its own creative, and its own conversion objective.
This sounds like more work. It is. But it produces three things that consolidated campaigns cannot:
- Hyperlocal audience targeting. Someone 2km from a Bandra outlet has different intent than someone 12km away. Outlet-level campaigns let you radius-target precisely.
- Proposition-specific creative. The offer at lunch is not the same as the offer on a Saturday evening. When each outlet has its own campaign, the creative can match the specific proposition being promoted.
- Clean CPL data. When you consolidate campaigns, poor performers hide behind strong ones. Outlet-level structure surfaces exactly which location, which creative, and which audience is driving cost efficiency.
The Objective Change
Switching from link click to WhatsApp conversation or instant form objectives was the second structural change. Link clicks measure traffic. Conversation objectives measure intent.
The CPL comparison is instructive. On link click campaigns, we were tracking cost per click, not cost per lead. When we moved to instant form or WhatsApp objectives, the actual lead count was suddenly visible, and the cost per genuine lead was measurable.
The best result in this campaign series: 586 WhatsApp conversations at Rs 15 CPL. Industry benchmark is Rs 95 to 125. That is not an optimisation, it is a structural difference.
The Rs 15 CPL result did not happen because of a perfect creative. It happened because the right objective, the right audience radius, and the right proposition were aligned at the outlet level.
How Creative Testing Fit In
Once the structure was right, creative testing became meaningful. We ran a staff-vs-product creative test across multiple outlets. Staff creatives show the team behind the food. Product creatives show the food itself.
The hypothesis was that staff creatives would build trust. The data said something different. Product creatives drove higher intent for delivery propositions. Staff creatives outperformed for dine-in occasions where the experience is part of the purchase decision.
This kind of finding is only visible when your campaign structure produces clean, comparable data. In a consolidated city-level campaign, the split is too noisy to be conclusive.
The Metric That Actually Matters
CPL is a useful benchmark but it is not the endpoint. A Rs 15 CPL is only valuable if the conversation converts. The real measure is cost per seated cover or cost per confirmed reservation.
What outlet-level campaigns give you is the ability to trace that path. You know which outlet, which proposition, and which creative produced a conversation. When the restaurant team tracks which conversations converted, you close the loop between ad spend and actual revenue.
Most F&B marketers stop at CPL. The ones who build trackable conversion paths from ad to cover are the ones who can confidently justify budget increases.
What to Take From This
- Move from city-level to outlet-level campaign structures before optimising anything else
- Use conversation or lead form objectives, not link clicks, if you want to measure actual intent
- Test creatives only after the structure is right, otherwise the data will not be clean enough to act on
- Define success as cost per conversion, not cost per lead