Boosting a post is the easiest thing Facebook lets you do, and the worst value it offers.
One button, a few hundred rupees, and Facebook shows your post to more people. What it does not tell you is that it optimised for likes, not enquiries. So you get engagement from people who will never buy, and the report still looks fine.
That gap — between running ads and managing them — is what this guide is about.
Why Meta and Facebook mean the same thing here
Facebook renamed the company Meta, so the ad system now covers Facebook, Instagram, Messenger and WhatsApp from one place. Business owners still say Facebook ads. Marketers say Meta ads. Same platform.
Practically, it means one campaign can reach someone on Instagram in the morning and Facebook at night, managed from a single account.
What boosting skips
When you boost, Facebook makes five decisions for you. Every one is worth making yourself.
- The objective. Boost optimises for engagement. A proper campaign optimises for messages, leads or purchases — and Facebook then finds a completely different set of people.
- The audience. Boost gives rough targeting. Real campaigns let you exclude existing customers, reach people who visited your site, or build audiences resembling your best buyers.
- Creative testing. Boost runs one post. A campaign runs several versions and lets the losers die.
- Placement control. Boost spreads budget across placements that may not suit your format.
- Tracking. The big one. Without the Pixel and conversion events, you never learn which ad produced the enquiry.
Same money, very different outcome.
The campaign types that work for local business
WhatsApp message ads
The strongest format for most businesses in Bihar. Someone taps the ad and a WhatsApp chat opens with a message already typed. No form, no landing page, no waiting.
It works because it removes every step between interest and conversation, and because your customers already live on WhatsApp. Set up quick replies for the questions you always get, so the first response goes out in seconds.
Lead form ads
The form opens inside Facebook or Instagram with name and number already filled in. Good for coaching admissions, hospitals and builders — anything where you need details before a conversation.
The trap: forms that are too easy to fill produce leads that are too casual. Add one qualifying question — budget, area, timeline — and the lead count drops while quality rises. That is a good trade.
Website conversions
For businesses selling online, or where the decision needs a proper page. Requires the Pixel and Conversions API working, otherwise the system optimises blind.
Retargeting
Reaching people who already watched your video, visited your site, or messaged and went quiet. Almost always the cheapest leads in the account, and the most underused campaign type in local businesses.
Creative is the biggest lever, not targeting
Most people spend their time on audience settings. Meta’s system is now good enough at finding buyers that targeting matters less than it used to. Creative matters more.
- The first two seconds decide everything. No logo animation. Start with the problem, the result, or a face talking.
- Real beats polished. A phone video of your actual shop often outperforms a designed graphic, because it does not look like an ad.
- Captions always. Most people watch on mute.
- Test three, not one. Three hooks for the same offer produce three very different costs per lead. Nobody can predict which wins.
- Refresh monthly. When frequency climbs and results fall, the ad is not broken — people have simply seen it too often.
Tracking, before you spend anything
Install the Pixel, set up conversion events, and verify them before the first campaign goes live.
The Conversions API matters too. Since Apple’s privacy changes, browser-based tracking misses a large share of conversions. The API sends events server-side, which keeps reporting closer to reality and keeps Meta’s system learning who your buyers are.
Skipping this is how accounts spend for months with nobody able to say which ad produced a customer.
Budget and the learning phase
Meta needs data before it improves. A new ad set stays in a learning phase until it gathers enough conversions, and during that period results are unstable and expensive.
There is a floor. Below roughly ₹3,000 a month the system cannot exit learning and the money leaks. If that is your ceiling, put it into local SEO instead and come back to ads later.
Do not panic in week one. Changing budget or targeting resets learning. The most common self-inflicted damage we see is an owner editing the campaign daily because day three looked bad.
What to watch, and what to ignore
Ignore reach, impressions and engagement rate. They are easy to make look impressive and tell you nothing about revenue.
Watch cost per result (per message, lead or sale), frequency (above 3 to 4 means fatigue), lead quality as reported by whoever answers the phone, and cost per customer once you know your close rate.
That last number is the only one that decides whether to spend more.
Common mistakes
- Too many ad sets on a small budget. Five ad sets sharing ₹5,000 means none of them learns anything. One or two is right at that level.
- Editing daily. Every significant change restarts learning.
- Nobody answering. Message ads replied to on Monday are wasted. Response speed decides more outcomes than targeting does.
- Running one creative for months. It worked, then quietly stopped, and the budget kept going.
- Judging by likes. An ad with few likes and cheap enquiries is the better ad. Every time.
Reading the numbers without fooling yourself
Meta’s dashboard is designed to make campaigns look successful. Three habits keep the picture honest.
Count leads your team actually received. If the platform reports forty and thirty were wrong numbers or people from another state, your real cost per lead is more than double what the dashboard says.
Watch frequency alongside cost. When frequency climbs past three or four and results fall, the ad is not broken — people have simply seen it too often. That is a creative refresh, not a targeting problem.
Treat platform-reported conversions as a direction, not a fact. Meta uses attribution windows that credit itself generously, including for people who would have found you anyway.
One number decides everything else: what did a customer cost, once you account for how many leads actually convert. Everything above it is diagnostics.
Frequently asked questions
How much should I start with?
From around ₹3,000 a month for one campaign in a smaller city. Competitive categories in a metro need more. Fund one campaign properly rather than three badly.
Do I need a website?
Not for WhatsApp or lead form campaigns — both work entirely inside Meta. A website helps for retargeting and credibility, but it is not a blocker.
Why did my ad account get restricted?
Usually a policy issue in the creative or an unverified business, not fraud. Common triggers: before-and-after images in health and beauty, claims about results, and personal attributes in the copy. Fix the cause, then appeal once, properly.
Facebook or Instagram?
Run both and let the system decide where your buyers respond. In most of Bihar, Facebook still reaches older audiences better while Instagram reaches under-35s. The split usually surprises people.
How long before it works?
Enquiries often arrive in the first few days. The first two weeks are the learning phase, where cost per lead is higher and unstable — normal, not failure. Judge it at week four, not week one.
Where to start
Pick one campaign type, get the Pixel working, write three different hooks for the same offer, and run it for four weeks without touching it. Then read the cost per result and decide.
If you would rather have it run properly, our Meta Ads service covers tracking, creative and weekly optimisation — and the account stays in your name. Already running ads? Give us access and we will send back where the budget is leaking, free either way.