A Straight Guide to Meta Ads for Small Businesses

Meta ads are simultaneously the easiest advertising platform to start on and one of the easiest to waste money on, because the barrier to spending is a credit card and the barrier to spending well is measurement you cannot see. Here is how the thing actually behaves.

Fix measurement before you spend anything

Optimising against broken signal is the most common way ad budget disappears, and it is invisible from inside Ads Manager — the dashboard reports confidently either way.

So before the first rupee: pixel installed and firing correctly, events mapped to real business actions, and the Conversions API sending server-side events. Since the iOS privacy changes, a browser-only pixel loses a large share of conversions, and the algorithm optimises toward what it can see. Server-side events with proper deduplication recover most of that. If you do one thing from this article, do this one.

Size the budget for learning, not for comfort

Meta's delivery system needs roughly 20–30 conversion events per week per ad set before it optimises reliably. Below that threshold you are not buying a cheap test — you are buying statistical noise and then drawing conclusions from it.

If your product costs ₹2,000 and converts at 2%, that threshold implies a daily budget you should calculate before you launch, not discover after a month of flat results. When the maths does not work, the answer is to optimise for an earlier event — add to cart, lead, signup — rather than to run an under-funded purchase campaign.

Creative is the lever now, not targeting

Meta's audience tools have narrowed and its algorithm has got much better at finding buyers on its own. The practical consequence: broad targeting plus strong creative now beats clever targeting plus weak creative, almost without exception.

Which means volume. Six or more genuinely different creatives per test cycle — not six colour variants of one image. Different angles, different formats, different hooks in the first two seconds. Static, carousel and short-form video all earn their place, and video usually wins on cost per result.

Keep the account structure boring

The instinct is to build an elaborate account with dozens of ad sets slicing the audience finely. This fragments your conversion data across too many ad sets to clear the learning threshold in any of them.

Start with one or two campaigns, a small number of broad ad sets, and your creative variety inside them. Consolidate rather than split. Add retargeting and lookalike audiences only once cold acquisition is working — retargeting a funnel nobody is entering is just paying to reach your existing customers.

Why it worked and then stopped

Two causes account for most sudden collapses.

• Creative fatigue. The same audience has seen the same ad too many times. Frequency climbs, click-through drops, cost per result rises. The fix is a refresh cycle, run on a schedule rather than in a panic.

• Broken tracking. Someone shipped a site change and the pixel stopped firing, or event match quality fell off a cliff. Nobody notices for weeks because spend keeps going out and the dashboard keeps reporting. Monitoring event match quality catches this in days.

Note that both are site-side problems as much as ad problems. That seam — between whoever runs the ads and whoever can change the website — is where most paid budgets quietly leak. It is the reason we run Meta ads and the store itself from one team.

What good reporting looks like

Spend, cost per acquisition, return on ad spend, and contribution — reconciled against your own order data rather than against what Ads Manager claims. Attribution windows mean the platform will always report more conversions than your bank account shows. Any report that does not acknowledge the gap is not a report.

Frequently Asked Questions

What should I budget to start?

Enough to clear the learning threshold for the event you are optimising toward — typically a daily figure derived from your conversion rate and order value, not a round number. Management for a small account starts at ₹30,000 a month plus the ad spend itself.

Meta or Google first?

Google if people already search for what you sell; Meta if you have to create the demand. We go through the decision properly in Facebook ads vs Google Ads.

Do I need a landing page or can I send traffic to my homepage?

A homepage converts paid traffic badly because it answers everyone's question and nobody's specifically. A page matching the ad's promise typically converts several times better, which usually dwarfs any targeting improvement.

Can you make the creative too?

Yes — static, carousel and short-form video, because creative volume is the main lever. If you have a brand team or an existing asset library we work within it and focus on angles and testing instead.