Meta Ads Targeting Options: A Complete Guide for Small Business

Last month, a bakery owner in Pune told me she had spent 18,000 rupees on Instagram ads and gotten exactly two enquiries. When I opened her Ads Manager, the targeting was set to "Women, 18 to 65, all of India, interested in Food." That audience was over 15 crore people. Her daily budget of 600 rupees was being spread so thin that Meta could not figure out who to show the ad to. We rebuilt the targeting from scratch: women aged 22 to 45, within 8 km of her shop, who had engaged with her Instagram page in the past 90 days. Same creative, same budget. She got 23 enquiries in the next ten days. Targeting is not a detail you configure once and forget. It is the single biggest lever in whether your Meta ads make money or burn it.

The three targeting layers in Meta Ads

Meta organizes its targeting into three layers, each building on the one before it. Understanding this hierarchy saves you from making the most common mistake: jumping straight to interests without considering whether a warmer audience exists.

1. Core audiences (manual targeting)

Core audiences let you define who sees your ad based on demographics, interests and behaviors. This is the layer most small businesses use by default because it does not require any existing data.

• Demographics. Age, gender, location, language, education level, job title, relationship status. Location targeting in India works at the city, pin code and radius level. For local businesses, I always use a radius of 5 to 15 km around the shop rather than selecting an entire city.
• Interests. Meta infers interests from pages people follow, content they engage with and apps they use. "Small business owners" or "Digital marketing" are interest categories you can layer. The trick is specificity: "Organic skincare" is better than "Beauty" because it narrows the audience to people with actual purchase intent.
• Behaviors. Purchase behavior, device usage, travel patterns. In India, "Engaged shoppers" and "Small business owners" are two behavior categories that consistently outperform broad interest targeting for B2B and ecommerce campaigns.

2. Custom audiences

Custom audiences are built from people who already know your business. They convert at two to five times the rate of cold core audiences because the person has already interacted with you. The main sources are:

• Website visitors. Anyone who visited your site in the past 1 to 180 days, tracked through the Meta Pixel. You can get granular: visitors who viewed a product page but did not purchase, or visitors who spent more than 30 seconds on your pricing page.
• Email and phone lists. Upload your customer database and Meta matches it against user profiles. Match rates in India typically land between 30% and 50%, depending on whether customers used the same email for Facebook.
• Video viewers. People who watched 25%, 50%, 75% or 95% of your video ads. A 75% viewer is far more qualified than someone who scrolled past after two seconds.
• Instagram engagers. Anyone who liked, commented, saved, shared or messaged your Instagram account. This is one of the highest-quality custom audiences because engagement signals genuine interest.

3. Lookalike audiences

Lookalike audiences find new people who resemble your existing customers. You pick a source audience (a custom audience), and Meta identifies users with similar traits and behaviors. Two things matter here:

• Source quality. A lookalike built from your top 100 paying customers will outperform one built from all website visitors. The source defines who Meta looks for, so feed it your best data.
• Percentage sizing. In India, a 1% lookalike covers roughly 3 to 4 million people. That is enough for most small business campaigns. Going to 3% or 5% increases reach but dilutes quality. I start every campaign at 1% and only expand to 2% if delivery slows down.

Advantage+ audience vs manual targeting

Meta introduced Advantage+ audience as a replacement for detailed targeting in many campaign types. Instead of you choosing interests and demographics, you provide "audience suggestions" and let Meta's algorithm find the right people. It sounds great in theory. In practice, it works well only when your pixel has strong conversion data, typically 50 or more conversions in the past 30 days. If you are starting a new account or running your first campaign, Advantage+ has nothing to learn from. It will spend your budget exploring broadly, which is expensive when your daily budget is 500 to 1,000 rupees. My approach for new clients: run manual targeting for the first four to six weeks, build up pixel data, then test Advantage+ in a separate ad set. Compare cost per result side by side. For most Indian small businesses I have worked with, manual targeting wins in the first month. Advantage+ catches up by month two if the creative is strong.

What changed in 2026: privacy-first targeting

Meta removed several detailed targeting options related to health, religion, political beliefs and financial status in late 2025. In 2026, the trend has continued. Interest targeting categories are broader and less specific than they were two years ago. The practical impact is that the "narrow your audience with five stacked interests" playbook no longer works as well as it used to. Meta is pushing advertisers toward broad targeting plus strong creative. The algorithm identifies who responds to your ad and finds more people like them. This works for brands with large budgets and high creative volume. For small businesses spending under 30,000 rupees per month, I still recommend starting with manual targeting, because you cannot afford the learning phase that broad targeting requires.

Targeting strategy for small budgets (under 30k/month)

Here is the exact framework I use for clients with monthly ad budgets of 15,000 to 30,000 rupees:

• Start with retargeting. Build a custom audience of website visitors (past 30 days) and Instagram engagers (past 90 days). Run your first ad set to this warm audience. Even if the audience is only 1,000 to 5,000 people, they already know you. Conversion rates from retargeting are three to five times higher than cold traffic.
• Layer one cold audience. Pick two to three specific interests that describe your ideal customer. Combine them with tight location and age targeting. Keep the audience between 2 and 5 lakh people. Run this as your second ad set with 60% of the remaining budget.
• Test a 1% lookalike. Once you have 100 or more people in your custom audience, build a 1% lookalike. Run this as your third ad set. After two weeks, compare cost per result across all three ad sets and shift budget toward the winner.

Common targeting mistakes

I see the same errors across almost every small business account I audit:

• Audience too broad. Targeting "all of India, 18 to 65" with a 500 rupee daily budget is the fastest way to waste money. Meta spreads impressions so thin that the algorithm never learns who your buyer is.
• Too many interests stacked. Adding 15 interests does not narrow your audience. It expands it because Meta uses OR logic, not AND. Each interest adds more people to the pool. Use two to four highly relevant interests.
• Never excluding audiences. If you are running a cold prospecting campaign, exclude your existing customers and recent converters. Otherwise you pay to reach people who already bought from you.
• Editing live ad sets. Changing the targeting on an active ad set resets Meta's learning phase. Duplicate the ad set, make your changes on the copy, and pause the original once the new one exits learning.

Frequently asked questions

What is the best Meta ads targeting option for small businesses with low budgets?

For budgets under 20,000 rupees per month, start with a custom audience of your existing customers or website visitors and build a 1% lookalike from it. This gives Meta the strongest signal about who to target without requiring a large audience size. If you do not have enough data for a custom audience yet, use core audiences with narrow interest targeting combined with location targeting. Keep your audience between 2 to 5 lakh people. Anything smaller limits delivery, and anything broader wastes budget on people who will never convert.

Should I use Advantage+ audience or manual targeting in 2026?

It depends on your data. Advantage+ audience works well when your pixel has recorded at least 50 conversions in the past 30 days because Meta's algorithm has enough signal to find the right people. If you are starting fresh or your pixel has fewer than 50 conversions, manual targeting with defined interests and demographics gives you more control. I run both for most clients: one ad set with Advantage+ and one with manual targeting. After a week, I shift budget to whichever delivers a lower cost per result. For most small businesses in India, manual targeting outperforms Advantage+ in the first month.

How often should I change my Meta ads targeting?

Do not change targeting every other day. Give each audience at least 7 days and 50 results before making a judgment. Meta's algorithm needs time to learn and optimize delivery. If you keep resetting the audience, you restart the learning phase each time and waste budget. Review targeting weekly. If cost per result is rising after two weeks, test a new audience in a separate ad set rather than editing the existing one. Editing an active ad set resets learning. Duplicating and adjusting is safer.

Related guides

• Facebook and Instagram ads for local business: the full walkthrough for setting up your first local campaign on Meta.
• How to write Meta ad copy: targeting gets the right people to see your ad, but copy is what makes them click.
• Why your Meta ads are not converting: if your targeting is right but results are still poor, the problem is usually the landing page or the offer.