Meta Ads Targeting Options: A Complete Guide for Small Business

Meta ads targeting is the set of audience tools, core demographics and interests, custom audiences built from your own contacts, lookalikes, and Advantage+ automation, that decide who actually sees your Facebook and Instagram ads. Get it right and a Rs 15,000 budget can outperform a competitor spending five times as much. Get it wrong and no amount of budget fixes it. A bakery owner in Pune once showed me an account spending Rs 600 a day targeted at "Women, 18 to 65, all of India, interested in Food," an audience of over 15 crore people. She had two enquiries from Rs 18,000 in spend. I rebuilt her targeting as women aged 22 to 45, within 8 km of her shop, who had engaged with her Instagram in the past 90 days. Same budget, same creative. Twenty-three enquiries in the next ten days. Nothing about her product changed. Only who saw the ad changed. This guide covers every targeting layer available in 2026, what actually works under Rs 30,000 a month, radius versus city targeting for Pune, exclusion lists, and a step-by-step setup you can copy into your own account.

How Meta ads targeting actually works

Every time your ad is eligible to show, Meta runs an auction among all advertisers targeting that same person at that same moment. Your targeting settings do not guarantee a person sees your ad. They define the pool of people Meta is allowed to consider, and Meta then decides who within that pool is most likely to take the action you asked for. This is why two accounts with identical targeting can get different results. The algorithm is also reading your creative, your bid and your conversion history, not just your audience definition.

Boosting a post from the Instagram or Facebook app skips almost all of these layers by default, which is one reason boosted posts so rarely reach the right audience. Targeting has quietly become less about picking the perfect interest and more about feeding Meta good source data and then getting out of the way. The three layers below, core, custom and lookalike audiences, plus Advantage+ automation, are the tools you have. The rest of this guide is about which combination to use and when.

Every Meta ads targeting layer for a small business

Core audiences: demographics, interests and behaviours

Core audiences are built from age, gender, location, language, interests and behaviours, and they are the default starting point because they need no existing data. Location targeting in India works at the city, pin code and radius level, and for any business with a physical location I default to radius over city, covered in detail below. Interests work best when specific: "Organic skincare" outperforms "Beauty" because it narrows to people showing real intent, not just a broad demographic. Behaviours such as "Engaged shoppers" and "Small business owners" consistently beat broad interest stacking for ecommerce and B2B accounts I run.

Custom audiences from your website, WhatsApp and customer lists

Custom audiences are built from people who already have some relationship with your business, and they convert at two to five times the rate of a cold core audience. Three sources matter most for a small business. Website visitors, tracked through the Meta Pixel, let you build an audience of anyone who viewed a page in the last 1 to 180 days, narrowed further to people who viewed a product but did not purchase, the foundation of what most people mean by retargeting. My Meta Pixel setup guide covers the installation if you have not done this yet. WhatsApp and customer lists, exported as a CSV of phone numbers or emails and uploaded under Audiences as a Customer list, typically match 40 to 60 percent of Indian numbers against real Meta accounts, and this list also becomes your most valuable exclusion and lookalike source. Instagram and video engagers, people who liked, commented, saved, messaged or watched most of a video, round out the set and are often the cheapest warm audience available because Meta already knows they pay attention to your content.

Lookalike audiences

A lookalike audience finds new people who statistically resemble a source audience you provide. Two decisions determine whether it works. Source quality matters more than source size: a lookalike built from 100 customers who actually paid will consistently outperform one built from 10,000 people who merely visited your website once. Percentage size matters second: a 1 percent lookalike in India typically covers 3 to 4 million people, which is plenty for almost any small business budget, and I only expand to 2 or 3 percent when delivery slows down on the 1 percent version.

Advantage+ audiences

Advantage+ audience replaced manual detailed targeting as the default in many campaign types. Instead of choosing interests yourself, you give Meta a small set of audience suggestions and let its algorithm find people based on your pixel and conversion history. It works well once your pixel has 50 or more conversions in the past 30 days, because the algorithm has real signal to learn from. Below that threshold it spends its early budget exploring broadly, which is expensive on a daily budget of Rs 500 to Rs 1,000. For a brand-new account, I run manual targeting for the first four to six weeks to build conversion history, then test Advantage+ in a separate ad set and compare cost per result before shifting spend.

Detailed targeting expansion

Detailed targeting expansion lets Meta go beyond the interests and demographics you selected whenever it believes doing so will lower your cost per result, and it is switched on by default in most campaign types now. For a brand-new account with no conversion history, I usually leave it on, because Meta has nothing better to learn from yet and a slightly looser audience beats an artificially narrow one. For an account with strong pixel data and a specific, high-value audience, a real estate developer targeting a narrow income band, for instance, I switch it off in at least one ad set to see what the defined audience alone delivers, then compare.

What works under Rs 30,000 a month

Below this budget, every rupee spent on the wrong audience is a rupee you cannot afford to waste, so the targeting framework has to be simpler and more disciplined than what a large brand would run. Here is the exact structure I use for clients spending Rs 15,000 to Rs 30,000 a month.

Run all three for at least seven days without touching them, then compare cost per result against the benchmarks in what is a good cost per lead for Meta ads in India and shift the following week's budget toward whichever ad set is cheapest, rather than pausing the others outright. For a business just starting out with no pixel data at all, the Facebook and Instagram ads for local business guide covers what to run in week one before any of these audiences exist yet.

Local radius vs city targeting for Pune businesses

Radius targeting draws a circle of a set distance, usually 3 to 10 km, around a pin you drop on the map. City targeting selects the whole municipal boundary, and Pune plus Pimpri-Chinchwad together is well over 7 million people. For a business with one physical location, a clinic in Kothrud, a salon in Baner, a restaurant in Viman Nagar, radius targeting nearly always wins, because it stops you from paying to reach someone in Hinjewadi who will never drive across the city for a haircut.

City-wide targeting earns its place in a narrower set of cases: a business that delivers or services the whole city, like pest control or AC repair; a brand-awareness campaign where reach itself is the goal rather than an immediate visit; or a citywide chain with multiple outlets, where you would otherwise need a separate radius ad set per branch. For everyone else, I start every new local account at a 5 to 7 km radius and only widen it if delivery is slow or frequency has already climbed past 3 within the smaller radius. Even then, I widen it a couple of kilometres at a time rather than jumping straight to the full city.

Exclusion lists that save your budget

The audience you keep out of an ad set matters as much as the one you put in it, and it is the most skipped step in accounts I audit. Set these exclusions once at the ad set level and they keep working without any further attention.

• Existing customers. Exclude your customer list custom audience from every cold prospecting ad set. Otherwise you are paying full price to advertise to people who already bought from you.
• Recent leads. Anyone who submitted a form or messaged you in the last 30 days should come out of the audience for new lead campaigns, so you are not paying twice for the same enquiry.
• Existing followers, for reach or trial-offer campaigns. If the goal is new customer acquisition, exclude your own page followers and Instagram engagers from that specific ad set, since they are more efficiently reached organically or through a separate retargeting ad set.
• Employees and competitors. A small but real source of wasted spend in any account that has been running a while. A short custom audience of known staff and obvious non-buyers, built manually, is worth setting up once.

None of these exclusions cost anything to set up, and together they typically save 10 to 20 percent of budget that was previously spent reaching people who could never have converted.

Why broad targeting plus strong creative often beats narrow interests in 2026

Meta removed several detailed targeting categories, related to health, religion and other sensitive attributes, in late 2025, and the interest list that remains is broader and less specific than it was two years ago. The old playbook of stacking five narrow interests to manufacture a "perfect" audience no longer performs the way it used to, because there is less precision left to stack. In its place, Meta is pushing advertisers toward a broad audience paired with Advantage+ automation, letting the algorithm find who responds and expand from there.

This genuinely works, but only once your account has fed it enough data, which is the qualifier most broad-targeting advice leaves out. An account with 50 or more conversions in the last 30 days and a handful of genuinely different creative concepts can often run one broad Advantage+ ad set that outperforms three narrow interest-based ad sets combined, because the algorithm is optimising on real conversion behaviour rather than a guess about what an interest label implies. An account with no conversion history yet has nothing for that automation to learn from, and broad targeting there just burns budget exploring randomly. My rule: new accounts start narrow with the framework in the Rs 30,000 section above; accounts with a proven pixel and at least three strong creative concepts should test one broad, Advantage+-led ad set against the narrow ones and let seven days of data decide. The creative matters more in the broad setup than the targeting does, since the algorithm is relying on engagement signals to find your buyer, and a mediocre ad gives it nothing to work with. My Meta ad copy guide and Instagram Reels guide cover the creative side of that equation.

How to set up Meta ads targeting step by step

This is the sequence I follow on every new account, from a brand-new Business Manager to a fully layered targeting structure. Each step has its own anchor if you want to send a specific one to whoever manages your account.

Step 1: Pick your objective and conversion location first

Targeting decisions depend entirely on what you are optimising for, so this comes before any audience work. Choose Leads or Sales in Ads Manager, then decide whether the conversion location is an instant form, your website or WhatsApp. My Facebook lead ads guide covers instant forms specifically if that is your conversion location; the targeting steps below apply to any of the three.

Step 2: Build custom audiences from your website, WhatsApp and customer lists

Install the Meta Pixel and Conversions API if you have not already, upload your customer list or WhatsApp contact export as a Customer list audience, and turn on an Instagram engagers audience from the past 90 days. These three become the raw material for everything else in this walkthrough, so build them even before you launch your first ad.

Step 3: Create a 1 percent lookalike from your best source

Once a custom audience has at least 100 people, ideally your customer list rather than all website visitors, build a 1 percent lookalike from it. This is usually the single highest-performing cold audience available to a small account, because it is built from people who actually paid rather than people who merely clicked.

Step 4: Set core targeting: radius, age and language for your city

For a physical location, set a 5 to 7 km radius rather than a city. Set age to match your real buyers rather than a generic 18 to 65. For mass-market categories in Pune, test one ad set in English and one in Marathi or Hinglish, since language-matched creative often lowers cost per result by 15 to 30 percent outside the premium end of the market.

Step 5: Add light detailed targeting and turn on Advantage+ expansion

Add two to four genuinely relevant interests, not fifteen, since Meta reads a stacked interest list with OR logic and every addition widens the audience rather than narrowing it. Leave detailed targeting expansion switched on for a new account so Meta can look slightly past your selections while it is still learning.

Step 6: Set your exclusion list

Exclude existing customers, recent leads from the last 30 days, and, for acquisition campaigns, your own followers. Apply this at the ad set level so it carries through automatically as you duplicate ad sets later. The exclusion list section above covers the full reasoning for each one.

Step 7: Launch three ad sets, wait seven days, then shift budget to the winner

Run retargeting, narrow cold and lookalike simultaneously with the budget split described in the Rs 30,000 framework above. Resist editing any of them for a full seven days, since every edit resets the learning phase and quietly resets your cost per result along with it. After seven days, compare cost per result across all three and move the next week's budget toward whichever is winning, rather than pausing the others outright, since a second-place ad set today can become useful again once the top one shows fatigue.

Common Meta ads targeting mistakes

• Targeting an entire city or state on a daily budget too small to get meaningful delivery across that many people.
• Stacking ten or more interests, believing it narrows the audience, when Meta's OR logic actually widens it.
• Never excluding existing customers from cold prospecting campaigns.
• Editing a live ad set's targeting instead of duplicating it, resetting the learning phase every time.
• Switching to Advantage+ audience before the pixel has enough conversion history to make it useful.
• Building a lookalike from every website visitor instead of from actual paying customers.
• Ignoring the placement and demographic breakdown reports, so a shrinking audience segment goes unnoticed for weeks.

If your targeting is correct by every measure above and results are still weak, the audience is rarely the remaining problem. Read why your Meta ads are not converting for the landing page, tracking and offer checks that usually explain it, and if cost itself is the concern rather than conversion, how to reduce cost per lead covers the other levers beyond targeting.

Frequently asked questions

What is the best Meta ads targeting option for a small business with a low budget?

For budgets under Rs 20,000 a month, start with a custom audience of your existing customers, website visitors or WhatsApp contacts, then build a 1 percent lookalike from it. That gives Meta the strongest signal available without needing a huge audience. If you do not have enough data yet for a custom audience, use core audiences with two or three specific interests plus a tight location radius, keeping the total audience between 2 and 8 lakh people. Anything smaller limits delivery and raises cost per result; anything broader wastes budget on people who were never going to buy.

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

It depends on your pixel data. Advantage+ audience works well once your pixel or Conversions API has recorded 50 or more conversions in the past 30 days, because Meta's algorithm has enough signal to find the right people on its own. Below that, manual targeting with a defined radius, age range and two to four interests gives you more control and usually a lower cost per result in the first month. My approach for a new account is to run manual targeting for four to six weeks, build up conversion data, then test Advantage+ in a separate ad set and compare cost per result before shifting budget.

What is the difference between radius targeting and city targeting on Meta ads?

Radius targeting draws a circle of a set distance, typically 3 to 10 km, around a pin you drop on the map, usually your shop or clinic. City targeting selects the entire municipal area, which for Pune plus Pimpri-Chinchwad is well over 7 million people. For a local business, radius targeting wastes far less budget because it excludes people who will never travel to you. City targeting only makes sense when you deliver to the whole city, such as home services with citywide coverage, or when you are running a brand awareness campaign rather than a lead campaign. I default every local client to radius targeting first and only widen it if delivery is slow or frequency climbs too fast.

How do I build a custom audience from my WhatsApp contacts?

Export your WhatsApp Business contact list or customer database as a CSV with phone numbers, then upload it in Ads Manager under Audiences as a Customer list custom audience. Meta hashes the numbers and matches them against user accounts, and match rates for Indian phone numbers typically land between 40 and 60 percent. Once the audience is built, you can advertise directly to it for repeat business, exclude it from cold campaigns so you are not paying to reach existing customers, or use it as the source for a lookalike audience. Refresh the list every 60 to 90 days so it keeps including your newest customers.

How often should I change my Meta ads targeting?

Do not change targeting every few days. Give each audience at least 7 days and 50 results before judging it, because Meta's algorithm needs that window to learn who converts. Resetting the audience restarts the learning phase and typically raises cost per result for several days afterward. Review performance weekly. If cost per result is climbing after two full weeks, build and test a new audience in a separate ad set rather than editing the live one, then pause the underperformer once the new ad set has its own stable data.

Related guides

• Meta Ads Learning Phase: How Long It Lasts and How to Exit
• Instagram ads cost in India: 2026 benchmarks by objective and industry to budget against once your targeting is set.
• Facebook and Instagram ads for local business: the full campaign walkthrough this targeting guide plugs into.
• How to write Meta ad copy: the creative half of the broad-targeting-plus-strong-creative approach.
• Why your Meta ads are not converting: for when targeting is right but the funnel after the click is not.
• Meta ads vs Google Ads for Pune businesses: which platform fits your budget and intent.

Your next step

Open your best-performing ad set and check two things: whether it excludes your existing customers, and whether it is a radius or a whole city. Those two settings alone explain a large share of the wasted budget I find when auditing new accounts. Fix them before building anything new. If you want a second opinion on your current targeting structure, send me the account and I will tell you which of the layers above is missing.