What is the minimum budget to run Meta ads in India?

Technically you can run at Rs. 100-200/day, but that rarely produces stable results because the ad set cannot reach the ~50-events-per-week threshold Meta needs to exit the Learning Phase. A more realistic testing budget is Rs. 500-1,000/day for a couple of weeks. Remember to add 18% GST on top of whatever media budget you set.

How long before I see results?

Lead and WhatsApp campaigns often show readable signal within 3-7 days; e-commerce purchase campaigns usually need 7-14 days to gather enough Purchase events to optimise. The real rule is "enough conversions to exit Learning," not a fixed number of days. Avoid major judgments in the first few days.

Facebook ads or Instagram ads?

They run through the same Meta Ads Manager and the same targeting. Leave Advantage+ Placements on so Meta delivers wherever it is cheapest for your goal, then read the placement breakdown after a week. As a generalisation, Instagram/Reels skews younger and visual; Facebook feed often does well for local services and older demographics, let the data, not the assumption, decide.

What is a "good" cost per lead in India?

It varies widely by industry, city, and offer, so the only honest answer is "below your break-even." Local services often see lower CPLs than B2B simply because intent and offer differ. Calculate what a customer is worth to you and what share of leads convert, then your acceptable CPL falls out of that maths rather than a benchmark.

How do I track whether my ads are really working?

Install the Pixel plus the Conversions API, define the events that matter (Lead, Purchase), and in Ads Manager track cost per result and ROAS against your margins. Then reconcile against your own backend, payment gateway, Shopify/WooCommerce orders, or CRM, because in-platform attribution and your ground truth will not always match.

Do Meta ads work for a purely local business with no website?

Yes. Engagement (Click-to-WhatsApp) and Leads (Instant Form) campaigns work entirely on Meta's own data and a phone number, no website required. Tight radius or pincode targeting plus fast WhatsApp follow-up is a common, effective setup for clinics, salons, tuition centres, and repair services.

Meta Ads for Small Businesses in India: Complete Budget and Strategy Guide 2026

Meta ads (Facebook, Instagram, and Messenger placements managed through one Ads Manager) are usually the fastest paid channel for an Indian small business to start generating enquiries, because they let you reach people who are not searching for you yet. The trade-off versus Google Search ads is intent: a Google searcher already wants a plumber; a Meta user is scrolling Reels and you have to interrupt them. That single difference shapes every decision below, from objective to creative.

This is a practitioner's walkthrough of the actual settings, menus, and numbers that matter for a business spending Rs. 10,000 to Rs. 1 lakh a month. Where you see a rupee figure, treat it as a typical range, not a guarantee, your real costs depend on industry, city, offer, and creative.

First, the account setup most businesses skip

Before a single rupee of spend, get these right in Meta Business Suite and Ads Manager. Skipping them is the single most common reason Indian SMB accounts underperform or get disabled.

Create a Business Portfolio (Business Manager), not a personal ad account. Go to business.facebook.com, create the portfolio, then add your ad account, Page, Instagram account, and Pixel inside it. A personal ad account can be lost with your profile and cannot be properly shared with an agency or staff.
Verify the business and set up two-factor authentication. Business verification (Business Settings → Security Center) reduces sudden ad-account restrictions and unlocks higher spend limits.
Install the Meta Pixel and the Conversions API. The browser Pixel alone now loses a meaningful share of events to ad-blockers and iOS tracking limits. Pairing the Pixel with the Conversions API (server-side) recovers signal. On WordPress/WooCommerce or Shopify the official plugin sets up both; on a custom site, use a Google Tag Manager container plus a server-side endpoint.
Set up the events you actually care about in Events Manager: Lead, Purchase, AddToCart, InitiateCheckout. Then configure Aggregated Event Measurement (Events Manager → your Pixel → Aggregated Event Measurement) and put your most important event at the top, because iOS users are capped at 8 prioritised events.

Without the Pixel firing a real conversion event, you cannot run a Sales/Conversion campaign properly, you cannot build website-visitor audiences, and you are effectively flying blind. This is non-negotiable.

Picking the right objective (and what Meta renamed them to)

Meta consolidated the old objective list into six outcomes under the ODAX framework. You will see these six when you click Create in Ads Manager:

Awareness — reach and brand recall. Useful for a launch or an event, rarely the right starting point for a business that needs leads.
Traffic — clicks to a website, app, or WhatsApp. Cheap clicks, but clicks are not customers; use sparingly.
Engagement — this now also houses the "Click to WhatsApp" / Messages options, which matter a lot in India (covered below).
Leads — Instant Forms (on-platform forms) or website leads. The workhorse for service businesses.
App Promotion — only if you have an app.
Sales — optimises for purchases/conversions; the default for e-commerce and D2C.

For most local and service businesses, start with Leads (Instant Form or Click-to-WhatsApp). For e-commerce and D2C, start with Sales optimised for Purchase. Do not start with Awareness, you will get a cheap-looking CPM and zero business outcomes, then conclude "Meta ads don't work."

Click-to-WhatsApp: the India-specific objective worth testing first

In India, a phone number on a landing page often converts worse than a chat. Meta lets you run a Click-to-WhatsApp ad (under the Engagement objective, optimised for conversations) where tapping the ad opens a WhatsApp chat with your business number, pre-filled with a message you write.

Why it tends to work here: there is no form friction, the lead lands in an app the user opens dozens of times a day, and you can reply with a catalogue, price list, or location pin. To set it up you need a WhatsApp number connected to your Business Portfolio (a free WhatsApp Business app number works; the WhatsApp Business Platform/API is for higher volumes). Suppose a Bhopal home-bakery runs a Click-to-WhatsApp ad with the pre-filled message "Hi, I want the festive hamper price list", every reply arrives as a warm conversation rather than a cold form fill. Treat this as a hypothesis to test against an Instant Form, not a guaranteed winner.

Setting a budget that can actually learn

Meta needs data to optimise. Its delivery system tries to exit the Learning Phase at roughly 50 optimisation events per ad set per week (50 leads, 50 purchases, etc., for the event you are optimising for). If your budget cannot realistically produce ~50 of your chosen event a week, the algorithm never stabilises and costs stay volatile. This is the real reason "Rs. 100 a day" usually fails, not that the number is small, but that it cannot reach the learning threshold.

Practical budget bands, framed as typical Indian-market starting points:

Testing / first 1-2 weeks: Rs. 500-1,000 per day. Enough to generate readable data on creatives and audiences within about a week. Below ~Rs. 300/day, expect noisy, unreliable results.
Local services (salon, clinic, tuition, restaurant): typically Rs. 500-2,000 per day on Leads or Click-to-WhatsApp.
Lead-based B2B (consultancy, coaching, agency): typically Rs. 1,000-3,000 per day; optimise for lead quality, not lead count, by adding qualifying questions.
E-commerce / D2C: typically Rs. 2,000-10,000+ per day, but the right number is whatever your ROAS and margins justify, not a fixed figure.

A useful sanity check before you start: work backwards from unit economics. If your average order value is Rs. 1,200 and your gross margin after product cost, shipping, COD/RTO losses and packaging is ~40% (Rs. 480), then any CPA below ~Rs. 480 is profitable. That number, not a guru's "good ROAS", tells you what to allow.

Campaign budget (Advantage+ CBO) vs ad-set budget

Meta's two options are Advantage Campaign Budget (set the budget once at campaign level and let Meta split it across ad sets, formerly CBO) and ad-set budgets (control each ad set yourself). Start with the campaign-level budget for simplicity. Switch to ad-set budgets when you specifically want to protect spend on a small high-value audience (for example, a tiny retargeting pool) that CBO would otherwise starve.

GST and billing note for Indian advertisers

Meta bills Indian advertisers through an Indian entity, so 18% GST applies on your ad spend. If you have a GSTIN, add it in Ads Manager → Payment Settings → Business Information so the tax appears on a proper invoice you can claim as input tax credit. Budget for the GST on top of your media spend, your Rs. 1,000/day plan is really ~Rs. 1,180/day out of pocket.

Building audiences that convert

"Everyone in my city, 18-65" is not targeting. Build in this order of warmth.

Custom Audiences (warmest, retarget these)

Website visitors (needs the Pixel): build separate audiences for high-intent pages, people who hit your pricing or product page convert very differently from a blog reader.
Customer list: upload existing customers' phone numbers/emails (hashed automatically). In India a phone-number list often matches better than email. Use this for repeat-purchase and upsell ads, and as a seed for Lookalikes.
Engagement audiences: people who engaged with your Instagram/Facebook Page or watched your video, built entirely from Meta's own data, so they work even with no website.
Lead-form openers/submitters and Instagram Reel viewers: excellent mid-funnel retargeting pools.

Lookalike Audiences (to scale)

Create a Lookalike from your best source, ideally actual purchasers or qualified leads, not all website traffic. Start at 1% (most similar) for your country, and only widen to 2-3% once the 1% audience is exhausted or you need more volume. A Lookalike built from a list of 1,000 high-quality customers beats one built from 50 random visitors every time, so feed it good data.

Broad / Advantage+ targeting

With a working Pixel and clean conversion events, Meta's broad and Advantage+ Audience targeting (where you give a suggestion and Meta finds buyers) often performs as well as hand-picked interests, because the algorithm has enough signal to find converters itself. If you are brand new with no data, start with 2-3 stacked, genuinely relevant interests plus location and age, then move toward broad as conversions accumulate.

Location targeting, the India detail people miss

In the location box, change the default "People living in or recently in this location" to "People living in this location" for delivery-based e-commerce, otherwise you pay to reach tourists and people just passing through. For a physical store, use the dropped-pin radius (you can go as tight as ~1 km in dense areas). For serviceable pincodes, you can paste multiple pincodes or upload a location list rather than targeting a whole city.

Creative: what actually stops the scroll in India

On Meta, creative is usually the biggest lever, you can fix a mediocre audience with a great ad far more easily than the reverse. General field truths that hold up for Indian feeds:

Shoot vertical 9:16 for Reels and Stories placements, and supply a 1:1 or 4:5 version for the feed. Let Advantage+ Placements deliver across all of them.
Win the first 1-3 seconds. Open with the offer, a question, or a relatable problem, not a logo intro. Most users decide to keep scrolling almost instantly.
Burn in captions/subtitles. A large share of feed video is watched on mute, so your message must land visually. For multilingual markets, Hinglish or the local language in on-screen text often outperforms formal English.
Raw and authentic usually beats glossy. A real owner talking to camera, an actual product demo, a genuine customer review (with permission) tends to feel native to the feed and avoids the "this is an ad" reflex.
Run 3-5 distinct creatives, not five tiny variations. Test different angles (problem-led vs offer-led vs testimonial), not just different fonts.

Formats by business type

E-commerce: Carousel for multiple SKUs or feature-by-feature; Advantage+ Catalog ads (formerly Dynamic Product Ads) once your product feed is connected, these auto-show the exact product a user browsed.
Services with no online checkout: a short demo/testimonial video driving to an Instant Form or WhatsApp.

Instant Forms: keep them short, qualify hard

Ask for Name, Phone, and one qualifying question (budget, location, or service needed). Every extra field lowers volume.
Choose the "Higher intent" form type (adds a review step) when you are getting junk leads; choose "More volume" when you need scale and your team can filter.
Connect the form to your CRM or a Google Sheet via a tool like Zapier/Make, or download leads from the Page's Lead Center, manual checking once a day means cold leads.

A simple funnel you can run from day one

Rather than one campaign, split spend by temperature. Rough allocation as a starting point, adjust to your data:

Prospecting (cold): ~50-60% of budget. Broad or Lookalike or interest audiences, optimised for your real conversion event. This is where you find new buyers.
Retargeting (warm): ~30-40%. Website visitors, video viewers, lead-form openers, and Instagram/Page engagers, with a stronger offer or social proof.
Retention (existing customers): a small slice. Your customer-list audience for repeat purchase and upsell, usually your cheapest conversions.

Keep retargeting audiences excluded from prospecting (and existing customers excluded from both) so you are not paying to re-acquire people you already have.

The mistakes that quietly drain budget

1. Editing or restarting ad sets constantly

Every meaningful edit (budget, audience, creative, optimisation event) resets the Learning Phase and costs you data. Set it up properly, then leave it alone for several days unless something is clearly broken.

2. Too many ad sets splitting the budget

Ten ad sets at Rs. 100/day each will all stay stuck in learning. Fewer, better-funded ad sets reach the ~50-event threshold faster. Consolidate.

3. Judging on day two

Give a campaign enough time and volume to exit Learning before you decide. For low-volume lead businesses, that can be one to two weeks. Reacting to a single bad day is how good campaigns get killed.

4. Ignoring mobile and page speed

The overwhelming majority of Indian Meta users are on mobile, often on patchy connections. If your landing page is slow or the form is fiddly on a phone, you have already lost most of the click. Test on an actual phone on mobile data, not just office Wi-Fi, before you spend.

5. No follow-up system for leads

The fastest way to waste ad spend is to generate leads and reply hours later. Speed-to-lead matters: have a WhatsApp template ready, route leads to a person or an auto-reply immediately, and track which leads actually become customers.

6. Policy trips that get ads rejected

Common Indian-SMB rejections: "before/after" body or skin claims, "personal attributes" wording ("Are you struggling with debt?"), unverified health/financial promises, and landing pages that do not match the ad. Read the ad, set a non-overlapping conversion budget, and keep the landing page consistent with the offer to avoid the review flagging it.

Measuring what actually matters

Ignore likes and reach as success metrics. In Ads Manager, customise columns to show the numbers tied to money:

Cost per result (CPL for leads, CPA for purchases) versus the break-even number you calculated from your margins. This is the headline metric.
Lead-to-customer rate: track in a sheet how many Meta leads actually close. 200 cheap leads that never buy are worse than 40 that do, your CPL can look great while the business loses money.
ROAS for e-commerce (revenue from ads / ad spend). Whether 2x or 4x is "good" depends entirely on your margin, not on a benchmark, so define your own break-even ROAS.
Frequency: if it climbs high on a small audience, your creative is fatiguing or your pool is too narrow, refresh creative or widen the audience.
Quality / Engagement / Conversion ranking: shown at the ad level. Repeated "Below average" usually signals a creative-to-audience mismatch, not a bidding problem.

One caveat on attribution: Meta's reported conversions can over- or under-count depending on your attribution window and iOS limits. Cross-check against your own backend (Razorpay/PayU/Cashfree dashboard, Shopify orders, or your CRM) before you trust the in-platform ROAS completely.

Scaling a winner without breaking it

Budget, gradually: raise the winning campaign/ad set by roughly 20-30% every few days. Doubling overnight throws it back into Learning and costs usually spike.
Audience: widen the Lookalike (1% → 2-3%), add nearby cities, or move toward broad/Advantage+ once you have conversion data feeding the system.
Creative: your best ads will fatigue, so keep a pipeline of fresh angles ready. Scaling is usually a creative problem, not a bidding one.
Duplicate vs raise: for aggressive scaling, some advertisers duplicate the winning ad set into a fresh campaign rather than editing the original, this keeps the proven ad set stable while the duplicate explores a bigger budget.

If you would rather have this set up and managed properly, our team does exactly this, see our services or get in touch.