Google Ads vs Meta Ads — which channel for a 2-crore brand?

Every second week, a founder somewhere between Gurgaon and Kochi sends me a version of the same question. "Our revenue is around two crore, we're ready to put a lakh a month into paid, should we go Google or Meta first?"

I have stopped answering the question as asked. It is the wrong question, and the founders who ask it are usually about to spend three months learning that the hard way. Let me try to explain what to ask instead, and then, because you did come here for a straight answer, what I actually recommend for most brands at that stage.

Why the channel question is the wrong question

The Google-versus-Meta framing assumes the two platforms are doing the same job with different labels. They are not.

Google Ads answers demand. Somebody is already typing "kids ergonomic study table under 15000" into a search box, and the ad you run is one of eight things they are choosing between. You are not creating the desire. You are winning the ten seconds of attention at the moment the desire has already surfaced. The ceiling on that traffic is fixed by how many people are looking for what you sell. In most Indian categories, that number is smaller than founders think.

Meta Ads creates demand. Somebody is watching a Reel of a dog wearing sunglasses, and a Story ad about the ergonomic study table interrupts the scroll. They were not looking for a study table thirty seconds ago. Whether the ad works depends on whether the creative can hold them for the four seconds it takes to remember that their daughter's back does hurt. The ceiling is the size of the audience you can plausibly interrupt, which for most consumer categories in India is much larger than what Google will send you.

If your product has substantial existing search demand — dental clinics, home loans, laptops, most B2B software — Google is doing the heavy lifting whether you like it or not. If your product is something people would want if they knew it existed, but nobody is searching for it yet — a new snack brand, a boutique jewellery line, a founder-led coaching program — Meta is doing the heavy lifting.

Most 2-crore brands sit somewhere in between, which is why they get confused. The question is not "which platform." The question is "which platform is doing which job for me, and how much of each do I actually need."

What a 2-crore brand's paid budget actually looks like

Ballpark, from working with about forty of them across seven years. A brand doing ₹2 crore a year in revenue can usually sustain paid spend somewhere between ₹75,000 and ₹2 lakh a month before the unit economics start to hurt. Below that you are subscale on both platforms. Above that, you are outgrowing the "2-crore brand" bucket and the maths starts changing.

At ₹1 lakh a month — the number in the founder's question above — you have about three thousand rupees a day of spend after platform fees. That is not enough to run five ad sets on Google, three campaigns on Meta, and a retargeting pool at the same time. You will spread yourself thin, none of the campaigns will hit statistical significance in a reasonable time, and you will spend the next quarter reading platform-generated recommendations that are noise, not signal.

The most common failure mode I see: brands split the budget fifty-fifty because it feels fair. Sixty thousand each way, plus a bit for retargeting. Everyone is busy. Nothing works well enough to scale. This is not a strategy. This is a compromise designed to avoid the real conversation.

What I actually recommend, by brand type

Three patterns cover most of what I see.

D2C consumer product, category with some existing search demand. Say a personal-care brand doing ₹2 crore, selling on Shopify with a small following. My default split at ₹1 lakh a month is roughly ₹65,000 to Meta, ₹25,000 to Google, ₹10,000 to retargeting. The Meta budget goes toward creative testing at the top of the funnel — five to eight new creatives a month, all short-form video, all shot on a phone. The Google budget goes almost entirely to branded search plus one narrow non-branded campaign for the highest-intent bottom-of-funnel keywords. If you are the Indian Golf Shop or a niche D2C brand in a low-search category, this is where you start.

Service business or professional practice with strong intent search. A chartered accountancy firm, a dental clinic chain, an ELV consultant. Flip the split. ₹70,000 to Google Search, ₹15,000 to Meta for retargeting and a very small awareness layer, ₹15,000 to Performance Max after month three, only if the account has enough conversion volume to feed the algorithm. The reason: people who need a chartered accountant are searching for one right now. Chasing them on Instagram before they've decided they need help is expensive theatre. The CA Ankur engagement is roughly this shape.

B2B software or high-consideration service. LinkedIn belongs in the mix here in a way it does not for the other two. My rough default: ₹40,000 to Google (search plus a narrow display remarketing loop), ₹35,000 to LinkedIn (single-image ads to a tightly-bounded ICP list), ₹15,000 to Meta for founder-led content and thought leadership, ₹10,000 to retargeting across everything. LinkedIn's cost per click is savage, but the quality of the person who clicks — director-plus, in the right function — is worth the tax if your average deal size covers it.

None of these splits are gospel. They are starting points. The interesting work happens in month three when the data starts telling you which of your assumptions was wrong.

Two accounts I watched go sideways this year

The first was a home-furnishings brand from Mumbai. Revenue about ₹1.8 crore, running ₹1.2 lakh a month split roughly evenly between Google and Meta. The founder was convinced Meta was the problem — the ROAS was 1.9, the Google ROAS was 3.4, obviously the answer was to move budget to Google. They had already made the move twice and each time revenue dropped.

The audit took a weekend. The Google account was harvesting demand that Meta was creating. When they cut Meta, Meta stopped creating demand, and two weeks later the Google searches for their brand name dropped 40 per cent. The Google campaigns kept running at a healthy ROAS on a smaller pool of searches, then that pool shrank, and total revenue tracked the shrinking pool down. The dashboards were lying by omission. We rebalanced spend back toward Meta, cut two vanity Google campaigns entirely, and revenue recovered inside six weeks.

The second was a B2B SaaS shop from Bangalore. Founder had been told LinkedIn was the answer and put ₹80,000 a month there for four months. The ICP list was correct. The creative was fine. The conversion rate on the landing page was 0.6 per cent. They had spent ₹3.2 lakh generating twelve inbound demos. The problem was not the channel. The problem was that the demo landing page was doing none of the qualifying, none of the trust-building, and none of the specific-outcome-storytelling that a director-plus visitor from LinkedIn expects to see before booking time on their calendar. We rebuilt the landing page in three weeks. The next month, on the same spend, they generated forty-two demos. The channel had been fine all along.

I mention both because most channel-mix questions are actually attribution questions, or landing-page questions, or offer questions, wearing channel-mix clothes. Fix those first.

The framework I actually use

Before I recommend a split, I ask the founder five things.

What proportion of your current revenue comes from people who searched your brand name to buy? If it is above thirty per cent, you have real brand demand and your Google account is probably harvesting it more than earning it. If it is below ten per cent, you have not yet built the brand demand and Meta is where you build it.

What is your gross margin? At sixty per cent plus, you can afford to test aggressively on Meta. At thirty per cent, you cannot afford much waste anywhere, and the higher-intent Google traffic is safer.

What is your creative-production capacity? Meta rewards volume — five new pieces a month is a floor, not a ceiling. If nobody in your team is going to make that many, the platform will underperform for you no matter what you spend. Google is much less creative-intensive.

Where does your customer actually decide? Some categories are decided in the search box. Some are decided in a WhatsApp group. Some are decided by a spouse looking at Reels. The channel that appears at the decision moment is the one that gets credit.

And the one nobody wants to answer: what is your patience horizon? Meta creative testing needs eight to twelve weeks to sort winning creatives from losers at ₹1 lakh a month of spend. If you and your investors need lift in six weeks, you are structurally set up to fail on Meta and should over-index to Google, take the smaller ceiling, and buy yourself the runway to do the Meta work properly later.

What most agencies get wrong here

Two things.

One, they optimise inside a platform without owning the mix across platforms. The Meta lead answers only for Meta ROAS. The Google lead answers only for Google ROAS. Neither is looking at whether the mix is right. Both will happily grow their own spend, because that is what their job is defined as. Somebody senior has to sit above the split and answer for total revenue, not platform ROAS. In most 2-crore accounts, that person is either the founder or an integrated agency that runs both platforms with one lead. If you have separate specialists on each, insist on a monthly conversation between them, chaired by you, about the split. It sounds obvious. Almost nobody does it.

Two, they measure the wrong thing. Platform-reported ROAS in 2026 is a fantasy. iOS attribution restrictions, cookie deprecation, and the growing share of purchases finished in-app or over WhatsApp mean the number your Meta dashboard shows and the number your Shopify dashboard shows will disagree by 30 to 50 per cent. The only ROAS that matters is the one calculated from your actual bank account. Build a simple weekly sheet: last week's paid spend across all platforms, last week's total revenue, last week's revenue minus your baseline organic revenue from the twelve months before you started paid, divided by spend. That is your real blended ROAS. Every other number is directional.

If you want a longer version of this argument, the paid-media page has our full point of view on measurement. The short version is: trust your bank, not your dashboards.

The straight answer, for the founder who asked

If you are a 2-crore consumer brand and you have to pick one to start, start with Meta, keep about twenty per cent aside for a narrow branded-search campaign on Google, and give the Meta creative testing three months before you judge it. That is roughly the pattern for most of the D2C shops we work with.

If you are a 2-crore services business and you have to pick one, start with Google Search on a small, ruthlessly-curated keyword list, put a light retargeting pool on Meta, and accept that the ceiling is what it is. You will grow more slowly than the D2C brand next door. You will also waste much less money.

If you are a 2-crore B2B software company and you have to pick one, do not start with either. Start with rebuilding your landing pages and your case studies. Then come back to the channel question. I promise you the answer will feel obvious by then.

A note on where paid fits inside the rest of the picture

Paid is a lever, not a strategy. If the product is not right, no channel will fix it. If the brand is invisible, no ROAS number will be meaningful. If your SEO foundation is weak, you will pay to acquire every visitor you get, forever. The brands we watch compound over three years are the ones that treat paid as one of three or four workstreams that all feed each other — paid, SEO, content, and CRO — not as a stand-alone growth engine.

If you have not read them yet, the two companion pieces to this one are the pricing post — which lays out what a serious paid-media retainer actually costs in India — and the Jagran Events case study, which walks through what the mix looked like for a very different kind of brand: not e-commerce, not B2B, but an events business at national scale. The numbers do not translate directly. The reasoning behind the split does.

Send us your account and current spend split via the contact form if you want a second pair of eyes on it. I read every brief, and if what you are running is fine we will tell you that too. Fewer things are more expensive in this work than fixing what did not need fixing.

— Shubhanshu