Notes for owners · Digital marketing
Meta ads or Google ads for an online store: where to spend the first ₹50,000
Google catches shoppers already searching for your product; Meta makes people want something they were not looking for. A hypothetical ₹50,000 plan shows what to test first, what counts as a result and when to switch.
The GullySales team · Updated 17 Oct 2026 · 6 min read
On this page
You have ₹50,000 for ads and two people telling you opposite things. If shoppers already search for what you sell by name, spend first on Google Shopping ads for your best sellers. You are catching people who have already decided to buy something like it. If nobody searches for it, or it sells on looks, spend first on Meta, because you have to make people want it. Either way, do not split the money evenly from day one. Run one platform until you know your cost per order, then add the other.
Search catches the shopper. Meta has to create one.
On Google, an online store's best ad is the product itself. Shopping ads show the photo, the price and your store name to someone who typed "cast iron dosa tawa" and is comparing. On Meta, the ad has to stop a thumb mid-scroll. A product shot on a white background often does not. A reel of a crisp dosa lifting clean off a seasoned tawa might.
So the two platforms need different things from you: a clean product feed for Google, and a steady supply of video for Meta. The general difference is in Google Ads vs Meta ads; for a store, ask which of the two you already have.
Before the first rupee: what has to work
- Purchase tracking on both platforms, firing once per order, with the order value. Place an order yourself. It should show up once, not twice. Our guide to Meta conversion tracking covers the duplicate problem.
- A product feed with correct prices, stock and sizes, if Shopping ads are in the plan. A feed showing a size you sold out of last week wastes the click.
- The product page and checkout, tested on a phone over mobile data, delivering to a pin code far from your warehouse.
- Your break-even cost per order, written down: the margin on an average order before any ad spend.
Skip any of these and the ₹50,000 buys clicks you cannot judge.
A hypothetical ₹50,000 plan for a cast-iron cookware store
Imagine a small Rajkot brand selling pre-seasoned cast-iron tawas and kadais on its own store. Its average order is about ₹1,600, with a margin before ads of about ₹650. Every figure here is hypothetical. The point is the order of testing, not what any platform will deliver.
Before the first phase, write down how many orders a campaign must bring before you will judge it. Say 20. It is your own bar, not a platform rule. Then check the bar against the money: at a ₹650 break-even cost per order, ₹25,000 covers about 38 orders, so 20 is within reach.
| Phase | Money | Where | What you are testing | Move on when |
|---|---|---|---|---|
| 1 | ₹25,000 | Google Shopping on the ten best sellers | Cost per order from shoppers comparing similar tawas and kadais | 20 orders, or the money is spent |
| 2 | ₹18,000 | Meta, two or three reels to a broad audience in the cities you ship to | Whether video makes new people want cast iron, and at what cost per order | 20 orders, with one reel clearly cheapest per order |
| 3 | ₹7,000 | Meta retargeting with the catalogue, to visitors and cart abandoners | Whether reminders bring back people who looked | The cost per order there sits well under break-even |
If a phase spends its money without reaching the bar, stop it and do not top it up. It has answered the question: the cost per order there is too high.
Google goes first here because people already search for cast-iron tawas, so Shopping gives you a cost per order close to the floor. Meta has to create the want. Judge it against that Google number, not against zero.
Search on your brand name is left out on purpose. Few people type a new brand's name, so that money would sit unspent. Add it once the search terms show buyers asking for you by name. If nobody searches for your product at all, swap phases 1 and 2.
What counts as a result?
Cost per order, against your break-even. In the example, a first order that costs less than ₹650 in ad spend pays for itself. Above that it loses money, unless the buyer comes back. Add-to-carts on Meta and clicks on Google are steps, not sales, so do not judge either platform on them. If many of your orders are cash on delivery, count delivered orders, not placed ones; when COD helps sales and when it eats the margin explains why.
Break-even return on ad spend (ROAS) is the order value divided by the margin before ads: about 2.5 here. Below that, each sale costs more than it earns.
The mistake: splitting the money evenly from day one
For example, suppose the same cookware brand's first attempt split ₹50,000 down the middle. ₹25,000 went to Google across five campaigns, and ₹25,000 to Meta across five ad sets aimed at different audiences. Each campaign got too little money to bring more than a handful of orders. At the end the founder had ten small campaigns, none with a cost per order anyone could trust, and a conclusion: "ads don't work for us".
The fix was fewer, larger tests, one after another. One Google campaign ran until it reached the 20-order bar. Then one Meta campaign with three reels was judged against that cost per order. Same money, and a decision at the end of it. Google reached its 20 orders at about ₹890 each, above the ₹650 break-even. That was a loss, but a known one, and the Meta reels were then judged against it.
When this plan is wrong
If your category's Shopping results are full of cheaper marketplace listings, your ad will sit next to them. Unless your product is plainly different, that money does more on Meta or on the marketplace's own ads. If you have no video and no way to make any, Meta is the wrong first spend. And if the store converts badly from every source, neither platform will rescue it; why your ads bring clicks and no orders is the place to start.
₹50,000 is enough to learn which platform and which products work only when your break-even cost per order is a small fraction of each phase's money. At ₹650 against a ₹25,000 phase, it is. With a ₹3,000 break-even, the same phase buys about eight orders, and you learn little. Test fewer products, or wait until you can fund a larger test.
What to do next
Write down your break-even cost per order, then type your best product into Google the way a buyer would and see who comes up. That tells you which platform gets phase one. How each product category is bought is covered in the e-commerce sales and marketing section, and performance marketing is the service for running plans like this. To plan it around your own margins and catalogue, tell us what you need.
Related services
