Set your daily budget to roughly your breakeven cost per acquisition, put all of it behind one campaign, and leave it alone for a fortnight. If your breakeven CPA is ₹550, that's ₹550 a day — about ₹16,500 for the month. Sizing a Google Ads budget for a new eCommerce store is not really a budgeting exercise; it's an arithmetic one, and the number falls out of your margin, not out of what you can afford to lose. The reason for the one-conversion-a-day floor is boring: Smart Bidding needs a meaningful number of conversions inside a 30-day window before its predictions are worth anything, and a store with no history has nothing to give it.
Below that floor, spend goes into learning and never comes out. We've watched ₹8,000-a-month accounts run for six months and produce a data set too thin to make a single decision from.
How to size a Google Ads budget for a new eCommerce store
Take a real product and work down the P&L. Say average order value is ₹1,800 on a skincare or apparel store:
- COGS at 55% → ₹990 gone, ₹810 gross
- Shipping, prepaid, ₹90
- Payment gateway at 2% → ₹36
- Packaging and inserts ₹15
- Blended allowance for returns and RTO, ₹120
₹810 minus ₹261 leaves ₹549 of contribution per order. That is your absolute ceiling on customer acquisition cost, and it means you break even, not that you make money. If you have no repeat purchase data yet — you don't, you're new — target 70% of it on first order. So ₹385 is the number you're actually aiming at, and ₹549 is the number at which you stop.
Now the budget. One conversion a day at breakeven is ₹549 daily, ₹16,470 for 30 days. Round to ₹18,000 so a couple of expensive days don't starve the back half of the month. For a store selling at $60 AOV in the US or UK with similar margins, the same maths lands somewhere around $30–$40 a day. Dubai stores with AED 250 baskets tend to sit higher because CPCs in the GCC are higher and the market is smaller.
If your contribution per order is under about ₹250, Google Search is a hard channel to win on in India at current CPCs. Say so out loud before you start, not in week five.
Days 1–3: spend nothing, fix the plumbing
Three days of setup with zero media spend is the highest-return part of the ramp.
Install the Google & YouTube channel on Shopify and push the feed to Merchant Center on day one, because disapprovals are slow. Missing GTINs, a price mismatch between the feed and the PDP, an unreachable returns policy page — each one costs a re-review cycle, and you can lose a week. Do it before you need it.
Set up conversion tracking twice and reconcile it: the Google Ads tag through the Shopify integration, plus GA4 with purchase events, then compare both against Shopify's own order count for a 48-hour window. They will not match exactly. If they're off by more than about 10%, find out why before you scale.
Turn on enhanced conversions. Then deal with COD, which is the part most new Indian accounts get wrong. If 60% of your orders are cash on delivery and 22% of those fail to deliver, then 13.2% of the orders you report to Google never became revenue. Feed Google the full order value and it will happily optimise toward whichever audience clicks cheapest and orders most carelessly. As a stopgap, send conversion value at 87% of order value. The proper fix is uploading conversion adjustments once delivery status is known, which is worth building by month two.
Blunt opinion: some stores are better off hiding COD from paid traffic for the first 30 days. You lose some volume. You get a clean signal, which at this stage is worth more.
Days 4–10: one Search campaign, bottom of the funnel only
Full ₹550/day into a single Search campaign. Not two. Not one per product line.
Eight to fifteen keywords, exact and phrase match, all of them the terms someone types when they have already decided to buy the category. "Buy cold pressed coconut oil online", "silk kurta set price", "stainless steel water bottle 1 litre". No broad match. No Dynamic Search Ads. Bidding on Maximise Clicks with a manual CPC cap set at roughly 15% of your target CPA — so ₹55–₹60 against a ₹385 target — because you want click volume on tightly controlled terms, and you want to know your own CPCs before you hand the bid to a machine.
Geography matters more than people expect in India. Delivery to tier-3 pincodes with 3–5 day transit converts worse and RTOs more. Start with the metros plus your five best-performing states from organic orders, and add out from there.
Seven days at ₹550 is ₹3,850 and roughly 65–70 clicks at a ₹55 CPC. At a 1.2% conversion rate that's less than one order. This is the uncomfortable middle of the ramp, and it's why the 30-day frame exists — a week is not a test, it's a warm-up. Check search terms daily and add negatives. That's the only lever you touch.
Days 11–17: add Performance Max, feed first
Around day 11, if search terms look sane and you've got at least a handful of conversions or add-to-carts, split the money. Raise total to about ₹800/day: ₹450 Search, ₹350 Performance Max.
Run PMax feed-only to begin with — no asset group, so no text or images for Google to spray across Display and YouTube while you have no idea what works. Feed-only keeps it close to Shopping inventory, which is where the intent is. Exclude your brand terms with a brand exclusion list so PMax doesn't eat cheap branded clicks and report them as wins. Set it to Maximise Conversions with no target CPA yet.
The awkward case: if your catalogue is under about 15 SKUs, PMax has very little to work with and often just becomes an expensive Shopping campaign. A Standard Shopping campaign is the better call there, and it gives you product-level bid control that PMax refuses to.
Days 18–24: change the bidding, not the budget
By day 18 you should have 12–20 conversions across both campaigns. That is enough to move Search to Target CPA, set at your actual observed CPA rather than your dream CPA. If you're getting orders at ₹620 and you set a ₹350 target, delivery collapses and you'll spend the next week wondering why impressions vanished. Set ₹600, then step it down 10% every four or five days.
Hold the budget flat through the bid change. Two variables at once and you learn nothing. We move budget in 20–30% steps and wait three days between moves, which is practice, not policy.
This is also the week to look at where the money is landing on your own site. A ₹55 click hitting a product page with a 3.4-second LCP on 4G is a donation. Check the paid traffic segment in GA4 for bounce and time-to-first-interaction, and if the PDP is slow, fixing it beats any bid adjustment you can make. Speed work on the landing template usually pays back faster than a campaign restructure at this stage.
Days 25–30: brand, remarketing, and an honest read
Add a brand Search campaign capped at ₹100–₹150/day. Yes, some of those clicks would have been free. You still want it, because competitors bidding on your name will show above your organic listing, and at ₹6–₹12 a branded click in India the insurance is cheap. Keep it in its own campaign so it never contaminates your non-brand CPA.
Add remarketing only if you've built an audience of 1,000+ in 30 days. Below that, list sizes are too small to serve and you're just carving up a budget that needed to stay concentrated.
Total for the month, on this shape: roughly ₹3,850 in week one, ₹5,600 in week two, ₹5,600 in week three, ₹4,900 plus brand in week four. Call it ₹21,000–₹24,000. That's the real cost of finding out whether Google Ads works for your store, and it is the cheapest version of that answer available.
What we don't run in the first 30 days
- Display and YouTube. Cheap impressions, no intent, and they'll flatter your assisted-conversion reports.
- Broad match with Target CPA and no history. It will find you searches you never imagined and charge you for all of them.
- Five campaigns at ₹300 a day each. This is the single most common self-inflicted wound. Every campaign learns separately, none of them get enough data, and month one produces a spreadsheet of noise.
- Target ROAS. It needs value data you haven't collected.
- Country expansion. One market, one currency, one shipping promise until the unit economics hold.
When a Google Ads budget is the wrong first spend
Open Keyword Planner and total the monthly search volume for your category terms in your target country. If it comes to a few hundred, nobody is searching for what you sell, and Search advertising cannot create demand that isn't being typed. That's a Meta prospecting problem, or a content and category-page problem. We'd rather tell a founder that in week zero than take a retainer for a channel that can't work yet.
Same verdict if your total monthly marketing budget is under about ₹15,000. Put it into category and comparison content and one Meta creative test instead. Google Ads with a starved budget is not a small version of Google Ads; it's a different, worse thing.
Day 31: three decisions
Pull the 30-day numbers and answer three questions with numbers, not feelings.
- Is CPA under breakeven? With our example store: under ₹549, scale by 25% and repeat the cycle. Between ₹549 and ₹800, keep spend flat and work on AOV and conversion rate rather than bids. Over ₹800 with 20+ conversions of data, this channel doesn't work at your current margin.
- Which campaign type earned it? If PMax carried the month, next month's split moves to 60/40 in its favour and you add an asset group. If Search carried it, expand keyword coverage before you touch anything else.
- What did the search terms teach you? Often the most valuable output of month one isn't orders. It's discovering that people search for your product by a name you don't use anywhere on your site.
If you want a second pair of eyes on the tracking setup and the landing pages before you switch on spend, our free store audit covers both. Bring your margin numbers with you; the conversation is short and useful when you have them, and vague when you don't.