Most budget guides for testing a new market give you one number: spend enough to hit 50 conversions a week per ad set, because that is what the platform's own learning phase wants. It is a fine rule if you sell a ₹1,500 skincare product with a two-week repurchase cycle. It falls apart the moment your AOV climbs, because most brands selling furniture, premium apparel, or anything over roughly ₹8,000 to ₹10,000 a unit simply cannot generate 50 purchase events in a week in a brand new market, and trying to force that volume usually means burning budget on the wrong audience just to hit an arbitrary number.
The fix is not a bigger budget. It is a different way of deciding when you have learned enough.
Why the 50-conversion rule breaks for higher AOV
Meta's stated learning phase threshold exists to give the algorithm enough data to optimize delivery. It is a genuinely useful number for high-frequency, lower-AOV categories where 50 purchases a week is a realistic, even modest, target. For a high-AOV brand, chasing that same number usually means one of two bad outcomes: spending far more than the test actually needs to answer the question, or forcing the algorithm toward cheaper, lower-intent traffic just to inflate the conversion count.
The better framing, borrowed from how experienced media buyers actually run creative and market tests, treats the learning phase as a delivery signal, not a testing signal. The real decision threshold is spend reaching roughly three to five times your target cost per acquisition (CPA) per ad set. At three times target CPA you get a fast, directional read. At five times target CPA you get a cleaner yes or no. Neither requires 50 purchases if your AOV and margin mean target CPA is high enough that five purchases already represents meaningful spend and a meaningful signal.
Start with CPM, not with a budget number
Before setting any budget, size the market itself. CPM (cost per thousand impressions) tells you what attention costs in that market, and it varies enormously by geography. Current benchmarks put Meta CPM in the US in the sixteen to twenty-three dollar range, against roughly one dollar fifty to three dollars in India, a gap of eight to fifteen times depending on the source and time of year. That gap alone should reset any assumption carried over from an India-based test.
The calculation sequence:
Step one: estimate impressions per dollar. At a twenty dollar CPM, one thousand dollars buys roughly fifty thousand impressions. At a two dollar CPM, the same spend buys five hundred thousand.
Step two: apply an expected click-through rate and conversion rate. US ecommerce Meta CTR typically runs around one and a half to two percent, with landing page conversion rate varying widely by category and price point, often one to three percent for cold traffic on a new market with no brand recognition yet.
Step three: translate that into cost per conversion, and check it against your economics before spending anything. Fifty thousand impressions at 1.7% CTR is roughly 850 clicks. At a 1.5% landing page conversion rate, that is about 13 conversions from one thousand dollars of spend, an implied cost per acquisition north of seventy five dollars before you have adjusted for a brand-new market's typically lower cold-traffic conversion rate.
If that implied CPA is already close to or above what your margin can support at your AOV, that is information worth having before you commit a real test budget, not after.
Two worked examples
Low-AOV brand (skincare, ₹1,800 AOV, healthy repeat rate). Target CPA of ₹700 based on existing India economics, adjusted upward for a colder new-market audience to roughly ₹900. Weekly budget to hit fifty conversions: (₹900 × 50) ÷ 7 ≈ ₹6,400 a day per ad set. This is realistic. The category converts often enough that fifty purchases a week is an achievable, meaningful number, and the 50-conversion threshold is the right one to use.
High-AOV brand (furniture, ₹18,000 AOV, considered purchase). Fifty conversions a week at a five thousand rupee target CPA would mean ₹35,700 a day, a budget most brands testing a brand new market have no business committing before they have any signal at all. Using the three to five times CPA threshold instead: at five thousand rupees target CPA, five times CPA is ₹25,000 of spend per ad set, enough for a directional signal from roughly five conversions, a realistic weekly test budget of ₹3,500 to ₹4,000 a day rather than ten times that. The test is smaller, faster to fund, and still tells you what you need to know.
Free Profitability Review
Not sure if your ROAS is actually profitable?
We review Meta, Google, GA4, and Shopify data together and tell you where the numbers are misleading you, before you spend another rupee scaling on a false signal.
Request a Paid Media Profitability ReviewWhat changes when you're testing India to US specifically
The CPM gap is the headline number, but it is not the only thing that moves. US audiences convert differently on cold traffic than India audiences do, partly because competitive density is far higher (more advertisers bidding for the same attention drives CPM up further, particularly in Q4 when costs typically run thirty to sixty percent above the annual average). Currency and payment readiness also need to be sorted before the first rupee of test spend goes out, a US-facing ad account, a payment gateway that handles USD cleanly, and a checkout that doesn't quietly assume INR-only shoppers.
None of this changes the calculation method above. It changes the specific numbers you plug into it. Always pull current CPM for the specific market and objective you're testing, since these figures move meaningfully year over year and even quarter over quarter, rather than anchoring to whatever number a case study used a year ago.
The test window
Seven days is a reasonable default for most categories. High-ticket or long-consideration products need ten to fourteen days, since a furniture shopper comparing options over two weeks will not show up in a seven-day window even if the campaign is working. Judging a considered-purchase category on a seven-day read is judging it before it has had a fair chance to convert.
Before you commit real budget
Run the CPM to cost-per-conversion math first, using current benchmark numbers for the specific market and platform. If the implied CPA already exceeds what your margin supports at your AOV, that is a pricing or positioning problem to solve before spending, not a budget size problem. The launch checklist for entering a new country covers the other pieces, localization, target CAC, and retention channel choice, that a budget calculation alone will not catch. And once you have real spend data, CAC payback period is the number that tells you whether what you're recovering is fast enough to sustain the test.
FAQ
Do I really not need 50 conversions to know if a market test is working?
Not for every category. Fifty conversions a week is Meta's own threshold for exiting the automated learning phase, and it is a realistic target for high-frequency, lower-AOV brands. For high-AOV or considered-purchase categories, spend reaching three to five times your target CPA per ad set gives a meaningful directional signal well before you would reach fifty purchases, and forcing volume to hit that number usually means testing the wrong audience.
How do I estimate CPM before I've spent anything in a new market?
Pull current published benchmarks for the specific country and platform, since CPM varies significantly by geography, competitive density, and season, and figures from even six months ago can be meaningfully out of date. Treat published benchmarks as a starting range to plan around, then correct against your own account's actual delivery once spend starts.
Should I use the same target CPA in a new market as I use at home?
No. Start from your home-market target CPA and adjust it for the new market's CPM and expected cold-traffic conversion rate, which is usually lower in a market where your brand has zero recognition yet. Using an unadjusted target CPA is one of the most common ways new-market tests get judged as failing when the budget was simply undersized for that market's real costs.
What if my implied cost per conversion is higher than my margin can support?
That is a signal to solve before spending a real test budget, not after. Either the price and bundle need to change for that market, or the category itself may not be viable there at the CPM and conversion rates you're seeing. Spending anyway to "see what happens" usually just confirms what the math already told you, at a real cost.
How long should I run the test before deciding it worked or didn't?
Seven days for most categories, ten to fourteen days for high-ticket or long-consideration products where a customer's decision cycle genuinely takes that long. A new market has no account history, so early data is noisier than an established market's, and judging it too early usually means judging it before it had a fair chance to convert.
Not sure what your test budget should look like for a specific market? Talk to us at growth@adtitudemedia.com.