If you already run Shopping ads, the question isn't "how do I launch a campaign." It's how to arrange the campaigns you have so spend flows to the products that earn, learning stays stable, and your reported ROAS tells the truth. That's what structure is for.
Most ranking guides stop at the hierarchy diagram. This one walks the hierarchy, the campaign-type portfolio, and the two things those guides skip: deriving your target from break-even margin, and keeping Performance Max from eating your brand.
The hierarchy every Shopping campaign inherits
Standard Shopping has three levels: the campaign, the ad group, and the product group. The campaign is the strategy layer — budget, bid strategy, and priority. The product group is where you split your feed into buckets you can bid on separately.
Shopping has no keywords. Your product title is the query-matching surface, so Google decides which searches trigger a product based on your Merchant Center feed data — titles, GTINs, product type, price, and availability. Front-load brand and product type in titles; a clean feed is the load-bearing input, not the bid.
Product groups let you subdivide by category, brand, product type, or a custom label you invented (say, margin tier). The practical move for an operating store: split your best sellers into their own product groups so you can bid them separately from the long tail.
Standard Shopping vs Performance Max: a portfolio, not a choice
Treating this as either/or is behind the field. In Optmyzr's study of 24,702 Performance Max campaigns, 82% of advertisers ran PMax alongside Shopping or Search, and accounts that split budget between the two posted the strongest returns in the sample. Run both; they do different jobs.
Standard Shopping is the manual-control counterpart: product-group bids, campaign priorities, negative keywords, and full search-term visibility. Performance Max is the automated one — you supply the feed, assets, and a target, and Google places across Search, Shopping, YouTube, Display, and more. For a deeper walkthrough of how PMax works, the Performance Max campaign overview covers the mechanics.
They now compete on Ad Rank in the same account — the old "PMax always outranks Shopping" behavior is gone. And despite the multi-channel pitch, smec (managing feeds for 350+ retailers) reports that 74–97% of PMax cost goes to feed-based Shopping-style ads. For a retailer, PMax is mostly a Shopping campaign wearing a bigger coat — so feed quality carries both.
Set your structure around break-even, not a benchmark
Here's the part the structure guides miss. Your target ROAS isn't a number you copy from a blog — it's arithmetic from your own margin:
Break-even ROAS = 1 ÷ contribution margin.
Say you sell a framed print at $48. Supplier and fulfillment run $19, your shipping share is about $4, and processing fees are roughly $2. Contribution = 48 − 19 − 4 − 2 = $23, a margin of 23 ÷ 48 = 47.9%. Break-even ROAS = 1 ÷ 0.479 = 2.09x, so your tROAS floor is about 209%.
Add a profit buffer — practitioners commonly use a multiplier of around 1.2 to 1.5 — and you'd set the target near 270%. That number is structural: it's the lowest return at which the campaign still clears a profit, and it belongs in the campaign setting, not a benchmark you half-remember.
Benchmarks are context, not decisions. LocaliQ's 2026 data put the retail-category search CPC at $4.14, and older WordStream figures compiled by Store Growers pegged Shopping clicks far below Search clicks (near $0.66 in 2016–2019 data — stale in absolute terms, directionally right). A $4 click is cheap at a 4% conversion rate and ruinous at 1% — that's your math, not the market's.
The target-vs-volume tradeoff that breaks most accounts
Raising your tROAS is a volume decision disguised as an efficiency decision. Google's own guidance on Target ROAS warns that setting a target far above what the account has achieved "may limit the amount of traffic your ads may get" — an aggressive target doesn't force efficiency, it declines auctions the system predicts won't clear the bar.
So structure your targets from trailing actuals: start at or slightly below your last-30-day real ROAS, then step up in roughly 10–20% moves. If you jump from 300% to 500% and spend collapses, that's the system working as designed, not a bug.
Volume also gates how many campaigns you can run. tROAS on Search and Shopping technically needs 15 conversions in the trailing 30 days, but the practitioner floor for stable bidding is higher — the widely cited field standard is around 50 conversions a month, and for PMax the 2026 consensus is roughly 30–50 conversions per campaign. Split a thin account into five campaigns and none of them learn.
Protect your brand terms from Performance Max
PMax chases the cheapest conversions, and branded searches are the cheapest of all. Left unprotected, it absorbs orders that were already coming from organic, direct, or your brand Search campaign, then reports a gorgeous ROAS for conversions it didn't create.
The cost is real. Practitioners estimate 8–15% of PMax budget leaks to brand queries in unprotected accounts, inflating apparent ROAS by about 15–30%. smec frames the mechanism bluntly: you end up paying "$1.50 for a click that you could have bought for $0.20".
The structural fix exists. Turn on account-level brand exclusions for PMax (available since 2024, covering variants and misspellings), run a dedicated brand Search campaign, and audit PMax search terms now that Google exposes them. Then judge PMax on non-brand performance, not blended numbers.
Zombie products and the forced-visibility campaign
PMax allocates toward proven converters, so a SKU with no conversion history gets no impressions, generates no data, and stays invisible — a loop that starves products that might sell. smec defines these "zombies" as products under about 10 clicks a month.
The structural answer is a dedicated Standard Shopping campaign that pulls zombies out and bids for clicks, forcing data collection. Budget it like R&D — it usually runs at a worse ROAS on purpose because it's buying information, not profit. Graduate the proven SKUs back into PMax once they have history.
How much structure your account can feed
Store Growers' widely cited budget tiers are a useful skeleton: under about $1K a month, favor one Standard Shopping campaign over PMax (not enough conversions to feed the machine); $1K–3K, one feed-only PMax; $3K–10K, one or two PMax campaigns if you clear 30+ conversions a month; above that, segment into PMax plus Standard Shopping.
The defensible logic underneath is conversion volume per campaign. Structure follows data, not ambition — don't build more campaigns than your order flow can keep out of learning mode. If this is all in service of a better return, the guide to increasing Google Ads ROAS goes deeper on the levers.
Where Victor fits
ROAS isn't profit — it ignores COGS, fulfillment, and fees, which is exactly why POD accounts can run a 5x and still lose money. That's the gap Victor closes.
Victor is the AI employee inside PodVector AI. He computes true per-order profit by seeing every order, every ad dollar, and every supplier cost from Printify, Printful, and Gelato together — so the break-even target above isn't a guess, it's a known number per SKU.
Victor operates your Google Ads directly: he pauses the campaign that fell below break-even, shifts budget toward the product groups carrying real margin, adjusts tROAS in measured steps, and adds the negative keywords that keep spend clean. He reprices thin-margin SKUs in Shopify and builds the Klaviyo flow that lifts repeat revenue, so paid acquisition doesn't carry the whole store. Put Victor on your account and he runs the structure, not just reports on it.
For the full picture of how paid search math fits an operating store, start with the Google Ads economics guide. If you'd rather hand the whole channel off, here's how a Google Shopping ads agency compares to running it in-house.
FAQs
Should I use Standard Shopping or Performance Max?
Both. In Optmyzr's 24,702-campaign study, 82% of advertisers ran PMax alongside Shopping or Search, and the split-budget accounts performed best. Use Standard Shopping for product-group control and forcing visibility on new SKUs, and PMax for scale once you have the conversion volume to feed it.
How many conversions do I need before splitting into more campaigns?
More than the technical floor suggests. tROAS on Search and Shopping needs 15 conversions in 30 days to function, but stable bidding is widely observed to need around 50 a month. Dividing a thin account into several campaigns leaves each one stuck in learning.
What target ROAS should I set?
Derive it from your margin, not a benchmark. Break-even ROAS = 1 ÷ contribution margin, so a 45% margin means a 2.2x floor; add a buffer and step toward it from your trailing actual. Setting the target far above what the account has achieved suppresses volume rather than creating efficiency, per Google's own guidance.
Why does Performance Max show a great ROAS while my revenue stays flat?
It's probably eating branded search. PMax chases the cheapest conversions, and brand terms are the cheapest — practitioners estimate 8–15% of PMax budget leaks to brand in unprotected accounts. Turn on brand exclusions, run a separate brand Search campaign, and judge PMax on non-brand orders.
Do I need a separate campaign for products that get no impressions?
Often, yes. PMax starves SKUs with no conversion history — smec calls products under about 10 clicks a month "zombies." Pull them into a clicks-bid Standard Shopping campaign to force data, then move the winners back into PMax.