Google Ads
A sudden ROAS drop in Performance Max usually has one dominant cause, not seven at once — the job is ruling causes out in the right order instead of guessing. Check whether tracking changed, then whether the drop is even real or just reporting lag, then feed changes, competition, recent bidding or budget edits, stock-outs, and a shift in brand versus non-brand traffic mix.
The same principle that applies to a PMax campaign with no sales at all applies here: PMax's automated bidding optimizes against whatever conversion value it's reported, and a tracking change often looks exactly like a performance drop. A theme update, a checkout change, an app install or removal, a consent banner edit, or a Google Tag Manager container change can all quietly break or degrade purchase tracking on the exact day performance appears to fall off a cliff. Check your Purchase conversion count against actual Shopify orders for the same window before looking anywhere else — see Google Ads conversion tracking for Shopify for what typically causes this.
Google Ads records a conversion against the date of the click, not the date the purchase actually completed, and conversions can take time to fully attribute and appear in reporting. The last three to seven days in any Google Ads view are almost always under-reported relative to where they'll settle once conversion lag catches up. If the “drop” you're seeing is concentrated in the most recent few days, wait for that window to finish reporting before treating it as a real trend rather than an artifact of how the data fills in.
Check Merchant Center Diagnostics for a recent jump in disapproved or unavailable products, a broken feed sync, or a recent attribute or category change that could have shifted how products are matched to queries. A feed that quietly loses a batch of previously-approved products doesn't announce itself in Google Ads reporting — it just shows up as fewer conversions from a smaller eligible catalog. Full detail on disapproval causes is in Google Merchant Center product disapprovals.
More advertisers bidding into the same auctions raises CPCs without any change on your end, which compresses ROAS even at a stable conversion rate. Seasonal demand shifts do the same in the other direction — a category with a sharp seasonal peak will often show a ROAS decline right as that peak passes, which can look like a tracking or account problem when it's actually a demand curve doing what demand curves do.
Check the account's change history before assuming the cause is external. Raising a Target ROAS pushes the algorithm toward fewer, higher-confidence auctions — that can look like a performance drop in aggregate even though the campaign is behaving exactly as instructed. Cutting budget mid-flight disrupts pacing and can force the algorithm to re-learn spend patterns. Either change made shortly before the drop is a strong candidate for the actual cause.
If a handful of bestsellers drove a disproportionate share of conversions and they went out of stock, PMax doesn't sit idle — it shifts spend toward whatever else in the feed is still available, including products with historically weaker conversion rates. Nothing about tracking, bidding, or the account changed; the campaign's best-converting inventory simply stopped being sellable. Cross-check inventory levels on your top products against the timing of the drop.
Brand search typically converts at a higher rate and lower cost than non-brand traffic, because the customer already knows what they're looking for. If brand's share of PMax traffic drops — from seasonality, a competitor bidding on your brand terms, or simply less brand demand in a given week — the blended ROAS falls even if non-brand performance itself hasn't changed at all. The Insights section in Performance Max surfaces some of this split; it's worth checking before concluding the whole campaign got worse, when it may just be carrying a harder mix of traffic than before.
Whichever cause looks most likely, it helps to confirm exactly when the drop started against your actual Shopify revenue, not just Google Ads' own reporting — the two don't always move in lockstep, and a drop that's sharper in Google Ads than in Shopify usually points back at tracking or reporting lag rather than a genuine sales slowdown. Attribix shows Google Ads' reported numbers next to actual Shopify orders day by day, which makes that comparison quick instead of a manual export exercise. See Google Ads conversion tracking for Shopify.
We'll trace the drop back to its actual cause — tracking, feed, bidding, or the market — against your real Shopify orders.