Most Shopify merchants run Meta and Google in parallel without checking whether they are paying twice for the same warm audience. The split is real money: branded search clicks and Meta retargeting audiences overlap more than either platform tells you. The customer that converts on a Meta retargeting ad would have converted on a branded Google search anyway, and you paid CPC on both.
This is not a tracking problem. Both platforms report the conversion accurately within their attribution model. The leak is upstream: the warm audience pool is shared, and neither platform deduplicates against the other.
Where the overlap lives
Two specific surfaces leak the most.
Meta retargeting audiences built on site visitors or cart abandoners typically include every shopper who already knows your brand. Anyone who clicked a Google search ad in the last 30 days is in that pool. Anyone who came in via email, organic, or direct in the same window is in there too. The audience is warm by definition.
Branded search campaigns on Google capture clicks from people typing your brand name. The intent is high; the CPC is low. These are buyers who already decided to buy from you. They would have found the homepage on direct, the storefront link in your bio, or a tab they left open last week.
The overlap is the customer who sees a Meta retargeting ad on Monday, clicks through, browses, leaves, and converts on a branded Google search on Wednesday. Meta reports a click and an attributed view-through conversion. Google reports a click and a last-click conversion. You paid CPC twice for a customer the platform charts as two distinct conversions.
How much it actually costs
Pulling numbers from a baby apparel brand on Shopify, US$11,000 a month in blended paid spend, the deduplication math:
- Meta retargeting: US$2,400 spend, 312 attributed conversions
- Branded Google search: US$1,200 spend, 287 attributed conversions
- Verified overlap (Meta retargeting click then branded Google click within 14 days, both with same hashed email): 184 customers
Of the US$3,600 spent on the two surfaces, US$1,420 paid for clicks from the same 184 customers. Net: 13 percent of total monthly ad spend went to duplicate-purpose retargeting.
This is not a tracking error. The platforms are telling the truth. You are buying the same customer twice.
How to detect it
Three checks. No tool required if you have a CDP, or even a server-side conversion API setup with hashed-email join keys.
Audience overlap on the warm side. Export the Meta custom audience for retargeting (last 30 days site visitors). Export the list of branded search converters from Google over the same window. Hash both on email and intersect. The percent overlap is your duplicate-purpose surface.
Sequence audit. Pull the 30-day order log. For each order, check whether the customer journey shows a Meta retargeting click followed by a branded Google click within 14 days. Either platform's first-touch report can do this if you have the join on hashed email.
Spend ratio sanity. If Meta retargeting spend exceeds 25 percent of total blended spend and branded search is also active, the overlap is structural. Cap one of the two and the math usually improves.
How to fix it
Two paths. The right one depends on which side is cheaper to scale.
Cap Meta retargeting, lean into branded search. If Google branded CPC is below US$1.50 and your CTR is above 8 percent, branded search is the cheaper closing surface. Cap Meta retargeting at the lookback window where conversion attribution is still cleanly Meta (typically 7 days site visit, not 30) and let branded search catch the rest.
Sequence the two with exclusion audiences. Build a Google Customer Match audience from branded search converters in the last 14 days. Exclude that audience from Meta retargeting. The Meta retargeting then targets only customers who have not yet hit branded search. The two surfaces stop competing for the same conversion.
The first move is faster and gives you a clean A/B. The second move preserves both surfaces and is the right answer if your category has long consideration cycles where the Meta retargeting genuinely warms the audience for branded search.
What you should not do
Two reflexes that look like fixes and are not.
Kill branded search. Branded search at low CPC is the cheapest paid channel you have. Killing it because of the overlap moves the conversions to direct and organic, where they would have happened anyway, but you lose the surface in the rare case the customer is on the fence.
Increase the Meta retargeting window to capture more. Wider windows attribute more conversions to Meta in the platform dashboard. They do not create more conversions. The overlap gets worse, not better.
What oddly does about this
oddly's reallocation algorithm surfaces this overlap automatically. The benchmark cohort data shows the typical ratio of branded search to Meta retargeting spend at your category and revenue band, and the dashboard flags the overlap when your blended ratio is above the cohort median. The recommended move is a one-tap cap on the over-spending surface with a 14-day re-evaluation window. The platform proposes; you decide. The reallocation card lives on the Steer tier.