Will running both campaign types cause them to bid against each other?

If you scope listing groups with mutually exclusive custom labels, no — each campaign only targets its assigned product subset, avoiding internal competition.

How many products should be in the control group?

Most merchants use their top 10-20 SKUs by revenue or margin, enough to generate meaningful search-term data without fragmenting budget too thin.

Can I add negative keywords directly to a PMax campaign?

Not through standard self-serve tools; Google requires negative keyword lists for PMax to be submitted through Google Ads support at the account level.

Should the control group use manual or automated bidding?

Enhanced CPC or Maximize conversion value are generally more practical than pure manual CPC once you have baseline conversion data, since they reduce the need for constant manual adjustment.

Standard Shopping Campaigns as a PMax Control Group

How to run a small Standard Shopping campaign alongside Performance Max to keep search-term visibility and a manual-bid benchmark.

Why keep a Standard Shopping campaign at all

Performance Max is a strong default for most Shopify catalogs, but it hides two things merchants often want: the individual search terms triggering clicks, and a manual-bid benchmark to sanity-check whether Smart Bidding's automated decisions are reasonable. Running a small, deliberately scoped Standard Shopping campaign alongside PMax gives you both back, without sacrificing the broader reach PMax provides on the rest of the catalog. Think of it as a control group in the experimental sense — a known, transparent baseline you compare the automated system against.

Which products belong in the control group

Scope the Standard Shopping campaign to your top 10-20 SKUs by revenue or margin, not your entire catalog — the goal is visibility into how your bestsellers are performing at the query level, not full parity with PMax. Use a custom label in your Merchant Center feed (set via the Shopify Google & YouTube channel's label rules, or a feed management app) to isolate this subset, so both campaigns can filter listing groups against the same label consistently as your catalog changes.

Setting up the listing group filter

In the Standard Shopping campaign's ad group, set the listing group to 'All products' then drill in and filter by your custom label, for example custom_label_0 = 'control-group'. In your PMax campaign's asset groups, add the inverse filter excluding that same label, so the two campaigns aren't bidding against each other for the same inventory. This mirrors how Google's own Shopping/PMax priority logic behaves by default, but making the exclusion explicit avoids relying on undocumented auction-priority edge cases.

Budget sizing for the control group

Keep the control group's budget modest relative to total spend — typically 20-30% of your combined Shopping/PMax budget is enough to generate meaningful search term data on your top SKUs without starving PMax of budget for the long tail of your catalog. For a $150/day total budget, that's roughly $30-45/day on Standard Shopping and the remainder on PMax.

What to compare weekly

Pull CPA and ROAS for the identical product set from both campaigns where possible — since PMax won't show per-product performance as cleanly, use the 'Products' report available at the campaign level in PMax (found under Campaigns > select PMax campaign > Products tab) to see performance for the excluded-then-reincluded set if you ever test full overlap, or simply compare your control group's CPA against your blended PMax CPA for a directional read. If Standard Shopping's manual-bid CPA on your bestsellers beats PMax's blended CPA by a wide margin, it may mean PMax is spending disproportionately on lower-intent placements for those same products elsewhere in the catalog.

Using search terms from the control group to improve PMax

The search terms report in Standard Shopping (Campaigns > Search terms, filtered to your Shopping campaign) reveals exact queries driving clicks and conversions on your top products. Feed high-performing query themes into your PMax asset group headlines and descriptions — if 'waterproof running shoes' converts well in Shopping but your PMax asset group headlines don't mention water resistance, that's a concrete copy improvement to make. This is one of the few reliable ways to get query-level insight back into a PMax setup that otherwise only shows aggregated categories.

Negative keywords learned here protect PMax too

Any consistently poor-converting or irrelevant query you identify and add as a negative keyword in the Standard Shopping campaign should also be submitted to Google Ads support as an account-level negative keyword list, which can then apply to PMax campaigns as well. This is currently the only way to exclude specific search terms from PMax, since campaign-level negative keywords aren't self-serve there the way they are in Standard Shopping.

Worked example: catching an inefficient placement

Suppose your control group of 15 bestsellers shows a blended CPA of $18 in Standard Shopping, while your PMax campaign's overall CPA across the full catalog (including those same 15 products, before you exclude them) sits at $27. After excluding the 15 from PMax and giving Standard Shopping the dedicated budget, if PMax's CPA on the remaining catalog stays around $24-26, that's evidence PMax wasn't spending inefficiently specifically on your bestsellers — the higher blended number was mostly the long tail. If PMax's CPA on the remaining catalog also drops sharply after the split, it suggests those bestsellers were previously absorbing disproportionate budget without matching efficiency.

When to fold the control group back into PMax

If after 6-8 weeks the control group's CPA is consistently within 10-15% of PMax's performance on comparable products, the manual overhead of maintaining a separate campaign may not be worth it — fold the excluded SKUs back into PMax's listing groups and let Smart Bidding manage them alongside everything else. Keep the search-term learnings and negative keyword list you built, since those remain useful even after the campaign structure simplifies.

Bid strategy choice for the control group

Use Enhanced CPC or Maximize conversion value rather than pure manual CPC for the control group once you have baseline data, since pure manual bidding requires constant adjustment as competition shifts. Enhanced CPC keeps your manual bid as a ceiling reference point (useful for the control-group comparison) while still letting Google adjust bids up or down based on predicted conversion likelihood for each auction.

Reporting cadence for a two-campaign structure

Review the control group weekly given its smaller scope and the value of catching query-level issues quickly, and review PMax on a biweekly cadence focused on the Insights tab and overall ROAS trend rather than daily fluctuations. Avoid the trap of over-managing the control group simply because its data is more visible — visibility isn't the same as needing more frequent changes.

Documenting the split for future reference

Keep a simple record (a shared doc or spreadsheet) of which custom label values are excluded from which campaign and why, especially if multiple people manage the account. Feed structures like this tend to be forgotten within a few months, and a new team member or agency picking up the account without this context can accidentally reintroduce full overlap by editing listing group filters without realizing the split was intentional.

Limitations of this approach

This control-group method gives directional insight, not a scientifically clean experiment — PMax and Standard Shopping can still influence each other's auction dynamics indirectly, and Google's cross-channel Smart Bidding models aren't fully transparent about attribution when both campaign types run simultaneously. Treat the comparison as a useful sanity check and a source of creative and negative-keyword intelligence, not as definitive proof of which system is 'better' in an absolute sense.