Skip to main content

Paid Advertising

Performance Max

Genuinely effective constrained, and genuinely expensive trusted. The default configuration is designed in Google’s interest, not yours.

Performance Max runs across every Google inventory from a single campaign, with limited reporting and few controls. The Nexclick applies the constraints that do exist — brand exclusions, asset group structure, audience signals — and adds the reporting Google withholds, because unconstrained it will buy traffic you already had.

Book a 20-minute callMonthly retainer · Paid Advertising from £750/month

Is this you?

What usually prompts the call

  • Performance Max has become most of your budget without anybody deciding that.
  • Its reported return looks excellent and total revenue has not moved.
  • You cannot see which search terms or placements it is buying.
  • It launched, search performance dropped, and nobody has compared the two.

What we do

The actual deliverables

Things that appear on an invoice, not adjectives.

Apply brand exclusions on the first day
Without them PMax absorbs branded search, reports it as its own conversion, and produces a return figure that flatters the whole account. The single most important setting, and it is off by default.
Structure asset groups deliberately
Split by product margin, category or audience rather than one group holding the entire catalogue. A single asset group is the version of Performance Max with no controls whatsoever.
Use audience signals as direction, not targeting
They influence where the campaign looks rather than restricting it, which is widely misunderstood. Feeding it your customer lists and genuine high-intent segments does measurably shift performance.
Extract the reporting Google does not surface
Scripts and custom reports exposing search terms, placements and asset-level performance. Without them you are approving spend you have no ability to inspect.
Exclude the placements that waste money
Mobile game inventory and content-farm placements, filtered continuously rather than once. The default inventory includes a great deal nobody would have chosen deliberately.
Fix the feed where products are involved
PMax leans heavily on the Merchant Center feed, so most of its shopping performance is determined before the campaign is even built. Feed work is campaign work here.
Test incrementality against what it replaced
Compared with the search and shopping campaigns it cannibalised rather than accepting its self-reported figures, which double-count anything it took from elsewhere.
Nominate the right conversion goal
It optimises hard towards whatever you name. A soft conversion nominated here produces an enormous volume of the wrong outcome very quickly, and the reporting will look excellent throughout.

Checklist

The Performance Max controls that do exist

You are told PMax has no controls. It has fewer than search, and these are the ones available. Most accounts we inherit have applied under half of them, usually including the first.

  1. 01Are brand terms excluded, so it is not re-buying traffic that would have arrived anyway?
  2. 02Is there more than one asset group, split by margin, category or audience?
  3. 03Have audience signals been supplied, including your own customer list?
  4. 04Are mobile app placements excluded at account level?
  5. 05Is there a placement exclusion list, and has anyone updated it this quarter?
  6. 06Is the campaign optimising towards a real conversion rather than a soft one?
  7. 07Do conversion values reflect actual margin rather than order value?
  8. 08Is new customer acquisition configured where repeat purchase matters?
  9. 09Is a search term report being extracted by script, since the interface will not show one?
  10. 10Are final URL expansion settings deliberate rather than left at default?
  11. 11Does the Merchant Center feed have complete titles, attributes and images?
  12. 12Are there enough assets that the platform is not simply repeating three of them?
  13. 13Has anybody compared performance against the campaigns PMax replaced?
  14. 14Has it been left alone for a full learning period, or edited every week?

How it works

Step by step, with timeframes

Timeframes are typical rather than guaranteed, and they assume we get account access and approvals when we ask.

  1. 01Week 1

    Audit or build

    An existing campaign reviewed against the control checklist, or a new one built with every constraint in place from the outset.

  2. 02Week 1–2

    Constraints and reporting

    Brand exclusions, asset group structure, audience signals, placement exclusions and the reporting scripts installed before spend scales.

  3. 03Week 2–5

    Learning period

    Left alone deliberately so the campaign can stabilise. This requires more patience than most accounts are given, and impatience here is the commonest cause of poor results.

  4. 04Ongoing, monthly

    Optimise and compare

    Asset performance reviewed, placements pruned, and an incrementality read taken against the campaigns running alongside it.

What you get

Reporting and ownership

  • Brand exclusions applied and verified, so the reported return is not inflated by your own name.
  • Asset groups structured by margin or category rather than one group holding everything.
  • Reporting on search terms and placements that the interface will not give you.
  • A placement exclusion list that is maintained rather than set once and forgotten.
  • An incrementality comparison against the campaigns Performance Max sits alongside.

Tools and platforms

  • Google Ads Performance Max
  • Google Merchant Center
  • Google Ads scripts for search term and placement reporting
  • GA4 for cross-campaign incrementality
  • Asset production across the full inventory range

Timeline

How long this actually takes

Two weeks to build or constrain, then a genuine four to six week learning period during which it should be left alone. The commonest mistake is editing it weekly so it never stabilises. Worth stating plainly: Performance Max is not a bad campaign type. Its defaults are simply set in Google’s interest rather than yours — brand traffic included, placements unrestricted, reporting minimal. Constrained properly it performs well and frequently better than the manual campaigns it replaced. Left on defaults it will find the customers you already had, charge you for them, and report an excellent return on ad spend while your total revenue sits exactly where it was.

Pricing model

Monthly retainer

Normally part of full account management rather than a standalone engagement, because PMax interacts with everything else running. Available on its own where it genuinely is the only campaign type you use.

Full pricing

Questions

Performance Max questions

Does Performance Max steal traffic from our other campaigns?

Yes, unless constrained. It competes for the same auctions as your search and shopping campaigns and, without brand exclusions, absorbs branded queries that would have converted anyway. The revenue then appears under PMax and disappears from search, which looks like PMax succeeding.

Why does Performance Max report such a good return on ad spend?

Partly because it works, and partly because it is claiming conversions that other campaigns or organic search would have produced. Branded traffic is the main culprit. Exclude brand terms and the reported figure usually falls sharply — which is the honest number rather than a decline.

Can we see which search terms Performance Max is buying?

Not in the interface, which gives categories rather than terms. A Google Ads script can extract far more detail, and we install one as standard. Running a campaign whose queries you cannot inspect is approving spend without visibility, whatever the reported return looks like.

How long is the learning period, really?

Four to six weeks with steady conversion volume, longer with less. Significant edits restart it, which is why accounts adjusted weekly never leave it. The discipline is genuinely difficult — the campaign performs worst precisely when the temptation to intervene is strongest.

Should we run Performance Max alongside standard shopping?

They compete, and Performance Max generally takes priority in the auction. Running both usually means the standard campaign gets starved. Some accounts run standard shopping for a defined product subset with PMax excluding it, which works but needs maintaining as the catalogue changes.

How many asset groups should we have?

Enough to separate things you would bid differently on — margin bands, categories, distinct audiences — and few enough that each has sufficient conversion volume to learn from. For most accounts that is three to six. One is no structure; twenty starves them all.

Is Performance Max worth running for a lead generation business?

It can be, with strong conversion tracking and offline import so it optimises towards qualified leads rather than form fills. Without those it produces high volumes of poor-quality enquiries very efficiently. For lead gen we would generally get search working properly first.

Tell us what you are trying to fix

A 20-minute call, no pitch deck. The Nexclick will tell you what we would do, roughly what it costs, and whether we are the right people for it.