Performance Max reporting has historically made it easy to see that something changed but hard to see what changed. Aggregate CPA rises. ROAS softens. The campaign insights tab offers vague explanations. Creative performance fluctuates. The channel timeline chart does not solve all of these problems, but it adds a specific diagnostic lens that the aggregate view systematically obscures: where is the budget actually going, and has that changed?
Why aggregate PMAX metrics hide channel drift
A PMAX campaign reporting £18 CPA this week versus £14 CPA four weeks ago tells you performance has deteriorated. It does not tell you whether the CPA change is driven by:
- –A shift from Search and Shopping (high-intent, high-CVR) toward YouTube and Display (lower-intent, lower-CVR)
- –Increased competition on the same channels at the same times
- –A creative performance decline within a specific channel
- –A seasonal conversion rate shift that is consistent across channels
- –A signal quality degradation affecting Smart Bidding across all channels
Each of these has a different fix. Treating a channel-mix drift problem as a bidding or creative problem will not resolve it. The timeline chart helps narrow down the diagnosis.
The five patterns marketers should look for
| Pattern | What it looks like | Likely interpretation |
|---|---|---|
| Search/Shopping share declining over 4+ weeks | Lower proportion of spend on these channels despite stable budget | PMAX exploring upper-funnel; investigate whether search impression share has also dropped |
| YouTube share spike | Sudden large increase in YouTube spend percentage | New video assets added, or algorithm increased video distribution — check if CPA followed |
| Display share creeping up quietly | Gradual increase in Display % without a trigger event | Algorithm filling spend with lower-competition inventory — may be waste if no creative strategy |
| All channels flat in proportion, but CPA rising | Channel mix unchanged, performance declining across all surfaces | Competitive pressure or seasonal signal quality issue — not a routing problem |
| Maps or Gmail sudden share increase | Unusual channel appearing in timeline for first time | Location asset or Gmail creative triggering new distribution — check conversion quality for these surfaces |
How to interpret shifts by objective
Lead generation campaigns
For lead gen, Search is almost always the highest-quality channel because it captures active intent. A shift away from Search toward Display or YouTube typically signals a conversion quality risk even if short-term conversion volume holds — because the users arriving from upper-funnel surfaces are at an earlier stage of the buying journey and less likely to submit a qualified form fill.
Watch for: Search share below 50 percent in a lead gen account that was previously Search-dominant. This is a flag, not an automatic problem, but it requires investigation into downstream lead quality alongside the channel data.
eCommerce campaigns
For eCommerce, Shopping is typically the highest-intent and highest-ROAS channel. YouTube and Display can contribute meaningfully for branded, visual, or impulse categories. A shift toward YouTube that does not correlate with a new video creative strategy is more likely to be waste than productive upper-funnel investment.
Watch for: Shopping share declining while Display or YouTube share rises in a product-focused campaign. This is often a signal that the feed is losing quality or that product creative is underperforming relative to video or image assets.
Lead gen example
A home renovation company runs PMAX for lead generation. In week one through four, Search accounts for 65 percent of spend and CPL is £32. In week five, a new video asset is added as a standard upload. By week eight, YouTube share has grown to 38 percent of spend. CPL has risen to £51.
The channel timeline chart makes this shift visible immediately. The account manager pauses the video asset, reviews asset performance for the video in the asset report, and finds the video has a very low CVR relative to text and image assets. Removing the video reduces YouTube distribution and Search share recovers over three weeks. CPL returns to £34. Without the channel timeline view, the video asset would have been evaluated in isolation without connecting it to the channel shift.
eCommerce example
An apparel retailer increases PMAX budget from £4,000 to £7,000 per month. The channel timeline shows Shopping share drops from 58 percent to 41 percent as Display absorbs the incremental budget. ROAS falls from 5.8x to 4.1x over four weeks. The retailer expected the increased budget to scale Shopping volume proportionally. The channel data shows the algorithm deployed the additional budget into lower-competition Display inventory rather than scaling Shopping spend.
Response: the team creates a separate feed-only PMAX campaign to concentrate Shopping-eligible budget separately, while keeping a smaller general PMAX campaign for broader distribution. Shopping ROAS recovers to 5.4x in the dedicated campaign while the general campaign handles upper-funnel distribution at an acceptable blended CPA.
What to compare alongside the timeline chart
- –Asset performance report: which asset types (video, image, text) have the strongest interaction rates — connects creative to channel distribution
- –Search impression share trend: if Search impression share is falling while Search PMAX spend is holding, competition may be the driver not channel mix
- –Conversion rate by landing page: if channel mix is stable but CVR is falling, the page is the issue not the channel
- –Audience signal performance: if audience signals are too broad or stale, the algorithm may expand to cheaper inventory as it struggles to find converting users efficiently
Bad drift vs acceptable exploration matrix
| Channel shift | CPA / ROAS trend | Verdict |
|---|---|---|
| Search ↓, Display ↑ (gradual) | CPA rising | Investigate and likely intervene — classic waste drift |
| YouTube ↑ (after new video creative) | CPA stable or improving | Acceptable — algorithm testing new creative, watch for 2 more weeks |
| Shopping ↓, Display ↑ (after budget increase) | ROAS falling | Intervene — scale Shopping separately, don't let Display absorb incremental budget |
| All channels proportionally stable | CPA rising | Not a channel issue — check competition, seasonality, creative freshness |
| Maps/Gmail appearing for first time | Conversions from these surfaces low | Audit placement quality — often low-intent for most service or eCommerce goals |
Weekly monitoring checklist
- Open channel timeline chart — note current channel share percentages for Search/Shopping vs upper-funnel channels
- Compare to the prior two weeks — has any channel shifted by more than 10 percentage points?
- If yes: check whether a creative asset was added, removed, or changed in the same period
- Check whether the CPA or ROAS trend correlates with the channel shift timing
- If the shift looks problematic: remove or pause the asset that correlates with the distribution change, or add a feed-only campaign to isolate Shopping intent