You launched Performance Max because Google promised it would find customers everywhere — Search, YouTube, Gmail, Maps, Display, all from one campaign. Weeks later the spend is real, the leads are thin, and the reporting feels like a black box that tells you almost nothing.
Before you pause it in frustration, know this: a struggling Performance Max campaign is almost always a feeding problem, not a broken tool. At ZenWeb, we run managed Google Ads for over 500 Malaysian businesses, and the same short list of gaps explains most poor results — thin conversion data, missing audience signals, and weak assets.
This guide explains why Performance Max underperforms, what failure actually looks like, and the exact order to fix it. The short video below sets up the problem before we dig in.
Source video: Performance Max Not Working? Do THIS on YouTube
Quick Answer: Performance Max is a goal-based Google Ads campaign that serves across Search, Display, YouTube, Discover, Gmail, and Maps from one setup, with Google’s AI controlling targeting, bidding, and creative mix. It underdelivers when you give it too little to learn from — automation only works as well as the inputs behind it.
You hand Performance Max your goals, your budget, your creative assets, and your data. In return, per Google’s own definition, it reaches across every Google channel from a single campaign. That is the appeal — and the risk. You trade manual control for automation, so the quality of what you feed it decides everything. Our guide to managed Google Ads covers where that control still sits.
You are not powerless inside the black box. You still shape results through:
Quick Answer: Most poor Performance Max results trace to a short list of input gaps — thin or misconfigured conversion tracking, too few creative assets, and no audience signals to guide the AI. Brand traffic inflating the numbers and a budget too small to exit learning round out the usual suspects.
Across ZenWeb-managed accounts, underperformance rarely comes from one dramatic fault. It comes from two or three quiet gaps stacking up. The chart below shows how often each cause appears in the struggling PMax campaigns we audit. A weak conversion setup leads the list — if that sounds familiar, our guide on Google Ads conversion tracking that isn’t working walks through the fix.
| Root cause | Share of struggling accounts |
|---|---|
| Thin or misconfigured conversion tracking | ~66% |
| Too few or weak creative assets | ~61% |
| No audience signals given | ~58% |
| No account-level negative keywords | ~49% |
| Brand searches inflating the numbers | ~43% |
| Budget too low to exit learning | ~37% |
Source: aggregated from ZenWeb-managed Performance Max accounts, Malaysia, 2024–2026. Shares exceed 100% because most accounts show two or three causes at once.
Notice the pattern: the top causes are all things you control before Google ever runs an auction. Account health matters too — a disapproved asset or ad quietly forces weaker creative to serve, dragging the whole campaign down.
Quick Answer: A failing Performance Max campaign shows a few clear tells: spend rising while real leads stay flat, cost per acquisition climbing past week six, conversions that look great on-screen but never turn into sales, and asset groups stuck on “low” ad strength. Spotting which one you have points you at the fix.
Before you change anything, read the symptoms honestly. Each one hints at a different underlying gap:
Not sure which of these is dragging your campaign down?
A quick account review reads the signals for you in minutes. See our Google Ads management →
Quick Answer: The more complete your assets and signals, the better Performance Max performs. A bare-minimum setup — few assets, no video, one conversion action, no audience signal — leaves the AI exploring blindly. A complete setup gives it the raw material to stabilise faster and spend efficiently.
Think of setup completeness as a ladder. Each rung gives Google’s AI more to work with, and the results follow. The table below shows what we see as accounts climb it, and why the top rung is worth reaching before you judge the campaign. Getting there is a core part of how we run Google Ads for Malaysian businesses.
| Setup level | What’s in place | Typical result |
|---|---|---|
| Bare minimum | A handful of assets, no video, one conversion action, no audience signal | Slow, volatile, and expensive |
| Partial | Full text and image assets, one audience signal, basic conversions | Improving, but results hit a ceiling |
| Complete | Full assets including video, strong audience signals, clean conversions with values | Stable, efficient, ready to scale |
Source: ZenWeb-managed Google Ads accounts, Malaysia, 2024–2026. Interpretive bands, not guarantees.
Google’s optimisation guidance points the same way: add the full set of headlines, descriptions, images, and at least one video, because a richer asset group can serve in more placements. A missing video alone locks you out of much of YouTube.
Quick Answer: Performance Max needs roughly six weeks to stabilise. The first two weeks are volatile learning, weeks three to four calibrate, and results usually settle by weeks five to six. Judging or overhauling the campaign before then resets the learning and starts the clock again.
Impatience is its own failure mode. Google’s guidance is to allow about six weeks for automated bidding to settle, and our accounts follow the same curve. The timeline below shows what to expect — and, just as important, what not to do at each stage. If budget is the thing holding learning back, our guide on a campaign limited by budget explains how to lift the cap without waste.
| Phase | What’s happening | What to do |
|---|---|---|
| Week 1–2 — Learning | Wide exploration, swinging cost per lead | Leave it alone; let it gather data |
| Week 3–4 — Calibrating | Google narrows to what converts | Small changes only; no overhauls |
| Week 5–6 — Stabilising | Performance settles into a range | Read results; refine assets |
| Week 7+ — Steady | A reliable baseline to judge | Scale budget; refresh creative |
Illustrative, based on Google’s ~6-week ramp guidance and ZenWeb-managed accounts, Malaysia, 2024–2026. Actual timing varies by budget and conversion volume.
Quick Answer: Fix Performance Max by repairing its inputs in order — conversion tracking first, then audience signals, then creative assets, then account controls like negatives and brand separation. Only after those are solid do you adjust budget and let it relearn. Fixing inputs beats fiddling with bids every time.
Work the steps in this order. The early ones do the heavy lifting; the later ones sharpen what the AI has already learned.
Quick Answer: Fixing conversion tracking and adding audience signals give the biggest lift for the least effort, usually within a week or two. Stronger creative pays off over a few weeks. Account negatives and brand separation are quick wins that sharpen quality fast. Budget changes help only once the inputs are right.
Not every fix moves the needle at the same speed or cost. The table ranks the main levers by effort, impact, and how soon you feel it — the same discipline that helps when a CPC is running too high.
| Fix | Effort | Impact | Time to see it |
|---|---|---|---|
| Fix conversion tracking | Medium | High | 1–2 weeks |
| Add audience signals | Low | Medium–High | 1–2 weeks |
| Strengthen creative assets | Medium | High | 2–4 weeks |
| Add account-level negatives | Low | Medium | Days |
| Separate brand into its own campaign | Low | Medium (clarity) | Days |
| Raise budget to exit learning | Low | Medium | 1–2 weeks |
Source: ZenWeb-managed Google Ads accounts, Malaysia, 2024–2026. Ranges are typical, not guaranteed.
Want these fixes run for you, in the right order?
As a Google Partner managing 500+ Malaysian accounts, we rebuild struggling Performance Max campaigns every week. Get a free Google Ads audit →
Quick Answer: The habits that keep Performance Max stuck are pausing it before learning finishes, trusting inflated brand conversions, starving it of budget, cramming everything into one asset group, and giving it no audience signal. Each treats the campaign as broken instead of under-fed.
Avoid these and the fixes above work far faster:
Performance Max is not a broken tool — it is a demanding one. It rewards clean conversion data, strong assets, and clear audience signals, and it punishes thin inputs by spending your budget while it guesses. Almost every struggling campaign we audit is under-fed, judged too early, or both.
Fix the inputs in order, separate brand from non-brand, give it six weeks, and read the results honestly. Do that and Performance Max earns its place beside your Search campaigns. If you would rather have the whole rebuild done for you, our team handles it through managed Google Ads, and our Google Ads agency turns around underperforming campaigns for Malaysian businesses every week.
Performance Max burning budget without the leads?
Book a free 30-minute session — we’ll check your conversion setup, assets, and audience signals, then hand you a concrete plan to get Performance Max delivering at the spend you already have.
Usually because it is under-fed. The most common causes are broken or thin conversion tracking, too few creative assets, and no audience signals to guide the AI. Brand searches inflating the numbers and a budget too small to exit learning are close behind. Most struggling accounts have two or three of these at once, so fix the inputs before touching bids.
Allow about six weeks. The first two weeks are volatile learning, weeks three and four calibrate as Google narrows targeting, and results usually settle by weeks five to six. Automated bidding needs that time to gather data, so avoid major changes during the window — each big edit resets the learning and starts the clock again.
It can. Performance Max may absorb branded and high-intent queries your Search campaigns used to win, which makes PMax look strong while Search coverage quietly drops. Watch your branded Search impression share, add brand as an account-level exclusion where it fits, and keep a dedicated brand campaign so you can see what each is really contributing.
Fill every asset slot. Google recommends a complete set of headlines and descriptions, landscape, square, and portrait images, and at least one video per asset group. A fuller asset group serves in more placements, and a missing video locks you out of much of YouTube. Completeness is one of the fastest ways to lift a weak campaign.
Rarely, and not as a first move. Pausing mid-learning throws away the data the campaign needs and guarantees you never see a fair result. Fix conversion tracking, add signals and assets, separate brand traffic, then give it a full six weeks. Only if it still underperforms with solid inputs should you rethink the campaign type.
Complete the form and our team will contact you to discuss your goals. Let’s grow your business.
Online
