Why your Performance Max campaigns are silently losing money.

Performance Max is dangerous out of the box for storage. Here are the five settings I find dialed wrong on roughly 80% of accounts I audit — each quietly burning 5–15% of monthly spend without anyone noticing.

Most storage operators running PMax don't realize that "let Google figure it out" is the most expensive setting in Google Ads. PMax was designed to maximize whatever signal you feed it — including the wrong signals.

I run PMax audits constantly, and the same five mistakes show up over and over. None are hidden. None require advanced expertise to fix. They just require knowing they exist.

Mistake 1: Conversion goal is "Lead form submitted"

Default PMax setup pulls "form submission" as the conversion goal. Sounds reasonable. It's catastrophic for storage.

Here's why: Google's algorithm optimizes for whatever you tell it. If you tell it "leads = form submissions," it'll happily deliver you 800 form submissions a month — most of which never sign a lease, many of which were spam or junk traffic, and the cost-per-real-lease quietly doubles.

The fix: Push offline conversions back to Google. When a lease is actually signed, that data flows back to PMax via Enhanced Conversions or offline conversion import. Google then optimizes for actual leases, not form fills. CPA drops 30–50% within 60 days every time we ship this fix.

Mistake 2: No audience signals attached

PMax can run with zero audience signals, but it shouldn't. Without signals, the algorithm starts cold and burns money learning who your customer is. With proper signals, it has a head start and converges 3–4× faster.

For storage, the signals that actually move the needle:

• Customer match list — your existing renter list, uploaded as first-party data
• In-market segments — "Moving services," "Real estate," "Home & garden"
• Custom segments — keywords like "moving boxes," "U-Haul," "apartment hunting"
• Detailed demographics — homeowners, parents with young kids, recently relocated

Signals are guidance, not constraints. PMax can still serve outside them, but it learns faster.

Mistake 3: Asset groups not segmented by unit type

One asset group with all your unit types lumped together is the default — and the wrong setup. Storage is more like e-commerce than service businesses: you have inventory (unit sizes), prices (per size), and conversion paths (online reservation per size).

Run separate asset groups by unit category:

• Small (5×5, 5×10) — entry tier, high volume, lower-value renters
• Medium (10×10, 10×15) — your bread and butter, highest LTV
• Large (10×20, 10×30) — lower volume, vehicles, business storage
• Climate-controlled premium — separate creative, separate landing pages

Different creative, different audiences, different bid strategies per group. Same total budget, materially better performance.

Mistake 4: Brand keywords not excluded

By default, PMax will gladly bid on your own brand name and report those conversions as PMax wins. They aren't. Those people were going to find you anyway — you just paid Google to be found.

For multi-facility operators, this leak is even worse. PMax will eat your branded search traffic across all locations and inflate its reported ROAS while cannibalizing organic.

The fix: Account-level brand exclusion list (it's a real feature, just buried). Add your facility names, parent brand, common misspellings. Forces PMax to find new demand instead of harvesting existing demand.

Mistake 5: No spend caps on individual asset groups

PMax loves to find the cheapest conversions and pile spend there. Sounds good — except the "cheapest conversions" are often Display network garbage that converts on form fills but never converts on leases. (See mistake #1.)

Without spend caps, you'll wake up to find 70% of your spend went to a single asset group serving Display ads to mobile gaming apps in markets you don't service.

If you're running PMax without monitoring placement reports weekly, you're flying blind. The algorithm is not your friend in week 1.

Pull placement reports every week for the first 90 days. Exclude obvious garbage placements aggressively — gaming apps, low-quality content sites, irrelevant geos. The cleaner the placement list, the higher the ROAS.

What to do this week

If you've been running PMax for more than 60 days and never touched these settings, your account is almost certainly leaking 20–40% of spend. The fixes take 2–4 hours of work and the results compound for as long as you run the campaigns.

The order to ship them:

• This week: Set up offline conversion tracking for actual leases signed
• Week 2: Add audience signals, segment asset groups by unit type
• Week 3: Brand keyword exclusion, placement audit + cleanup
• Ongoing: Weekly placement audits, monthly creative refresh