We have a slightly uncomfortable theory after auditing more than 80 Google Ads accounts in Spain in the last 18 months: about 70% of advertisers in this country are quietly funding Google's quarterly results. Not because their products are bad. Not because the platform is broken. But because of seven repeating, fixable mistakes — and they all live in the boring parts of the account that nobody likes to open.
This is the exact 7-step audit we run on every new client account. We've seen accounts cut wasted spend by 40% in 30 days using only what's below.
Step 1. The Search Terms graveyard
Open Reports → Search terms and sort by Cost. We expect to find at least one of three horrors:
- Brand searches of competitors sucking 12–18% of budget — and not converting.
- Pure informational queries ("qué es CRM", "cómo funciona…") with cost-per-conversion 4x your average.
- Generic 1-word queries ("ofertas", "promo", "barato") that match too broad and tank quality score.
A real client (B2B SaaS, Madrid): €2,840/month was going to "que es un crm gratis" and similar. They sell a paid CRM. We added 47 negatives in one afternoon. Following month: same conversions, −31% spend.
Audit action: dump the last 90 days of search terms, build a negatives list of at least 30 terms, push it to all campaigns. Repeat monthly.
Step 2. Conversion tracking — is it actually right?
This one hurts. We routinely find:
- Page-load conversions ("thank you page visited") counted as Purchases. Bots reach those pages too.
- Phone-click conversions counted, even though nobody actually picks up the phone.
- Multiple conversion events firing on the same form, double-counting leads.
- No "primary" vs "secondary" conversion distinction, so the algorithm is optimizing for the wrong thing.
A logistics company in Valencia was reporting €42 CPL. After we cleaned up the tracking and switched to verified form submissions only, it became €118. That's the real CPL — and it's the one Google's algorithm needs to optimize against.
The painful truth: bad conversion data is worse than no conversion data, because Smart Bidding will happily burn budget chasing fake events.
Audit action: verify every conversion in GTM debug mode, mark only one as Primary, send actual value (not 1,1,1) where possible.
We love Pmax. We also see it eat budgets alive when:
- It's not given a meaningful audience signal
- It's launched without brand-term exclusions (so it gets credit for brand traffic that would have closed anyway)
- The asset groups have just 1 image and 1 headline (Google needs at least 5 of each for diversification)
A Spanish ecommerce client — €11k/month in Pmax. We added brand-term exclusion and increased asset diversity from 1 image to 8. Same 30 days later: same revenue, −22% Pmax spend. The "performance" was just absorbing organic brand demand.
Audit action: add brand_terms as an account-level negative keyword list. If Pmax ROAS drops, you've just found out how much of it was inflated.
Step 4. The Quality Score blind spot
QS isn't an SEO vanity metric. It's a direct discount on your CPC. A QS of 7+ vs 4 can halve your cost.
Three things drive it:
- Expected CTR — fix by tightening ad groups (1 theme = 1 ad group)
- Ad relevance — fix by including the keyword in the H1 and the description
- Landing page experience — Core Web Vitals on the landing page directly impact this
We routinely see Spanish clients sending paid traffic to their homepage for every ad group. That's a QS-3 strategy.
A real estate agency in Barcelona had ad groups for "piso obra nueva Sant Cugat", "piso obra nueva Gracia", "piso obra nueva Pedralbes" — all pointing to /. We built three matching landing pages. CPL went from €87 to €34 in 6 weeks.
Audit action: for each top-spending ad group, ensure there is a dedicated landing page with the matching H1 and three relevant on-page sections.
Step 5. The day/hour heatmap nobody looks at
Open Reports → Time → Day of week, Hour of day with conversions and cost.
We have never run this report on a Spanish account and not found at least 15% of waste. Common patterns:
- B2B accounts spending heavily Saturday and Sunday nights
- Restaurants and clinics still bidding at 03:00–06:00 with zero conversions
- HR/recruitment accounts spending all of August on autopilot
A B2B client in Madrid: 23% of their monthly spend was happening between 22:00 Friday and Sunday morning. Conversion rate during that window: 0.4%. We added an ad schedule cutting bids to −80% during that window. CPL down 19% the next month.
Audit action: apply ad scheduling whenever a 4-hour window has clearly worse conversion rate than the daily average, sustained over 60+ days.
Step 6. Auction-time signals (the underrated lever)
Most accounts only optimize on keyword + match type + bid. Auction-time bidding strategies use signals 100x richer:
- Device + browser + OS
- Demographics (age, household income tier — yes, available in Spain via "Detailed Demographics")
- Audience (in-market, custom-segments based on competitor URLs visited)
- Geo radius down to postcode
- Weather (for verticals like ecommerce, dining, travel)
For a ski rental shop in the Pyrenees we layered in weather + temperature signals with smart bidding. Conversion rate on snow-forecast days went up +38% and they re-allocated budget away from no-snow days. ROAS doubled in one season.
Audit action: if your account is purely keyword-driven, you're at maybe 40% of the platform. Move to value-based smart bidding with at least 2 audience signals per campaign.
Step 7. The landing page (where 80% of the leak happens)
This is the one nobody on the ads team can fix without a developer, which is exactly why nobody fixes it. Symptoms:
- LCP > 3.0s on mobile. Spain has great 4G/5G; if your LCP is bad, it's purely your code.
- Form has more than 5 fields. In Spain, anything above 4 fields drops conversion ~20% per extra field.
- No social proof above the fold. Spanish buyers are heavy on social validation — 12+ reviews next to the form is non-trivial.
- Sticky button on scroll missing on mobile. Every B2C below €100 should have one.
- No alternative CTA for non-converters (WhatsApp, phone, "send me a quote").
A clinic in Madrid with a perfectly-optimized Google Ads account was still doing €74 CPL. We rebuilt only the landing page (Next.js, 1.4s LCP, 3 form fields, sticky WhatsApp button, 8 patient reviews above the fold). Same ads, same budget. CPL: €31 in week 2.
This is why we always tell clients: a great ads team without a great landing page is a Ferrari with the parking brake on.
The full audit checklist (steal this)
If you're going to audit your own account next Monday morning, here's the order:
- Search terms → build 30+ negatives.
- Conversion tracking → one Primary, verified, with value.
- Pmax brand exclusion + asset diversity (5+/5+).
- Quality Score → fix any keyword below 5.
- Day/hour heatmap → schedule.
- Audience signals → at least 2 per campaign.
- Landing page → LCP < 2.5s, 4 fields max, social proof above fold.
Run the full thing in one session (it takes 4–6 hours). Re-run quarterly.
Want us to do this for free on your account?
If you're spending €1,500+/month on Google Ads in Spain and want a written audit with concrete actions and projected savings, we'll do it for free. No commitment. We just want to talk to you afterward, and frankly, we're confident enough in the value that we don't worry about giving it away.
Request a free Google Ads audit or explore our paid-media services.
The 70% who never fix the boring stuff will keep funding the platform. We'd rather you funded your growth instead.