Google Ads Audit: where your budget goes to waste
We examine what your account delivers, and act where the auction is decided. The result: prioritised actions, your wasted spend quantified in euros, a plan your team can act on immediately.
Quick check at no cost · Three evidenced findings, in writing · Read-only access is enough
Lukas Tippmann and Jordi Moll review your account themselves. Both were Account Strategists at Google in Dublin.
Afterwards, if you want: the full audit at a fixed price from €3,500 →
What determines where your ads show and what you pay
Google does not sell to the highest bidder. Which factors count besides the bid depends on the campaign type, and only in Search does Google report them as three separate ratings.
In Search campaigns, Google calculates an Ad Rank for every search. Google names six factors for it and publishes no formula. Only one of them is your bid.
- In your hands
- Bid
- Quality of your ads and landing page
- Impact of your assets
- Set by the auction
- Ad Rank thresholds
- Competitiveness of the auction
- Context of the search
Google assesses quality in three parts. That puts five things in your hands:
All three quality components are relative, not absolute. Google rates them above average, average or below average, compared with the other advertisers serving on the same keyword. So there is no fixed target you reach once: what counts as good enough is set by your competitive environment, and that keeps moving. This is why we assess your account against your actual auction competition, not against a benchmark table.
What it means
How to spot a problem
Where we step in
What the report does not show: how many clicks on brand searches would have come through your organic result without an ad. An incrementality test answers this; since late 2025 Google offers it for smaller budgets too, though only through a Google account representative. Which searches beyond your keywords cost money now that broad match and AI Max have a say is shown in the search terms report.
Google publishes neither a formula nor a weighting for these factors, which is why we deliberately give no percentages. The figures in circulation come from third-party reverse-engineering of the visible 1–10 Quality Score. That score is a historical average: in the auction, quality is recalculated for every single query, including context such as device, location and time of day. The six factors are listed in Google’s help page on Ad Rank.
Shopping ads have no keywords. Google matches a search to your products through the product data in Merchant Center: title, description, category and identifiers such as the GTIN. Besides the bid, what counts is how well a product fits the search. Google does not show separate quality ratings here the way it does in Search.
The levers
- Product data: titles, GTIN and images decide which searches a product can appear for. If price or availability differ from the shop, Google disapproves the product.
- Bids by product value: an item with a high margin can carry a different cost per click than one with a low margin.
- Campaign setup: negative keywords and campaign priority set for which searches, and from which campaign, a product is bid on. If the same products also run in Performance Max, Ad Rank has decided which campaign serves since October 2024.
How to spot a problem
- Products are disapproved in Merchant Center or barely serve.
- Many clicks come from very generic search terms, and few of them buy.
- Your prices sit above the benchmark price Merchant Center shows for the same product.
- Contribution margin does not keep pace with revenue.
Where we step in
- Shopping feed & Merchant Center: which products do not serve at all, and why
- Search terms & negative lists
- Bidding strategy & budget pacing
Performance Max is one campaign for all Google channels: Search, Shopping, YouTube, Display, Discover, Gmail and Maps. Google picks the channel, the bid and the ad combination itself, guided by the goals you set. If a search matches an exact match keyword in a Search campaign, the Search campaign takes priority.
The levers
- Conversion goals and values: Performance Max learns from what you report as success, optionally with higher bids for new customers.
- Product data and asset groups: the material Google builds the ads from.
- Audience signals: they speed up learning but do not limit who sees the ads.
How to spot a problem
- A large share of results comes from searches for your brand name.
- Costs rise faster than the number of new customers.
- Nobody can say how much each channel contributes, even though Google now provides a channel report and a search terms report for it.
Where we step in
- Performance Max & cannibalisation: overlap with Search and Shopping, split by channel, search terms
- Conversion tracking & data quality: what the campaign actually optimises for
- If most of your budget sits here, we tell you upfront which part of the audit applies.
Demand Gen shows image and video ads on YouTube, including Shorts, in Discover, Gmail and the Google Display Network. There is no search query: the ad reaches people before they actively look for what you offer. Audiences and the bid goal steer whom it reaches; Google does not report a quality rating.
The levers
- Audiences: customer lists, lookalike segments and custom segments built from search terms or websites.
- Image and video material: it competes with videos, posts and emails for attention.
- Bid goal: clicks, conversions or conversion value.
How to spot a problem
- The “Conversions” column also contains engaged-view conversions: conversions Google credits to a watched video (for example after 10 seconds of viewing) rather than a click.
- Reach grows without more people searching for your brand. Google reports these searches as “Attributed Branded Searches”, but only once Google has activated the metric.
- The same creatives have been running for months.
Where we step in
- Audiences & signals
- Conversion tracking & data quality: how much rests on views rather than clicks
- Ad assets & Ad Strength
Video campaigns on YouTube aim for reach, views and engagement. Depending on the format, you pay per thousand impressions or per view. A view counts once someone has watched 30 seconds (the whole video if it is shorter) or interacts with the ad. Google does not publish a formula for how video ads are weighed against each other.
The levers
- Format and bid type: skippable ads, bumper ads, Shorts; price per view or per thousand impressions.
- Audience and environment: who sees the ad, and next to which content.
- The first seconds: skippable ads can be skipped after five seconds. By then it is decided whether anyone stays.
How to spot a problem
- Views are reported, but nobody has checked whether the videos work. Google measures this with Brand Lift (a survey on ad recall and awareness) and Search Lift (extra searches for your brand compared with a control group). Neither is open to every account; both run through a Google account representative.
- Reported conversions rest mostly on views.
- The ads run next to content that does not fit your brand.
Where we step in
- Audiences & signals: audiences and placements
- Conversion tracking & data quality: what views demonstrably achieve
What we examine
Each category comes with a specific finding and evidence taken straight from your account.
Account structure & campaigns
Cleanly separated by goal, brand and match type, or subsidising one another?
Conversion tracking & data quality
Are micro signals and real conversions mixed together? Are values, fallbacks and consent correct?
Search terms & negative lists
Which queries burn budget without effect, and are the right exclusions missing?
Bidding strategy & budget pacing
Does the strategy match your target CPA or ROAS? Where is budget over- or under-allocated?
Performance Max & cannibalisation
Is PMax absorbing traffic you could have won more cheaply? How is the channel split?
Shopping feed & Merchant Center
Feed quality, item IDs and custom labels by margin: the basis for profitable Shopping. First we count which products are not served at all because of disapprovals.
Lead quality & CRM feedback
Do MQL and SQL signals flow back into bidding? Is Google optimising for leads or for pipeline?
Audiences & signals
Real prospecting or just automated signals? How clean is the audience data?
Ad assets & Ad Strength
Are there enough assets and variants for the systems to optimise meaningfully?
Landing pages & relevance
Which pages tie up budget without converting? Does the promise match the page?
Form flow & lead capture
Where do prospects drop off? Are fields, multi-step and qualification built correctly?
Competition & auction dynamics
Who is taking your impressions, and where is displacing them realistic?
Change history
Who changed what, and when: which edit explains the kink in the curve?
Value model & its source
Does the conversion value in the account still match the calculation it came from?
The insights you can expect
The questions that usually go unanswered in day-to-day operations.
How much budget is wasted?
Wasted spend as a hard euro figure and a percentage, not a gut feeling.
Is the conversion data sound?
Or is the account long since optimising on distorted signals and wrong values?
Where do channels cannibalise each other?
Are you paying for traffic you would have won anyway?
What target ROAS makes sense?
Derived from margin and order value.
What may a Sales Qualified Lead cost?
What does a qualified lead cost you today?
Which campaigns deliver volume only?
Leads that never become opportunities, yet still tie up budget.
Which landing pages cost the most?
Without converting, ranked by the budget they absorb.
Who are you losing the auction to?
And in which segments is displacing them realistic?
What is possible with more budget?
A scenario, including the point where efficiency starts to break down.
From feed to contribution margin
What decides profitability in a shop, beyond the campaign interface.
The feed is the foundation
Titles, attributes and item IDs decide whether Shopping serves at all.
Custom labels by margin
Not every product deserves the same bid: steer by contribution margin.
Brand searches
How much are you paying for clicks from people who were already searching for you?
Weigh CLV against new customers
Repeat buyers are worth more, growth needs new customers. The balance depends on your business model.
From click to pipeline
What happens between the ad and the deal, and where pipeline is lost.
Form flow
Every required field costs enquiries: build multi-step and qualification properly.
Lead scoring
Not every lead is worth the same: score before handing over to sales.
Enhanced Conversions for Leads (ECL)
Real SQL signals back into bidding: Google optimises for pipeline instead of form fills.
Closed-loop reporting
Measurable end to end from click to close, straight out of the CRM.
Typical findings
What we see again and again, mostly where the account knows the revenue but not the cost side. A range, not a promise.
Real case · anonymised
What we found inside one shop account
One account from our practice, anonymised. All eleven findings come straight from the account, over a 90-day window.
- Channel overlap96.0%
PMax bought traffic from Search and Shopping: clicks the account would have had anyway. YouTube, Discover, Gmail and Maps went practically unused.
- Waste37.5% / 69.0%
Share of Search and PMax spend that ran on search terms without a single conversion.
- Steering metricROAS distorted
Micro signals and real conversions sat in one conversion bucket, a newsletter sign-up counted like a purchase. Bidding followed the wrong number.
- Feed control55.3%
Share of PMax delivery without product data: Shopping without the control it is built for.
- Margin steering0 labels
No custom labels by contribution margin: products at 8 and 48 percent margin ran on one bid.
- Brand protectionunprotected
Brand traffic was bought expensively through PMax.
- Customer economicsno split
New and returning customers ran in the same campaigns: repeat buyers bought at new-customer prices.
- Remarketingnot separated
Cart abandoners and product viewers were never addressed separately: the cheapest audience in the account went unused.
- Landing pages43.5%
Share of spend against pages rated below average. That hits twice, on Quality Score and on conversion rate.
- Landing page match3 of 9
Campaigns led to category instead of product pages: intent and destination did not match.
- Purchase intent1 stage
Problem, solution and purchase stages ran in one campaign at one bid.
No promise of results: these are findings, not outcomes. What comes of them depends on execution and market.
Typical pattern · not a single account
What we find in advanced lead-gen setups
The problem is rarely missing tracking, most accounts have that. It is the layers above: differentiated lead values, modelling, attribution windows and value rules. There is no single case here by design: sales cycles, deal values and CRM setups differ too widely.
- Goal definitiontwo target metrics
Campaigns optimise for MQL, sales measures SQL. Two target metrics that never meet.
- Value steering1 instead of 6
One lead value for six lead types: bidding cannot tell a €5k from a €500k enquiry.
- Conversion weighting1 weight
Whitepaper download and demo request count the same: hard and soft conversions not separated.
- Timeline30 vs. 74 days
Attribution window shorter than the sales cycle: the model never sees the close.
- Bid balancemicro dominates
Small signals outweigh the offline import. Bidding follows volume, not closings.
- Value rules0 of 3 levels
Region, device and new versus returning customer do not enter the bid.
- Form flowfriction, no return
Required and optional fields not separated: built for neither minimal friction nor pre-qualification.
- Spam protectionno honeypot
No field validation, no honeypot: bot submissions count as conversions.
- Buying committee1 message, 6 roles
Practitioners, procurement and leadership see the same ad.
- Channel qualityDisplay unfiltered
PMax delivers cheap form fills via Display that go nowhere in sales.
- Deal support0 stages
Live deals without their own messaging: no audiences from CRM stage, no support through to signature.
- After the closeneither out nor in
Existing customers neither excluded nor addressed for expansion: the cheapest revenue in the account is left untouched.
Illustrative values drawn from several accounts, not figures from a single case. The audit shows which of them apply to your account.
Six questions for any audit, ours included
Audits rarely fail for lack of findings. Mostly they fail on findings that do not hold up. These questions show the difference, whoever wrote the report.
No accusation towards others. These are the places where we ourselves have been wrong before. That is why no finding of ours goes into the report unless it passes these six questions.
The growth plan
The findings become a sequence, including the quality factors that bring your cost per click down.
Foundation
Targets from the economics
Target ROAS or CPA derived from margin, order value and close rate.
Foundation
Account structure, in stages
What demonstrably works keeps running for now. Rebuilds happen step by step, handing budget over rather than switching things off.
Foundation
Signals & bids
Conversion hygiene, correct value passing, a staged path to tROAS/tCPA.
Foundation
CRM feedback into bidding
Enhanced Conversions for Leads and offline import deliver real SQL signals.
Quality
Expected CTR & ad relevance
Ads that match search intent: the cheapest lever on cost per click.
Quality
Landing Page Experience
Deliver on the promise, load time and structure, which feeds straight into Quality Score.
Channels
Channel mix & brand protection
Every campaign with a clear role, so brand searches are not bought at a premium.
Growth
Testing roadmap
A/B tests on bids, creatives, audiences and pages, in a clear order.
Growth
Budget scaling
Growth in stages with milestones and a scenario for each spend level.
Growth
Reporting that counts
Steering on contribution margin or pipeline value.
No overhaul in one go. Every structural change resets learning phases, so we never change more than necessary and never everything at once. What measurably works stays in place. Rebuilds run in stages, with parallel delivery and a gradual handover of budget, and every stage has a checkpoint. If a step falls short of expectations, we step back rather than push on.
What you receive
Audit report
Findings per category, each with evidence from the account.
Wasted spend in euros
Quantified, not “there is potential here”.
Prioritised actions
Ranked by impact × effort, in a clear sequence.
Growth plan
Structure, signals, channel mix and budget path including a scenario.
Flat-fee proposal
Transparent, with no percentage markup on your ad spend.
Walkthrough call
We go through everything together, pushback expressly welcome.
The audit report is complete: every finding with evidence, your wasted spend in euros, and every action described so concretely that your team can implement it without us. Working with us afterwards adds no withheld knowledge: it adds the execution: we build the structures, manage the learning phases and keep steering, on request as part of our ongoing Google Ads management.
Free quick check, full audit at a flat fee
Two stages, so you never buy blind.
No cost · real substance
The quick check
We go through your account beforehand with read-only access and check the places where budget is most often lost. In the call you get three concrete findings from your account, each with the exact place we found it and a first estimate of the scale behind it.
You get the findings in writing, on a single page, with the place we found them and their scale, so you can pass them on internally. They are yours to keep, even if you commission nothing afterwards. The difference to an intro call: we don’t talk about your account, we have looked inside it.
from €3,500 · fixed price, named upfront
The full audit
Every finding backed by evidence straight from the account, every measure prioritised and described so concretely that your team can implement it without us. For Search campaigns it also states what share of your budget runs on keywords with below-average expected CTR, ad relevance or landing page experience. The fixed price starts at €3,500 net, depends on account size and is set before we start. If you start ongoing work within three months, we credit it in full against the first month. Results typically within a week of access being granted.
What the one page looks like
Three findings, biggest lever first. Each one says where in the account it sits and roughly what it costs. We discuss the fix in the call. Plus the fixed price for the full audit.
Sample with made-up example values. Your quick check shows findings from your own account.
Quick check: three findings from your Google Ads account
Form submitted, per 100 startedMobile41%Desktop67%
Search terms without a single conversion
- Where
- Campaign “Generic” › Search terms
- Size
- approx. €1,800 a month
Newsletter sign-ups count as conversions
- Where
- Goals › Conversions
- Size
- 31% of all reported conversions
The daily budget runs out by noon
- Where
- Campaign “Category” › Search lost impression share (budget)
- Size
- 22% of possible impressions
Fixed price for this account’s full audit: €3,800 netLukas Tippmann · Jordi Moll
What we need from you
Very little, and nothing that puts you under obligation.
Read-only access
Read-only access to Google Ads, optionally GA4. We change nothing without your approval. Before any access we sign a confidentiality and data processing agreement (Art. 28 GDPR) with you, including for the free quick check.
30 minutes of your time
A short conversation about goals, margins and where you stand.
Basic economics
Margin, or the value of a qualified lead, without it any ROAS statement is worthless.
What prospects want to know before the quick check
There is none. The quick check costs us preparation time inside your account: worth it to us, because we see for ourselves whether there is enough leverage for a full audit. If there isn’t, we say so openly, and you keep the three findings either way, in writing. We also only take on audits where the leverage we find clearly exceeds our fee.
One page with the three findings: what we found, where exactly in the account, and the scale behind it, plus our assessment of whether the full audit is worth it. Deliberately short enough to forward internally without having to explain anything.
Twelve areas: account structure, conversion measurement, search terms, bidding strategy, Performance Max, shopping feed, lead quality, audiences, ads, landing pages, forms and competition, each backed by evidence straight from your account.
A fixed price from €3,500 net that depends on account size and is set before we start. It does not change afterwards. We name the exact amount in the free quick check. If you start ongoing work within three months of the audit, we credit the audit fee in full against the first month.
No, the fixed price quoted is a net price. Tulos is an EU company based in the Republic of Cyprus, so we invoice businesses holding a valid VAT ID without VAT: under the reverse-charge procedure the tax liability shifts to you. You declare the VAT and reclaim it as input tax in the same step, so nothing is added on top for businesses entitled to input tax deduction. In Switzerland it comes out the same way, via Swiss acquisition tax.
The quick check: 30 minutes. The full audit: typically within a week of access being granted.
Read-only access to Google Ads, optionally GA4. Two minutes of effort, revocable with two clicks. Before any access we sign a confidentiality and data processing agreement, including for the free quick check.
It pays off for running campaigns with a few months of data history, ahead of budget decisions, or around an agency switch. It doesn’t pay off yet for brand-new accounts without reliable data or mid-rebuild. We tell you that openly in the quick check too.
No. Read-only access can be revoked with two clicks and there are no follow-up obligations. Whether you commission the full audit after the quick check is entirely your decision.
Ready for a look inside the account?
Get your Google Ads account checked: the quick check is diagnosis, not sales: at the end you hold three evidenced findings from your own account. Reply within 24h on business days.
Prefer writing? Send an enquiry
Briefly describe your account and goals. Reply within 24h on business days.
Form not showing? Email us directly at hello@tulosgrowthlab.com.
Book the quick check
Lukas Tippmann and Jordi Moll: the two of us look into your account, nobody else.
- Pick a slot and grant read accessRead-only on Google Ads, with the Art. 28 GDPR data processing agreement signed first.
- We look into the account beforehandThree evidenced findings are on the table when we talk, not slides.
- The findings are yoursIn writing on one page, plus the flat fee for the full audit, no surprises.
Calendar not showing? Book directly at meetings‑eu1.hubspot.com or email hello@tulosgrowthlab.com.