Cet article fait partie du guide Optimisation Google Ads : 10 Leviers Pour Ameliorer Vos Resultats
Google Ads offers hundreds of metrics. Impressions, clicks, CTR, CPC, CPM, CPA, ROAS, conversion rate, Quality Score, impression share, average position, lost impression share... Most advertisers drown in numbers without knowing which ones actually matter.
The truth: 80% of Google Ads decisions are made on 5-6 KPIs. The rest is noise.
The KPI Hierarchy: From Impression to Customer
Google Ads metrics form a funnel. Each stage filters the previous one:
Impressions (your ad was seen) Clicks (someone clicked) Conversions (the visitor took action: form submission, call) Qualified leads (the conversion is a real prospect) Customers (the prospect signed)
The further up the funnel you look, the more impressive the numbers — and the less useful they are. 100,000 impressions are worth nothing if they generate zero customers.
Vanity Metrics: What to Ignore (or Almost)
Impressions
Impressions measure how many times your ad was displayed. It's a visibility indicator, not a performance indicator.
When to look at them: to diagnose a volume problem (if impressions drop, your budget or keywords are the issue).
When to ignore them: when judging profitability. A million impressions that generate zero leads have no value.
Average CPC
CPC (cost per click) is useful for budgeting, not for performance. A $10 CPC that generates customers at $200 is more profitable than a $1 CPC that generates nothing.
When to look at it: to check that your bids are competitive and that your Quality Score isn't catastrophic.
When to ignore it: when deciding if a campaign works. CPC alone says nothing about profitability.
Impression Share
The percentage of impressions you received out of the total available. An 80% share means you missed 20% of opportunities.
When to look at it: to understand growth potential and identify losses due to budget or ranking.
When to ignore it: as a goal in itself. Having 100% impression share on unprofitable keywords is a bad idea.
The 6 KPIs That Actually Matter
1. Cost Per Conversion (CPA)
Formula: Total Spend / Number of Conversions
CPA answers the question: "How much does each lead or sale cost me?" It's the baseline metric for evaluating profitability.
Benchmark: varies enormously by industry. In B2B services, $50-200 is common. In e-commerce, $5-30. What matters isn't the absolute number but the ratio to customer value.
Trap: a low CPA isn't necessarily good. If you're optimizing on micro-conversions (page view, scroll), your CPA will be low but your real leads will be expensive. Make sure you're measuring the right conversions with reliable conversion tracking.
2. ROAS (Return On Ad Spend)
Formula: Revenue Generated / Ad Spend
ROAS is the king KPI in e-commerce and for any advertiser who can assign monetary value to conversions. It answers: "For every dollar invested, how much do we get back?"
When to favor ROAS over CPA: when your conversions have different values (variable cart sizes, different contract sizes). CPA treats all conversions as equal — ROAS weights them by value.
3. Conversion Rate
Formula: Conversions / Clicks x 100
Conversion rate measures your landing page effectiveness. It answers: "What percentage of paid visitors become leads?"
Search benchmark: 3-5% is average, 5-8% is good, 10%+ is excellent.
If your conversion rate is below 2% in Search, the problem is likely your landing page, not your ads. Above 8%, focus on volume (more clicks).
4. CTR (Click-Through Rate)
Formula: Clicks / Impressions x 100
CTR measures your ad's attractiveness. It directly impacts Quality Score and therefore CPC.
Search benchmark: 3-5% average, 5-8% good, 8%+ excellent.
CTR is an intermediate KPI — it doesn't measure profitability directly, but it conditions everything downstream.
5. Cost Per Qualified Lead (CPQL)
Formula: Total Spend / Number of Qualified Leads (SQLs)
This is the KPI that 90% of advertisers don't track — and it's the most important one in B2B/services.
The difference between a lead and a qualified lead: a submitted form vs a prospect who has a real need, a budget, and a timeline. Out of 100 form submissions, 20-30 are qualified leads. The other 70 are noise.
CPQL requires CRM integration. That's where offline conversions come into play. Without a CRM, you're optimizing on form fills. With a CRM, you're optimizing on potential customers.
6. Customer Acquisition Cost (CAC)
Formula: Total Spend / Number of Customers Signed
The ultimate KPI. How much does a new customer cost you via Google Ads. It's the only number your C-suite cares about.
CAC requires end-to-end tracking: from the Google Ads click to the signed contract in your CRM. It's complex to implement but transformative for decision-making.
Building a Reporting Dashboard
Weekly Report (Operational)
Frequency: every week. Audience: the media buyer.
| KPI | What to look for |
|---|---|
| Spend | Budget consumed vs planned |
| Clicks | Traffic volume |
| CTR | Ad quality |
| Average CPC | Bid trends |
| Conversions | Lead volume |
| CPA | Cost per lead |
Monthly Report (Tactical)
Frequency: every month. Audience: the marketing manager.
Everything from the weekly report plus:
- ROAS or CPQL
- Conversion rate by campaign
- Average Quality Score
- Impression share and losses (budget vs ranking)
- Top keywords by performance
- Budget allocation by campaign
Quarterly Report (Strategic)
Frequency: every quarter. Audience: leadership.
- CAC (customer acquisition cost)
- Revenue generated vs spend (actual ROAS)
- CRM pipeline attributed to Google Ads
- Comparison with other acquisition channels
- Budget allocation recommendations
Reporting Mistakes
1. Looking at Metrics Without Time Context
A $150 CPA this week means nothing without the trend. Is it rising or falling? On what volume? Always display KPIs with their trajectory.
2. Comparing Incomparable Campaigns
A brand campaign (your company name) will always have a better CPA than a generic prospecting campaign. Comparing them is meaningless. Segment brand vs non-brand in your reports.
3. Optimizing on Micro-Conversions
If you're tracking page views as conversions, your CPA will be $2 and your director will be thrilled. Until they discover those "conversions" generate zero business. Define conversions on actions with real business value.
4. Ignoring Assisted Conversions
The last-click model attributes all value to the final click. But an informational keyword that initiates the customer journey has value, even if it doesn't "convert" directly. Review conversion path reports in GA4.
KPIs in Your Optimization Strategy
Reporting isn't an end in itself. Each KPI points to an action:
- Low CTR — improve your ad copy
- Low conversion rate — improve your landing page
- High CPA — check your Quality Score and bidding strategy
- Declining ROAS — review targeting and audiences
- Dropping conversions — verify conversion tracking and Consent Mode
Google Ads optimization is a cycle: measure, analyze, act, re-measure. Good KPIs tell you where to act. Bad KPIs give you the illusion that everything is fine.
Not sure which KPIs to track, or is your reporting not leading to action? Book a free diagnostic — we'll build a dashboard that connects your Google Ads data to your business results.
Articles du meme cluster
Annonces Google Ads : Rediger des Textes Qui Convertissent
Annonce Google Ads : RSA best practices, formules de titres, CTA, strategie de pinning, et methodologie de test. Redigez des annonces qui cliquent et convertissent.
7 min de lecture
Consent Mode Google Ads : RGPD et Tracking
Consent Mode v2 Google Ads : fonctionnement, basic vs advanced, impact sur les conversions, modelisation, et mise en conformite RGPD. Le guide pratique.
6 min de lecture
Conversions Avancees Google Ads : Mise en Place
Enhanced Conversions Google Ads : ce que c'est, pourquoi c'est indispensable post-cookies, et comment les configurer via GTM, API ou tag manuel.
6 min de lecture