How to Measure Digital Marketing ROI: the only metrics that actually matter

How much did you invest in digital marketing last year? And how much did you get back? If you cannot answer in under 10 seconds, you have a problem — and you are not alone. 90% of Italian SMBs do not have a structured system for measuring marketing return on investment. They spend, hope and then decide "by feel" whether to continue or cut.

This article is not yet another list of marketing metrics: it is an operational guide for building a measurement system that tells you, with numbers, whether your marketing is generating profit or burning money. It is the same framework we use with every +Click client.

#Why 90% of businesses measure ROI the wrong way

The first problem is confusing activity with results. "We published 20 posts this month" is not a result, it is an activity. "We generated 47 qualified leads at a CPA of 12 euros" is a result. The difference sounds trivial, but 90% of the marketing reports we see talk about activities, not results.

The second problem is the lack of connection between marketing spend and revenue. The agency sends you a report with impressions, clicks and CTR. The sales team closes contracts. The accounting system records revenue. But nobody connects point A to point C: how much of that revenue comes from digital marketing?

• Agency report: talks about impressions, reach, engagement, CTR. These are inputs, not outputs.
• Sales CRM: talks about leads, deals, closings. But it does not know where the leads came from.
• Accounting: talks about revenue, margins. But it does not know which marketing channel contributed.
• The CEO: asks "how much is marketing earning me?" and nobody has a precise answer.

The framework we propose solves exactly this: connecting every euro spent on marketing to the euro that returns. Not perfectly — perfection in attribution does not exist — but with an acceptable margin of error and data-driven decisions.

#The 5 metrics that actually matter: CAC, LTV, ROAS, CPA, margin

There are hundreds of metrics in digital marketing. But to decide whether your marketing is working, you need five. All the others provide context — useful, but not decisive.

1. CAC — Customer Acquisition Cost

How much does it cost you to acquire a new customer? Not a lead, not a contact: a paying customer. CAC is calculated as: total marketing spend divided by number of new customers acquired in the period. If you spend 3,000 euros per month on marketing and acquire 15 new customers, your CAC is 200 euros.

CAC alone tells you nothing: you need to compare it with customer value (LTV). A 200-euro CAC is excellent if the customer is worth 2,000 euros over time. It is disastrous if the customer buys once for 150 euros.

2. LTV — Lifetime Value

How much is a customer worth across the entire relationship? Not the first purchase: the total of all purchases over time. For a restaurant, a regular customer is worth 50 euros per visit times 2 visits per month times 24 months, equalling 2,400 euros. For a dentist, a patient is worth the initial treatment plus annual check-ups plus referrals.

The LTV-to-CAC ratio is the king metric. The healthy benchmark is 3:1 — every euro invested to acquire a customer should return as 3 euros over time. Below 2:1, your marketing is not sustainable. Above 5:1, you are probably under-investing and leaving growth on the table.

3. ROAS — Return on Ad Spend

ROAS is the ratio between revenue generated by advertising and ad spend. If you spend 1,000 euros on Meta Ads and generate 4,000 euros in tracked sales, your ROAS is 4x (or 400%). Important: ROAS does not account for product cost, agency fees or your time — it is gross by definition. For real profit, look at the margin.

4. CPA — Cost per Action (or acquisition)

CPA measures how much a specific action costs: a lead, a booking, a download, a sale. It is more granular than CAC because you can measure it by channel, by campaign, by keyword. If your Google Ads campaign generates 30 leads from 600 euros of spend, the CPA is 20 euros per lead.

CPA becomes powerful when you compare across channels: Google Ads costs you 20 euros per lead, Meta Ads 15 euros per lead, email marketing 3 euros per lead. But be careful: CPA alone is not enough — a 3-euro lead that never converts is worth less than a 20-euro lead that closes at 40%.

5. Net margin post-marketing

Net margin is the definitive metric: revenue generated by marketing minus cost of product/service minus marketing cost (ads plus agency plus tools). If a customer is worth 500 euros in revenue, the product costs you 200 euros and the marketing to acquire them cost 100 euros, the net margin is 200 euros (40%). Everything else is noise.

#Vanity metrics: the trap of numbers that look good in reports

Followers, impressions, reach, engagement rate: these are the numbers that make the monthly agency report look impressive but do not pay invoices. We are not saying they are useless — they provide context. The problem is when they become the primary KPIs.

• Followers: 10,000 followers who never buy are worth less than 500 followers who convert. The number alone says nothing about the business.
• Impressions: your ad was "seen" 100,000 times? Maybe. Or maybe it was scrolled past in 0.3 seconds 100,000 times.
• Engagement rate: a post with 500 likes and zero sales is not a marketing success, it is free entertainment.
• Reach: you reached 50,000 people. Then what? How many took action? Reach without action is noise.
• Time on page: useful for understanding content quality, irrelevant if the content does not lead to a conversion.

Vanity metrics have a legitimate use: understanding whether your communication works at the awareness level. But they belong in the report as context, not as results. The result is: leads, customers, revenue, margin. As we explain in our guide on data-driven social media strategy, even organic social must answer to measurable business objectives.

#Attribution models: last click is dead, what to use in 2026

Attribution is the process of assigning credit for a conversion to the different touchpoints that generated it. The customer found you on Google, then saw an Instagram post, then clicked an email, then purchased. Who gets the credit?

Why last-click no longer works

The last-click model assigns 100% of the credit to the final touchpoint before the conversion. In 2026 this is like giving credit for a goal only to the player who made the assist — ignoring everyone who built the play. The average customer journey has 6-8 touchpoints before conversion. Last-click ignores the first 5-7.

Practical consequence: with last-click, email marketing looks like the best channel (because it is often the last touch) and brand awareness looks useless (because it is the first). In reality, without the initial brand awareness, there would be nobody to email.

Models that work in 2026

1. Data-driven (DDA): Google's or Meta's algorithm analyzes all conversion paths and assigns credit based on the real impact of each touchpoint. It is the most accurate model, but requires volume (minimum 300 conversions per month for Google, 50 per month for Meta).
2. Position-based (U-shape): 40% credit to the first touchpoint, 40% to the last, 20% distributed among the middle ones. A good compromise for lower-volume accounts.
3. Time-decay: more credit to touchpoints close to the conversion, less to distant ones. Works well for short decision cycles (e-commerce).
4. Marketing Mix Modeling (MMM): for budgets above 20,000 euros per month, a statistical model that analyzes the impact of each channel on overall sales. Expensive, but the only one that captures offline impact too.

For most SMBs we recommend: DDA if you have the volume, position-based if you do not. And in any case: look at the macro data too. If you spend 5,000 euros per month on marketing and revenue has grown by 20,000 euros per month since you started, the ROI is there — even if granular attribution is not perfect.

#How to build a measurement framework in 5 steps

You do not need a master's in data analytics. You need 5 clear steps, implemented once and maintained with 30 minutes per week.

1. Define conversions: what counts as a "result" for your business? Form submitted? Phone call? Booking? Purchase? Define a maximum of 3 primary conversions (those that generate revenue) and 2-3 secondary ones (micro-conversions that indicate intent).
2. Install tracking: Google Tag Manager, Meta Pixel plus CAPI, Google Analytics 4 with custom events. Every conversion must be tracked digitally — if you do not track it, it does not exist.
3. Connect the data: your CRM needs to know which channel every lead came from. Use UTM parameters on every link, integrate the CRM with Google Ads and Meta Ads, tag sources. Data must flow from click to sale.
4. Calculate metrics weekly: CAC, CPA by channel, ROAS by campaign, LTV of customers acquired in the quarter. No complicated spreadsheets needed: 5 numbers updated every week is enough.
5. Decide with data: every month, look at the numbers and decide — which channels to scale? Which to optimize? Which to cut? The decision must be based on data, not feelings.

#Dashboards with Looker Studio: what to monitor and how

Looker Studio (formerly Google Data Studio) is free, connects natively to Google Ads, Analytics, Search Console, Sheets and through connectors to Meta Ads, CRMs and virtually any data source. It is the tool we use for all +Click Ads client dashboards.

The ideal dashboard structure

A useful dashboard can be read in 30 seconds. If it takes longer, it is too complex. The structure we use:

• Row 1 — Key KPIs: total spend, total leads, average CPA, average ROAS. Four big numbers at the top, with comparison to the previous month.
• Row 2 — Trends: line chart with spend and conversions over the last 12 weeks. You need to see the trend, not the single day.
• Row 3 — Channel breakdown: table with Google Ads, Meta Ads, organic, email. For each channel: spend, leads, CPA, ROAS.
• Row 4 — Top performers: the 5 campaigns or keywords with the best CPA. These are the ones to scale.
• Row 5 — Worst performers: the 5 campaigns with the worst CPA. These are the ones to optimize or kill.

The advantage of Looker Studio is automatic updates: you connect the sources once and the dashboard refreshes itself. No more manual monthly reports. The +Click team configures these dashboards as part of the initial setup for every project.

#Real numbers: how we measure ROI for our clients

Theory is fine, numbers are better. Here is how we apply this framework with real cases.

• F&F Autoservice: 199 leads generated, CPL of 1.07 euros, average closing cycle 14 days. Effective ROAS calculated on average client value (car sale) exceeds 1200%. Tracking: Meta Pixel plus CAPI plus UTM plus integrated CRM.
• Selfiestreet: CPI of 0.44 euros, growth of plus 1100%. ROI here is measured on downloads plus activation plus in-app revenue. Average active user LTV: the numbers justified doubling the ads budget in the second quarter.
• Sabina Autodemolizioni: 40M-plus total views, 17M on TikTok. But the KPI that matters is the cost per parts request — 30% below the sector benchmark. Views are the means, not the end.
• Estethya Beauty: in the beauty sector the decision cycle is longer. We measure qualified leads (who book a consultation), closing rate and average treatment value. CPA must be read in the context of LTV.

I do not send you a report with likes. I send you a report with how much you spent, how much you earned and what to do next month to earn more. If I cannot give you those numbers, the problem is mine.

— Niccolo Giuseppetti, founder +Click

Every month, for every client, we produce a report with: spend by channel, leads by channel, CPA by channel, overall ROAS, actions for the following month. All in a Looker Studio dashboard accessible 24/7. Total transparency is a non-negotiable condition of our method. You can explore our approach on the case studies page.

The 5 most common ROI measurement mistakes

1. Confusing correlation and causation: revenue grew, marketing is active, but maybe it is seasonality. Use incrementality tests to verify.
2. Measuring too early: ad campaigns need 4-6 weeks to stabilize. Judging ROI after 7 days is like judging a real estate investment after one week.
3. Ignoring opportunity cost: if your internal team spends 20 hours per month managing marketing, those hours have a cost. Include it in the ROI calculation.
4. Not calculating LTV: measuring only the first purchase underestimates the real value of marketing. A customer acquired today might buy for the next 3 years.
5. Changing everything at once: if you change channel, budget, creative and landing page in the same month, you will never know what worked. Change one variable at a time.

#FAQ on measuring ROI