The Honest Guide to Measuring Marketing ROI

Everyone wants ROI data. Very few businesses measure it properly. This guide walks through the metrics that actually matter, the ones that mislead you, and how to set up honest reporting.

Most marketers can't measure their ROI accurately. This isn't ignorance — it's that marketing attribution is genuinely hard, and most off-the-shelf tools oversimplify it. A customer might see your Facebook ad, read your blog three weeks later, click a Google Ad the following month, and then convert from a direct visit. Which channel gets the credit?

The attribution problem

The customer journey is rarely linear. Someone might interact with your brand 7–8 times before making a purchase — across different channels, devices, and time periods. When your analytics tool credits a conversion to a single source, it's usually making a guess based on the last click, the first click, or some other simplified model.

This creates a predictable problem: channels that appear early in the journey (like brand awareness content) look ineffective, while channels that appear just before conversion (like branded search) look disproportionately valuable. Businesses make budget decisions based on this data and wonder why cutting the 'ineffective' channels damages results.

The ROI formula

The formula is simple. The difficulty is in the numerator: accurately determining which revenue is attributable to which marketing activity. For e-commerce businesses with short buying cycles and trackable transactions, this is relatively manageable. For service businesses with long sales cycles, multiple touchpoints, and offline conversations, it's much harder.

Metrics that matter vs vanity metrics

Attribution models explained

• Last-click attribution: 100% credit to the final touchpoint before conversion. Simple but overvalues bottom-of-funnel channels.
• First-click attribution: 100% credit to the first touchpoint. Overvalues awareness channels.
• Linear attribution: equal credit spread across all touchpoints. More balanced but still imprecise.
• Data-driven attribution (available in GA4): uses machine learning to assign credit based on actual contribution. Most accurate but requires significant data volume.
• Time decay: more credit to touchpoints closer to conversion. Good for short sales cycles.

Setting up measurement properly

Before running any campaign, define what success looks like in revenue terms. Not 'more brand awareness' — what's the target cost per lead? What's the target cost per acquisition? What conversion rate do you need from landing page to enquiry? Having these benchmarks in place before you spend a penny means you can evaluate performance honestly.

• Set up Google Analytics 4 with proper conversion events (not just pageviews)
• Connect your CRM to your ad platforms so you can see which leads became customers
• Use UTM parameters on every link in every campaign so traffic sources are correctly attributed
• Track micro-conversions (e.g., scroll depth, video views, form starts) to understand the funnel
• Build a simple monthly dashboard that answers: leads generated, cost per lead, leads to customers, revenue attributed

Key takeaway

Perfect attribution is an ideal, not a reality. The goal isn't to build a flawless attribution model — it's to measure well enough to make better decisions than your competitors who aren't measuring at all. Start with the basics, be consistent, and improve the model over time. A directionally correct measurement system beats no measurement system every time.