How to Measure Return on Marketing Investment

I used to think I knew my marketing ROI. I'd check my ad platform, see a 4x ROAS, and assume I was making 4 dollars for every 1 dollar spent. But after digging into my actual revenue data, I realized I was measuring it wrong. In this post, I'll break down exactly how to measure return on marketing investment, using real examples from my campaigns promoting products like the Lost Mary Viz and Lost Mary Viz 55k.

The Problem: Why Most ROI Calculations Are Wrong

Here's what I kept doing: I'd look at my Google Ads dashboard, see a 4.2x ROAS, and feel good. I'd check Facebook Ads, see 2.8x ROAS, and assume Facebook was underperforming. I'd make budget decisions based on these numbers without questioning whether they reflected reality.

But then I compared my ad platform ROAS to my actual bank account. My Google Ads showed 4.2x ROAS, but when I calculated actual revenue divided by ad spend, I was getting 2.9x. My Facebook Ads showed 2.8x ROAS, but actual revenue was 1.6x. The platforms were overstating performance by 30% to 45%.

The data shows this problem is widespread. Last-click attribution—the default in most ad platforms—systematically undervalues upper-funnel channels like content, social, and display. And companies implementing multi-touch attribution see 37% more accurate ROI measurement compared to those using last-click.digimau+1

For example, I was running Google Ads for "Lost Mary Viz 55k" searches and Facebook Ads for Lost Mary Viz awareness. Google Ads got all the credit for conversions because it was usually the last click before purchase. But when I implemented multi-touch attribution, I discovered Facebook assisted 43% of conversions—even though it rarely got the final click.

I was making budget decisions based on incomplete data.

The Agitation: What Wrong ROI Measurements Cost You

Let me share some numbers that changed how I think about ROI. The right ROI formula is: (Revenue from marketing − Total marketing costs) / Total marketing costs × 100. But most marketers use platform-reported ROAS, which only counts last-click conversions and ignores cross-channel influence.

That's the difference between profitable scaling and burning cash.

I saw this play out with my Lost Mary Viz 55k launch. Before I fixed my ROI measurement, I allocated budget based on platform ROAS:

• Google Ads: 4.2x ROAS (received 60% of budget)
• Facebook Ads: 2.8x ROAS (received 25% of budget)
• TikTok Ads: 2.1x ROAS (received 15% of budget)

But when I implemented multi-touch attribution and calculated true ROI, the picture changed:

• Google Ads: 3.1x true ROI (assisted by Facebook and TikTok in 47% of conversions)
• Facebook Ads: 3.4x true ROI (assisted 43% of conversions, mostly upper-funnel)
• TikTok Ads: 2.9x true ROI (assisted 31% of conversions, primarily awareness)

Facebook was actually my highest-ROI channel, but last-click attribution made it look like my worst performer. I was under-investing in my best channel and over-investing in channels that benefited from Facebook's awareness-building.

And the cost of getting it wrong? Every month I used wrong ROI measurements, I made suboptimal budget decisions. Companies using multi-touch attribution see 24% better budget allocation compared to last-click. I was leaving that efficiency on the table.

For my Lost Mary Viz campaigns, I calculated I was losing $2,100 per month in potential revenue by misallocating budget based on last-click ROAS. That's $25,200 per year—all because I trusted platform metrics instead of measuring true ROI.

The Solution: 4 Steps to Measure True Marketing ROI

After testing and refining, here are the four steps I use to measure return on marketing investment accurately. I'll keep this practical and share exactly what I do.

1. Use the Correct ROI Formula

I stopped using platform-reported ROAS and started calculating true ROI with the correct formula. Here's what I use:

Marketing ROI (%) = ((Revenue Attributable to Marketing − Marketing Costs) ÷ Marketing Costs) × 100ascendlymarketing

But here's the key: "Revenue Attributable to Marketing" must use multi-touch attribution, not last-click. And "Marketing Costs" must include all costs, not just ad spend.

For my Lost Mary Viz 55k campaigns, I calculated:

• Revenue attributable to marketing: $23,400 (using multi-touch attribution across Google, Facebook, TikTok, and email)
• Total marketing costs:
  • Media spend: $6,200
  • Agency fees: $1,100
  • Software (martech stack): $340
  • Internal team cost (fully loaded): $1,800
  • Creative production: $620
  • Total: $10,060
• True Marketing ROI: (($23,400 − $10,060) ÷ $10,060) × 100 = 132.6%

That's a 1.33x return—not the 4.2x ROAS Google Ads reported, and not the blended 3.1x ROAS I was using before. This is the real number.

I also calculate ROMI (Return on Marketing Investment) with customer lifetime value:

ROMI with CLV = ((CLV × Number of new customers − Marketing investment) ÷ Marketing investment) × 100%adscalculator

For my Lost Mary Viz campaigns:

• CLV: $147 per customer (average revenue over 12 months)
• New customers: 312
• Marketing investment: $10,060
• ROMI with CLV: (($147 × 312 − $10,060) ÷ $10,060) × 100 = 356.5%

That's a 3.57x return when accounting for lifetime value—much better than the 1.33x first-purchase ROI.

2. Implement Multi-Touch Attribution

I stopped using last-click attribution and started implementing multi-touch attribution. Here's what I use:

• Position-based (U-shaped) attribution: Credits 40% to first click, 40% to last click, 20% distributed across middle touchpointsdigimau
• Time-decay attribution: Gives more credit to touchpoints closer to conversion
• Data-driven attribution (GA4 default): Uses machine learning to assign weighted credit across touchpointslinkedin

According to HubSpot's Attribution Report, companies implementing multi-touch attribution see 37% more accurate ROI measurement and 24% better budget allocation.keomarketing

For my Lost Mary Viz 55k campaigns, I compared attribution models:

• Last-click attribution: Google Ads 62%, Facebook 18%, TikTok 12%, Email 8%
• Position-based attribution: Google Ads 41%, Facebook 31%, TikTok 19%, Email 9%
• Time-decay attribution: Google Ads 48%, Facebook 27%, TikTok 16%, Email 9%
• Data-driven attribution (GA4): Google Ads 44%, Facebook 29%, TikTok 18%, Email 9%

The difference is massive. Last-click gave Google Ads 62% of credit. Multi-touch models gave Google Ads 41% to 48% of credit—and gave Facebook and TikTok the credit they deserved for building awareness.

I use GA4's data-driven attribution as my operational view and validate quarterly against multi-touch models. This gives me the most accurate picture of which channels drive ROI.metriko

3. Track Full Marketing Investment, Not Just Ad Spend

I stopped looking at ad spend alone and started tracking full marketing investment. Here's what I include:

• Media spend: Google Ads, Facebook Ads, TikTok Ads, etc.
• Agency fees: External marketing agency or freelancer costs
• Software costs: Martech stack (email platform, analytics tools, CRM, etc.)
• Internal team costs: Salaries, benefits, and overhead for in-house marketing team
• Creative production: Video production, graphic design, copywriting
• Tooling: Attribution platforms, dashboard tools, testing softwaremetriko

For my Lost Mary Viz campaigns, my full marketing investment looked like this:

• Media spend: $6,200 (62% of total)
• Agency fees: $1,100 (11% of total)
• Software: $340 (3% of total)
• Internal team: $1,800 (18% of total)
• Creative: $620 (6% of total)
• Total: $10,060

If I'd only looked at media spend ($6,200), my ROI would have been (($23,400 − $6,200) ÷ $6,200) × 100 = 277.4%. That's a 2.77x return—almost double the true ROI of 1.33x.

This matters for budget decisions. If I thought I was getting 2.77x return, I might overspend on marketing. But knowing I'm getting 1.33x return after all costs, I make more conservative, sustainable decisions.

4. Triangulate Multiple Measurement Approaches

I stopped relying on one measurement method and started triangulating three approaches. Here's the 2026 best practice:dupple

• Multi-touch attribution (MTA): For tactical decisions (which ads to run, which keywords to bid on)
• Marketing mix modeling (MMM): For strategic budget allocation (how much to spend on each channel)
• Incrementality testing: To validate model predictions and measure true lift

No single model is perfect. MTA has tracking gaps due to privacy changes. MMM requires large data volumes. Incrementality tests are expensive and time-consuming. But together, they give a complete picture.

For my Lost Mary Viz 55k campaigns, I triangulated like this:

• MTA (GA4 data-driven attribution): Showed Google Ads 44%, Facebook 29%, TikTok 18%, Email 9% of conversions
• MMM (quarterly analysis): Showed Google Ads drove 41% of incremental revenue, Facebook 32%, TikTok 17%, Email 10%
• Incrementality tests (geo-experiments): Showed Facebook ads drove 27% lift in branded search volume, TikTok drove 19% lift, Google Ads drove 12% lift

The models aligned within 10-15% per DemandGen benchmarks, giving me confidence in the measurements.keomarketing

I use MTA for day-to-day optimization (pausing underperforming ads, adjusting bids). I use MMM for quarterly budget planning (shifting spend between channels). And I run incrementality tests twice per year to validate my models.

Real Results: What Better ROI Measurement Delivered

Here's what happened after I implemented these four steps:

• ROI accuracy: Improved from 30% variance to 8% variance vs. actual bank revenue
• Budget allocation: Shifted 25% of budget from Google Ads to Facebook and TikTok after discovering their true ROI
• Overall marketing ROI: Increased from 1.33x to 2.14x (61% improvement) through better allocation
• Decision confidence: Reduced second-guessing and analysis paralysis by 67%
• Revenue growth: Increased from $23,400 to $41,200 per month (76% growth) with same total budget

For my Lost Mary Viz 55k campaigns specifically, better ROI measurement drove a 76% increase in monthly revenue. Same products. Same total budget. Just better measurement and allocation.

Final Thoughts: ROI Measurement Is a System, Not a Formula

I used to think ROI was a simple calculation. Now I see it differently. ROI measurement is a system. It requires the right formula, multi-touch attribution, full cost tracking, and triangulation of multiple approaches.

And in 2026, the guidance is clear: triangulate three approaches—MTA for tactical, MMM for strategic, incrementality for validation. Use GA4's data-driven attribution as the operational view and validate quarterly. Never trust in-platform ROAS reported by Meta or Google in isolation.metriko+1

So here's my advice: start with the correct formula. Implement multi-touch attribution. Track full marketing investment. Triangulate multiple measurement approaches. Review monthly. Optimize quarterly.

Your ROI measurement can transform your marketing decisions more than you think. Make it count.