Skip to content
English
Contact us
  • There are no suggestions because the search field is empty.

Marketing Measurement from Zero

[DRAFT - Pending Review]

Stage: Getting Started

Target: Anyone with zero measurement knowledge

Reading time: ~12 minutes

Purpose: You now know what digital marketing is. This module explains why measuring it is so hard — and how the industry is solving it.

Why Is Measuring Marketing So Hard?

Imagine you are walking down a busy high street. You see a TikTok ad for a skincare brand on your phone. Two days later, you Google the brand name. A week after that, you get a retargeting ad on Instagram. Finally, you click an email with a discount code and buy.

Four different marketing channels touched you before you bought. So who gets the credit?

This is the multi-touch problem — and it is the central challenge of marketing measurement. In the real world, customers do not see one ad and immediately buy. They see dozens of touchpoints across days or weeks before converting. Figuring out which touchpoints actually mattered is incredibly difficult.

Last-Click: The Default (and Why It Is Broken)

Most measurement tools use last-click. It is simple: whichever ad or link the customer clicked immediately before buying gets 100% of the credit.

In our example above, the email with the discount code gets all the credit. TikTok, Google Search, and Instagram get nothing — even though they all played a role.

Why is this a problem?

  • It massively overcredits demand capture channels (Google Search, retargeting, email) — the ones that catch people who were already ready to buy
  • It gives zero credit to demand creation channels (TikTok, YouTube, Meta awareness campaigns) — the ones that actually made people want the product in the first place
  • The result: brands cut spending on the channels that build demand, over-invest in the channels that capture it, and wonder why overall growth stalls

Think of it like this: Last-click is like giving all the credit for a football goal to the player who tapped it in from one yard out — ignoring the midfielder who created the chance and the striker who beat three defenders to set it up.

The Privacy Shift

The old approach of tracking every individual click and building a user-level journey is no longer viable at scale due to privacy changes. The industry needed a fundamentally different approach.

Enter: Media Mix Modelling (MMM)

MMM is a statistical approach that uses aggregate data — total spend, total impressions, total conversions — to figure out the relationship between marketing activity and business outcomes.

Instead of tracking individual users (which is increasingly impossible), MMM looks at patterns across the entire dataset. It asks: "When we increased TikTok spend by 20%, what happened to total revenue? What about when we paused YouTube for a week?"

The strengths of MMM:

  • Privacy-safe — works on aggregate data, no individual tracking needed
  • Cross-channel — can measure every channel in one model, including ones that do not generate clicks
  • Impression measurement — can credit channels for the views and impressions that influence purchases, not just the last click

The traditional weakness: Old-school MMMs (run by consultancies like Ekimetrics or Analytic Partners) are quarterly. They take months to deliver results. By the time you get the answer, the market has moved. You cannot make daily budget decisions on quarterly data.

Fospha's Daily MMM: The Evolution

This is where Fospha comes in. Fospha took the principles of MMM and made them daily, automated, and granular.

What makes Fospha different:

  • Daily refresh — models retrain every day, not every quarter
  • Ad-level granularity — not just "Paid Social works" but "this specific TikTok campaign is your top performer"
  • Full-funnel — measures both clicks AND impressions, so brand-building channels finally get fair recognition
  • Glass box — transparent methodology. You can see how every number is calculated, unlike black-box tools

How it works (simplified):

  1. Start with GA4 data — the baseline, but known to be incomplete
  2. Run four click measurement models — ElasticNet, XGBoost, Google's Last Click, and Google's DDA. Average the results (like four people independently checking each other's math)
  3. Add post-purchase data — discount codes and surveys capture channels that clicks miss (influencers, podcasts, word-of-mouth)
  4. Reconcile — GA4 undercounts sales by 20-40%. Fospha reconciles against your ecommerce platform's actual order data
  5. Measure impressions — using machine learning, credit the brand-building channels (TikTok, YouTube, Meta awareness) that created the demand but never received a click
  6. Output — a fully reconciled, cross-channel, daily, ad-level view trusted by marketing, finance, and leadership

You will dive much deeper into each of these steps in Stage 2 and Stage 3.

Why This Matters

Without proper measurement:

  • Brands overinvest in lower-funnel channels because they look like they are working (they are just capturing demand that brand channels created)
  • Brands underinvest in brand-building because it looks like it is not working (it is — you just cannot see it with last-click)
  • CFOs cannot trust marketing data because every platform self-reports inflated numbers
  • Teams make budget decisions on incomplete data — and waste significant spend as a result

With proper measurement (what Fospha provides):

  • Every channel gets fair credit for the role it actually plays
  • Budget decisions are backed by data, not gut feel
  • Marketing, finance, and leadership all trust the same source of truth
  • Brands are on average 30% more efficient than the market

Key Terms from This Module

Last-Click — Credits 100% of the conversion to the last ad/link clicked. Simple but deeply flawed.

Multi-Touch Attribution (MTA) — Tracks individual user journeys across touchpoints. Increasingly broken by privacy changes.

Media Mix Modelling (MMM) — Statistical approach using aggregate data to measure marketing effectiveness. Privacy-safe and comprehensive.

Daily MMM — Fospha's approach: MMM principles applied daily with ad-level granularity. Not quarterly, not batch — always on.

Impression Measurement — Crediting channels for the views and impressions that influenced a purchase, even when the customer never clicked the ad.

Full-Funnel Measurement — Measuring marketing impact from first impression to final purchase, across every channel and every stage of the customer journey.

Quick Self-Check

Before moving on, you should be comfortable answering:

  • Why does last-click overcredit demand capture channels?
  • What changed with iOS 14.5 and why does it matter for measurement?
  • What is MMM and how is it different from MTA?
  • Why is Fospha's Daily MMM different from traditional quarterly MMMs?
  • Why does underinvesting in brand-building hurt overall growth?

What is Next

In Stage 2, you will meet Fospha properly — our products, how we measure, and what makes us different. Module 2.2 (Core 101) takes you through the full Model Waterfall step by step.