Attribution Modelling and Minestrone

In the early days of advertising and media, the world was simpler—at least on the surface. Campaigns were built channel by channel: outdoor, radio, print, TV. Each played a role, each had a moment, and together they worked toward a shared outcome.

Fast forward to today, and while the channels have multiplied, the principle hasn’t changed. What has changed is our ability—and expectation—to measure how everything works together.

And like a recipe for a classic winter soup, this where attribution comes in. Again.

Attribution is often defined as assigning credit to touchpoints along the path to conversion. But that definition feels increasingly incomplete.

Today, attribution isn’t just about credit. It’s about understanding experience.

A modern customer journey isn’t linear—it’s fragmented, dynamic, and often invisible. A user might:

• Discover a brand via short-form video
• Research through search and reviews
• Engage with social content over weeks
• Return via a branded search or direct visit
• Convert after developing the intent to act or receiving a well-timed email or offer

And that’s just what we can see.

With cross-device behaviour, privacy restrictions, and platform fragmentation, much of the journey now exists in partial view. The challenge is no longer just which channel drove the sale, but:

What combination of interactions created enough momentum to convert—and how do we improve that over time?

Because understanding the journey is only valuable if it leads to better decisions, and better decisions compound into growth.

For years, last-click attribution dominated because it was easy. It gave us a clear answer—even if it was the wrong one.

But the industry has moved on.

Platforms like Google have shifted toward data-driven attribution models, while tools like Google Analytics 4 have removed last-click as the default mindset entirely.

Why? Because last click is like giving all the credit for a minestrone soup to the last ingredient added.

It ignores everything that came before—the base, the timing, the balance.

And in today’s environment, that’s not just simplistic—it’s misleading.

The biggest shift since this article was first written isn’t just technology—it’s privacy.

With the decline of third-party cookies, the rise of consent frameworks, and increasing platform walled gardens, attribution has become less deterministic and more probabilistic.

We’re now working with:

• Aggregated data instead of user-level tracking
• Modelled conversions instead of exact paths
• Platform-reported performance with limited transparency

This has forced a reset.

Rather than chasing perfect attribution (which no longer exists), leading marketers are focusing on directional truth—building confidence through multiple data sources rather than relying on a single version of reality.

The conversation has evolved from which attribution model is best to how measurement works as a system.

Today, that system might include:

• Platform attribution (Meta, Google, TikTok)
• Analytics-based attribution (GA4 or similar)
• Marketing Mix Modelling (MMM) for channel-level impact
• Incrementality testing to understand true lift
• First-party data strategies to reconnect journeys

No single method tells the full story. But together, they create a more complete picture.

Much like our minestrone.

But more importantly, this system creates the foundation for something else—continual improvement.

Because when measurement is connected, it doesn’t just explain performance. It enables you to refine it, iteration by iteration.

AI has accelerated optimisation—but also introduced opacity.

Campaigns now optimise in real time using signals we can’t fully see or control. Bidding, creative rotation, audience targeting—all increasingly automated.

This creates a tension:

• Performance improves
• Transparency decreases

Attribution, in this context, becomes less about explaining every decision and more about validating outcomes.

Are we driving incremental growth?
Are we investing in the right channels?
Are we improving efficiency over time?

These are the questions that matter now.

And answering them consistently is what enables performance to scale.

Interestingly, the future of attribution looks a lot like the past.

In traditional media planning, we understood that impact came from:

• The sequencing of channels
• The layering of messages
• The timing of activity

We didn’t credit a single touchpoint—we understood the system.

Today’s best-performing strategies take a similar view. They recognise that:

• Channels don’t operate in isolation

• Creative and context matter as much as placement

• Experience across touchpoints drives conversion—not just exposure

Attribution isn’t just a reporting tool. It’s a strategic layer that informs:

• Budget allocation
• Channel roles within the funnel
• Creative and messaging strategy
• Customer experience design

It underpins how Paid, Owned, and Earned media work together—not as silos, but as a connected ecosystem.

Like any great minestrone, the outcome depends on balance, timing, and quality of ingredients.

Your ingredients today might include:

• Search intent
• Social engagement
• Content and storytelling
• UX and site performance
• CRM and lifecycle marketing

Individually, none of these guarantee success.

But together—sequenced, optimised, and measured—they create something far more powerful.