What is attribution-agnostic advertising?

You and your agency run ads to drive foot traffic to store locations or a stall at a huge international expo. You got the usual mix - some social for activation and retargeting from DOOH, some search, maybe radio + TV if budgets allow. You’ll know it works if the daily per store averages increase or if more leads are collected at the stall. The whole methodology rests on being able to optimize the media mix and creative variations towards those that drive key metrics. 

But what if you won’t be able to measure those key metrics? What if there is no way to attribute it back to all the channels in your mix? Some would think your media plan should go into trash and your agency should be fired. Some would trust their gut and just throw more money at search. 

I don’t think signal loss is a media problem. Is it a measurement problem? Not really. It’s a strategy problem. 

Stupid Sniper’s Dilemma 

I get a sniper rifle with a 60x magnification scope and parallax adjustment that lets me reach a target over a 1.5km away. I shoot cans in my back yard with it cause it’s much easier to check if I hit the target. Later I bark at strangers despite having a dog. Who in the right mind would call me smart after all that?

You could guess that I’m using this shooting’n’barkin’ metaphor to point a finger at the shortsightedness of a lot marketers and their team who are stuck optimizing their strategies, performance, and creatives towards what’s easier to measure instead of what they actually need, and burning stacks of cold, hard cash doing it. At the same time, hiring agencies they don’t use for intended purposes (who needs a dog that barks, right?), while spiraling into insolvency.

I didn’t invent that - System1, Peter Field, and Adam Morgan analysed 55,000+ ads against IPA effectiveness data to find it takes 2.6X more media spend to achieve the same market share growth if your creative is dull. 

Dull is safe, because dull is measurable. That’s why Analytic Partners' research found that last-click attribution overestimates the impact of paid search by 190% while undervaluing brand-building by 90%.

The 6ft under, where the brands doing that are going, is also easily measurable.

All thermometers, no doctors

I love this concept in systems engineering called fault tolerance. You don't design a system and pray to god it doesn't fail. You design it to function when components fail. That’s why fixing measurement, adding more third party pixels, and clean rooms, data lakes, and more and more data-driven solutions are missing the forest for the trees. 

They are treating a symptom. And it’s a symptom of the wrong disease. 

The problem isn't that attribution is inaccurate (although it is, and often simply sucks, but I digress). The problem is that attribution was never designed to be a strategic decision-making tool. It was designed to be a diagnostic. Somewhere along the way, we promoted a thermometer to CEO and then wondered why the patient got sicker.

Attribution can tell you what happened. It cannot tell you what to do. 

Back to fault tolerance - what if we applied the same logic to advertising strategy? Before I scare away every single serious CMO and Marketing director, this is not about making all campaigns zero-click, anti-data, pro-brand print ads. I’m talking about a strategy designed to achieve its objectives regardless of whether attribution is accurate, a strategy where the question "but can we track it?" is irrelevant. Our strategic architecture doesn't depend on the answer.

I’m not a revolutionary here, and this isn't radical. It's how every brand that lasted more than 20 years was actually built. Coca-Cola didn't grow through attributed click paths. Neither did Salesforce, Caterpillar, or IKEA.

Attribution-agnostic strategy isn't new. What's new is that we forgot it was the default.

Attribution-agnosticism for beginners

Every marketing strategy resolves into five resource allocation decisions. Attribution-dependent strategy resolves them one way. Attribution-agnostic strategy resolves them differently.

Where to spend. Not "which channel has the best attributed ROAS" but "where do our category buyers actually spend their attention?" The input is media consumption data and competitive share of voice. Your Looker studio/Tableau report with last click attribution will not have an answer to it. Remember: 30% of search clicks are actually driven by other forms of marketing. The channel getting the credit isn't the channel doing the work.

How much to spend. Not "what's our marginal ROAS threshold" (although a good question) but "is our share of voice above or below our share of market? (a much better question)" If you want to grow, your voice must exceed your weight class. Punching above your weight is not a figure of speech here - brands with negative excess share of voice lost market share in 80% of cases studied.

What to say. Not "which message gets the highest CTR" which is a moronic premise on multiple levels, but "will someone remember this in six months when they actually have the need?", “will people start calling all tissues Kleenex and all diapers Pampers? ” System1 found that 92% of ads that score highest for long-term brand building also drive above-average short-term sales. You don't sacrifice activation by building memory. It never worked that way. Yet, 47% of US ads and 52% of UK ads generate zero emotional response, performing the advertising equivalent of setting a heap of money on fire to not even send a message.

Who to reach. Not "who's showing intent signals on the platform" but "are we reaching the full category buying pool: all the fine folks  who'll buy in 6, 12, 24, 48 months?" Your future customers aren't clicking your retargeting ads. They don't know they need you yet. At the rate I’m citing System1 here they would have to give me an affiliate link - performance advertising targets a niche group of those already interested, but when the pool of prospective buyers dries up, only brands that invested in broad reach have replenished it.

When to change course. Not "did this week's campaign metrics improve" but "have the leading indicators of brand health moved?" The System1/IPA Compound Creativity study found that the most creatively consistent brands grow market share 2X faster over five years with the same media spend. Constant optimization is the enemy of compounding. Brands that maintain the same creative agency within a 5-year period produce higher creative quality and grow their advertising distinctiveness. 

TL;DR - stop mistaking noise for signal in your strategies.

The Stack

Each of these decisions has a specific, measurable input, while none of them require the entire attribution stack to be correct. And most of them are available from tools you already have.

Les Binet's Share of Search metric, the share of organic Google searches for your brand divided by total category searches, is free via Google Trends (rush to use it before Google sunsets it as well), correlates strongly with market share, and can predict market share movements up to a year in advance. If your brand gets 400 searches/month and the category total is 4,000, your Share of Search is 10%. Track that against competitors quarterly, feel very smart, confident, and sexy. 

Combine that with competitive share of voice analysis (LinkedIn Ad Library, SpyFu, Pathmatics), pipeline quality metrics your CRM already tracks (average deal size, sales cycle length, win rate), and direct traffic trends — and you have a measurement stack that's harder to game, harder to misattribute, and more predictive of business growth than anything in your attribution dashboard.

This doesn’t sound like the strategist telling you “trust me, bro,” does it?

Which Way, Marketing Person?

Keep moving the chairs on the Titan submersible, optimising a broken system and paying more every year for diminishing returns. Watch the spark leave your creative team, as each next ad gets duller, your targeting gets narrower, and your buyer pool shrinks. Ride the doom loop all the way down.

Or design a strategy that works whether attribution is perfect, broken, or nonexistent. Invest in memory structures that compound. Build creative that doesn’t suck and gets remembered. Reach the buyers you can't currently track. Measure what matters instead and plan to stay in business for decades.

One approach gets more expensive every year. The other compounds.

If your strategy collapses when attribution doesn’t work, you never had a strategy - you had an arbitrage. And based. doesn’t deal with arbitrage, only strategy.  

‍

Sources & Further Reading

• The Multiplier Effect — WARC, System1, Analytic Partners, Prophet, Bera.ai (2025)
• The Extraordinary Cost of Dull — System1, Peter Field, Adam Morgan / eatbigfish (2023)
• Compound Creativity: The Magic of Creative Consistency — System1, IPA (2024)
• Defeating Dullness in B2B Advertising — System1 (2025)
• Share of Search as a Predictive Measure — Les Binet, EffWorks Global / IPA (2020)
• The Long and the Short of It — Les Binet & Peter Field, IPA (2013)
• Marketing in the Era of Accountability — Les Binet & Peter Field, IPA (2007)‍
• Binet & Field B2B Research — LinkedIn B2B Institute (2019)