Performance marketing didn't kill risk, it just required a committee. Liquid Death built a $1.4 billion brand on creative that would never have survived one, argues Marketing Director, Dhiren Patel.

Can brands still take risks in performance-driven systems?

The short answer is yes, though modern day performance marketing is actively working against you.

Performance marketing has delivered exactly what it said on the tin. It made marketing

accountable, measurable and predictable, and that has delivered genuine value. The problem is what comes with it - systems built to reward what works will always filter out what does not, and over time you stop exploring. Creativity does not disappear; it just becomes something you need permission (and a lot of senior approval) for.

Risk has been quietly domesticated

Most organisations have not eliminated risk, they have simply redefined it. It shows up only once performance is stable and targets are comfortably hit. What follows is pre-approved, defensible experimentation, safe enough to survive budget reviews. Which is not genuine risk by definition.

The data gap nobody talks about

MMM models, attribution platforms and incrementality testing are all useful, though none of them are complete. Several things that genuinely move the needle simply do not show up in a dashboard. Examples of which include:

● Brand awareness that builds gradually over time

● Cultural relevance that drives organic word of mouth

● The lengthy delay between someone seeing an ad and eventually converting

When partial data gets treated as the full picture, assumptions go unchallenged and the range of ideas making it through approval shrinks accordingly.

Why everything looks identical

Scroll any social platform for five minutes and the pattern is obvious. Same hooks, same format, same pacing but a different logo. The logic is the same and the approach delivers efficiency, predictable returns and something easy to sign off. That advantage is eroding faster than most teams realise, because as AI-driven optimisation becomes standard, templated creative is no longer a differentiator. Left unchecked, AI accelerates uniformity by learning from past performance rather than pushing forward. Ideas that actually cut through are always the hardest to justify internally, making them the first to get killed. 

However, one brand, Liquid Death, and its success springs to mind in clearly demonstrating how taking risk can deliver in a world of performance driven systems.A canned water brand, Liquid Death was built purely on creative which no performance committee would have ordinarily approved (skulls and heavy metal designs coupled with killer slogans like "Murder Your Thirst").

With no conventional ad formats and no category playbook, it had hit a $1.4 billion valuation by March 2024, doubling in value within two years, on the back of $333 million in annual revenue (Bloomberg, March 2024). The product is ubiquitous and available freely in the western world, the only real differentiator was a risky creative.

 What you actually need to do differently 

There are three practical things worth considering:

● Ringfence a portion of budget specifically for experimentation and protect it when pressure hits

● Stop evaluating every idea against short-term attribution, because some things take timeto show up

● Back ideas that are directionally right, even when the numbers are not there yet 

The simplest test: if an idea can be fully justified before testing, it probably is not genuine risk. The current performance systems are designed to eliminate uncomfortable ideas, not surface them. Taking real creative risk inside organisations means building a deliberate override and nobody else, including AI, can build that for you.