Taste Boundary Experiment
Build creative taste by testing controlled risks and learning where the boundary lies.
- Difficulty
- Advanced
- Time to result
- ~ongoing to results
- Steps
- 6
- Confidence
- 98%
The experiment treats the boundary of acceptable creative risk as something learned through controlled exposure. The marketer names a dimension to stretch—format, tone, topic, personality, or theatricality—and ships a reversible test executed well enough to provide fair evidence. Before release, the team defines the desired audience, useful response, and unacceptable downside. Afterward, it separates productive attention and memorability from rage views, confusion, or brand misalignment. A test that falls flat can often be left behind because audience memory is short, while a successful unconventional execution may create disproportionate recall. Each result moves the perceived boundary: the next experiment becomes slightly bolder, narrower, better contextualized, or avoids the area entirely. The method develops taste through deliberate feedback while rejecting the idea that all publicity is valuable.
Origin
Patrick Campbell used his unconventional churn keynote and a controversial Bud Light data post to explain how creative risk and feedback calibrated his taste on Marketing Against The Grain.
Core principles
- 01Taste includes knowing when an idea goes too far or not far enough.
- 02Boundaries become clearer through shipped experiments, not internal debate alone.
- 03A failed experiment is often forgotten faster than teams expect.
- 04A successful distinctive experiment can create disproportionate memorability.
- 05Risk should be controlled through competent execution and bounded exposure.
- 06Feedback must distinguish valuable attention from rage or misalignment.
How to run it
- 1
Choose the Boundary
Identify the specific convention you want to challenge, such as formal tone, predictable format, restrained personality, or topic selection. Explain why crossing it might improve attention or connection.
Pro tip Test one boundary at a time so the lesson remains interpretable.
Watch out Provocation without a strategic reason is not a useful experiment.
- 2
Bound the Downside
Select a reversible format, limited audience, or low-cost release. Define legal, ethical, reputational, and relational limits that must not be crossed.
Pro tip Use an event session, pilot series, or isolated post before changing the entire brand.
Watch out Do not use vulnerable people or misinformation as experimental material.
- 3
Define Valuable Feedback
Specify the audience and reactions that would indicate productive distinctiveness. Also identify warning signals such as rage views, audience confusion, or loss of trust.
Pro tip Include a downstream measure such as qualified conversations or message recall.
Watch out Engagement alone cannot distinguish approval, outrage, and mockery.
- 4
Execute Competently
Produce the experiment well enough that a failure tests the idea rather than avoidable sloppiness. Preserve clarity about what the brand stands for.
Pro tip Rehearse unconventional live formats and verify data behind sensitive claims.
Watch out A badly executed risk supplies little evidence about whether the underlying concept was sound.
- 5
Read the Full Response
Analyze reach, audience composition, sentiment, qualitative comments, recall, and business effects. Separate short-lived backlash from durable damage or useful learning.
Pro tip Wait long enough to observe whether the response persists beyond the initial spike.
Watch out Do not dismiss legitimate criticism merely because some online reactions are transient.
- 6
Move the Boundary
Decide whether to repeat, intensify, soften, contextualize, or abandon the approach. Record the specific lesson for the next experiment.
Pro tip Adjust only the dimension implicated by the evidence.
Watch out A single success does not make every larger risk safe.
In the wild
Campbell wanted to explore a keynote that felt more like a show. He delivered a story-driven presentation while saying only the word “churn” with different inflections. The experiment was unconventional but rehearsed and tightly related to the company's expertise.
→ The presentation became highly memorable and made the company a major topic of conversation at the conference.
Campbell published a data-oriented analysis of the Bud Light controversy to stretch beyond SaaS topics. The post earned substantial attention, including rage views and backlash, revealing that politically charged material could pull his personal brand toward an audience response he did not necessarily want.
→ He learned to add context, alter the treatment, or avoid similar studies rather than blindly repeating the engagement spike.
Common mistakes
Treating All Attention as Validation
Rage views may inflate reach while moving the brand away from the audience and identity it wants.
Taking an Unbounded Risk
An irreversible launch or severe ethical, legal, or reputational exposure is not a controlled taste experiment.
Learning the Wrong Lesson
A failure caused by weak execution should not automatically prove that the creative boundary itself was wrong.
Is it for you?
Best for
It is best for established marketers and brands that need more personality but can run reversible, observable experiments.
Not ideal for
It is not ideal for irreversible campaigns, regulated claims, vulnerable communities, or risks with severe legal or reputational consequences.
From the transcript
“And so I think it's doing more stuff like that earlier on, but the taste comes in, as you said, and like knowing where in…”
“But if it works out really well, I get a ton of people to come up to me and like, oh, that was awesome.”
“And so now I learned a little bit of like, oh, really controversial thing. I need to have a little bit more of this or…”
From the episode
$200M Founder Reveals Why Ai Can’t Replace Top 10% Of Marketers (#135)