Confirmation Bias: Your Approval Chain Has the Same Blind Spot Callaway's Did. Here's how to fix it.

Your CEO is not your customer. Your board is not your customer. You are not your customer. Stop confusing internal applause for external validation

Last week, golf brand Callaway released a co-branded ad with content company Good Good Golf to promote a new driver. In the ad, Good Good's co-founder Garrett Clark pushes over brand influencer Alexis Miestowski while she attempts to grab a Callaway driver, before standing over her and threateningly saying : “Do not touch my new driver.”

The backlash was immediate and severe. Within days, Callaway had ended its partnership with Good Good entirely, retailers pulled Good Good merchandise from shelves, and the PGA Tour postponed a related broadcast.

When Callaway's CEO publicly apologised, he confirmed the company had reviewed and approved the ad before it went out — signed off by people who thought it was a good idea.

Read that again: The ad was produced, approved, and released by people at both companies, and no one along the way stopped to ask if getting attention for depicting violence against a woman was the attention they wanted.

The ad survived the room it was born in, but died in the room it was sent to.

Nobody who approved it was seeing it for the first time, cold, with no backstory.

Good Good's CEO admitted the ad was meant to be a parody of the horror film "Obsession." Inside that frame, the push wasn't violence — it was a reference, a wink, a bit of humour everyone on set understood.

But nobody outside that circle, scrolling for the first time, past a 60-second driver promo on social media, arrives with the context of the ad.

That's exactly the Multi-Audience Messaging Problem in action. The internal team was audience 1 (full context, inside joke, already agreed). The public was audience 2 (zero context, sees it once, cold), and what they saw was a man shoving a woman to the ground and warning her not to touch his belongings.

So how does a message survive a room it was never built for?

There's a name for the mechanism behind it: confirmation bias — the tendency to seek out, favour, and remember information that supports what you already believe, while dismissing or downplaying anything that doesn't.

In 1960, psychologist Peter Wason ran a simple test. He gave participants the number sequence 2, 4, 6, and asked them to figure out the rule behind it by proposing their own three-number sequences and getting a yes/no on whether each one fit.

Most people guessed the rule was "even numbers increasing by two" — and then spent every guess trying to confirm it: 8, 10, 12. Yes. 14, 16, 18. Yes. Confident, they announced their answer.

They were wrong. The actual rule was simply "any ascending numbers." Almost no one tested a sequence designed to break their theory. They only tested guesses built to succeed.

Why comms leaders are especially exposed

Be honest - you probably wrote the strategy. Or you were close enough to it that you already believe in it before anyone reviews a word.

The friendly review loop. 

So you run a draft past your closest allies instead of your critics. People who already trust you. Who already get it. Who already want it to work. And of course they'll approve it. (Sound familiar?)

Reading silence as agreement. 

You read a lukewarm response as "they'll come around," instead of a warning. Silence is rarely agreement. It's often people deciding it isn't worth the risk to disagree out loud with the CEO. (Everyone does this.)

Most likely will land badly: “The team's feedback on the new structure has been overwhelmingly positive.” (Based on three conversations with people who already supported it.)

Most likely to land better: “We've heard support from leadership and some real concerns from the floor about workload — here's how we're addressing both.”

Explaining away the outlier. One dissenting voice in ten gets written off as "just one person" instead of treated as an early signal. The outlier is often that one person who might just be the first to say what everyone else is thinking. (You probably already know who that person is.)

Four ways to break the bias in your approval chain

1. Find the person with zero context, and use them. Not a colleague who's loosely aware. Someone who wasn't in the room when the decision was made, doesn't know the backstory, and will read exactly what's on the page. This could be a manager/committee/skimmer from your homepage.

That's why it's important to anchor in your audience's world, not yours. Good Good's team was anchored in their own world — the reference, the inside joke, the shared context of the shoot. They never tested it anchored in the world of someone scrolling past a driver promo with none of that. The message has to make sense starting from what the reader already knows, not from what the room already agreed on.

2. Test with your critics, not your allies. Before you send, ask: who's most likely to have a problem with this, and have I shown it to them? If the answer is no, you haven't tested it — you're still at the rehearsing stage.

3. Give the outlier real weight. One person raising a concern isn't noise to filter out. Ask what they're seeing that the rest of the room might be avoiding saying.

4. Write down what would prove you wrong, before you send. Decide in advance what a bad reception would look like — specific reactions, not just "if it goes badly." It's much harder to explain away a signal you already agreed, in writing, would count as one.

Callaway didn't skip the review process. That's the part worth remembering. The ad went through people, multiple times, and every one of them said yes. They asked people who already had the reference, already had the context. And already agreed.

It was never about "reviewing more carefully." It was always about "reviewing differently." (Something Callaway's second apology addresses). Because somewhere in that approval chain, one person without the context, without the inside joke, without a reason to already agree would have seen it in five seconds. The bias wasn't in the ad. It was in who got asked to look at it.

------------------

Messaging designed to survive the rooms you're not in

Hi, I'm Vivien,

I help global B2B communication leaders and teams design messaging that survives across stakeholders, buyers, and employees — so critical decisions move faster, updates don't get ignored, and your credibility stays intact.

Sending something with real stakes attached — a restructuring, a layoff, a strategy shift? Get Executive Comms support

Trying to fix how your team tests messaging before it ships? Build your Key Messaging Library