Most content gets safer as it moves through an organization. Sometimes that reflects good judgment. Other times, it’s just risk avoidance.
Research on two-sided messaging suggests admitting something unfavorable can increase credibility depending on what you disclose and how.
But the same idea that makes a company more believable can also undermine trust or hand a competitor a useful attack line.
What separates a reasonable risk from a damaging one? How do you decide what to reveal, how far to go, and what support the idea needs to merit the risk?
1. What does the risk get you?
Start with the reason to say the uncomfortable thing at all.
Does it make an important claim more credible? Explain a tradeoff? Differentiate your point of view? Give the audience information they need to make a better decision? Reveal expertise they couldn’t see in the rated G version?
If removing the risky part costs the idea nothing, there may not be a reason to incur the risk.
Domino’s 2009 Pizza Turnaround campaign shows what it looks like when the risk earns it place.
Built around criticism of its flagship product, the recipe relaunch campaign repeated customer comments comparing its pizza to cardboard with ketchup.
That’s a significant risk for a global pizza brand to take in its own advertising.
But the criticism gave the turnaround claim something concrete to address. “We made our pizza better” is ho-hum. Revealing exactly what customers thought was wrong made the promised improvement more credible.
The same principle can show up in smaller ways: acknowledging a product limitation in an explainer, naming the downside of an otherwise attractive approach, or even saying “we don’t know yet” where your evidence stops.
2. Are you revealing a problem—or joining a conversation?
Negative information carries a different risk when the audience already knows it.
Research on two-sided advertising specifically identified existing negative beliefs, competitor attacks, and unfavorable publicity as situations where acknowledging the negative can be useful.
If prospects are already encountering a weakness in G2 reviews, silence doesn’t create a world where that weakness doesn’t exist. Silence leaves everyone else to interpret the weakness for you.
If you don’t like what is being said, change the conversation.
In Mad Men, Don doesn’t propose that the client deny the controversy around demolishing Penn Station. He reframes what the project represents: not the destruction of old New York, but the beginning of a new one.
Sometimes the opportunity isn’t avoiding an uncomfortable conversation or conceding the criticism. It’s putting your hands on the frame and changing what the known facts mean.
That doesn’t automatically make disclosure the right call.
Mild dissatisfaction, an existing rumor, and a widely documented product problem create very different situations—but before rejecting an idea because “people might think X,” establish whether people already think X.
3. Are you blowing the interview?
There’s a reason interview coaches tell candidates not to answer “What’s your greatest weakness?” by naming a skill required for the job.
You’re auditioning for a role in your audience’s budget. Don’t give them a reason to allocate resources somewhere else.
The pratfall effect points to why. In a 1966 experiment, a small blunder made a highly competent person more likable—but made a mediocre person less likable. The weakness only helped once competence was already established.
LinkedIn brand marketer Keith Browning imagined how the same principle could work in B2B.
Maybe an app loads more slowly because the company applies more rigorous security, or a business loan takes longer because the lender is more serious about responsible underwriting.
In both cases, the apparent weakness helps explain a strength the buyer values.
Now take a risk closer to the core promise. Imagine a company considering a headline like How we grew our NPS 30+ points in 12 months.
The improvement is impressive. The starting point also gives a competitor an easy attack. There must have been a lot to fix!
The closer the uncomfortable fact gets to your core promise, the higher the stakes.
4. What state is the uncomfortable thing in?
“Admit your flaws” is lousy advice on its own because the uncomfortable thing isn’t always a flaw.
Before deciding how to talk about the uncomfortable thing, classify what you’re actually dealing with:
- Fixed: “Our onboarding process was causing customers to drop off. We changed it, and here’s what happened.”
- Improving: “This still takes longer than we want. Here’s what we’ve changed so far and what remains.”
- Accepted tradeoff: “We deliberately limit this capability because maximizing it would compromise something we consider more important.”
- Unresolved: “We’re seeing this pattern. We don’t know yet whether A or B is causing it, but here’s what we’re testing.”
- Outside your control: “We can improve the conditions that make your company discoverable in AI answers. We cannot guarantee a particular model will cite you.”
Semrush makes its control boundary explicit. The platform can’t guarantee citations because AI answers are nondeterministic.
The state of the problem changes what the disclosure asks the audience to believe.
Fixed problems can demonstrate learning, an accepted tradeoff can sharpen positioning, and an unresolved problem can show intellectual honesty—but only if the company is clear about what it knows and what remains open.
Buffer offers a supreme example. The company had already made transparency part of its operating culture when it experimented with completely transparent employee feedback. It backfired.
People became less willing to give constructive feedback publicly. Anyone could pile into a conversation, and recipients became understandably more defensive.
So Buffer ended the experiment and returned constructive feedback to private channels.
Then it did the most Buffer thing possible and published When We Took Transparency Too Far, an article explaining why the experiment failed.
The lesson wasn’t “transparency bad.” Buffer learned that one of its values still required boundaries.
5. What gives people a reason to stay confident?
A learning story with a measurable outcome has an obvious obligation: show the outcome. If you’re going to tell people something was wrong and then say you fixed it, you have to provide evidence that makes your case.
But calculated risk isn’t one content format, and the thing that preserves confidence won’t always be a metric. It could also be:
- a mechanism
- a compensating strength
- a product boundary
- corrective action
- customer evidence
- an unusually well-supported argument
- a clear account of what you know, what you don’t, and what comes next
Databox used its company blog this way in a retrospective about developing its Benchmark Groups product.
The company describes assumptions that proved wrong, users abandoning its onboarding, reluctance to connect data, repeated redesigns, and later realizing it was still missing something and needed to rethink the approach.
These admissions don’t sit there unsupported. The piece shows the team’s reasoning as it evolves, turning the company’s experience into source material.
That’s what makes this kind of transparency so brand-building. Readers don’t just see what the company knows. They see how it knows, how it learns, and what happens when reality disagrees with it.
6. What’s the strongest reasonable case against you?
Before publishing something risky, try to make the opposing argument better than your most anxious stakeholder might.
Not the most malicious interpretation imaginable. More like: what could a smart prospect, customer, competitor, journalist, employee, or critic reasonably conclude from what you’re about to publish?
Remember the NPS headline? The hostile read is easy:
Company admits its product had serious problems.
Okay—now what? Would customers lose trust? Would competitors use it in deals? Would prospects start asking sales about it?
That last outcome isn’t necessarily a disaster if you’re having conversations.
A risky piece may surface an objection that previously lived only in the buyer’s head. It may give sales a harder question and a much better answer.
Or it may hand people a negative takeaway that’s much easier to repeat than your explanation.
You won’t know exactly which way the wind will blow before publication, but you can make sure the organization’s prepared for the conversation if the idea escapes the tidy context you built around it.
Robert Rose makes a related point in his piece on mining brand flaws for stories. In one consulting example, the company didn’t lead with the weakness. It used the exercise to prepare a consistent response in case someone else raised it.
Pressure-testing the risk doesn’t mean talking yourself out of the idea. It means knowing what conversation you’re inviting—and whether you’re prepared to have it.
7. How much exposure does the idea actually need?
The risky thing doesn’t always need its own piece of content. Sometimes it’s a sentence acknowledging a limitation, or a caveat beside a claim. Occasionally, it’s an entire product-development retrospective.
Other times, the risk starts with the headline.
How we grew our NPS 30+ points in 12 months does more than promise a result.
Before the reader clicks, it promises a confession. It reveals that there was substantial room to improve.
A safer headline could preserve the customer-experience and product lessons while burying the uncomfortable premise. But that would also bury the reason the story is interesting.
That’s why I prefer to treat transparency as an editorial posture rather than a deliverable.
A company like Buffer can make that posture visible across years of publishing. Another company might use it selectively when a limitation materially helps the audience understand a product. Databox uses its own operating experience as material for teaching.
You don’t need a Here’s Our Mistake series. You need enough room in the content strategy for reality to occasionally survive the approval process.
Take risks like an adult
There’s no score where six favorable answers mean the risk is officially safe. If a rubric could remove uncertainty, it wouldn’t be much of a risk.
These questions make the tradeoffs easier to see. Start with what the uncomfortable information adds and whether the audience already knows it.
Then look at how close it sits to the competence you’re selling, what state it’s in, and what will preserve confidence once you say it.
Pressure-test the strongest reasonable criticism. Decide what happens if prospects start asking sales about it. Choose how much exposure the risk actually needs.
After all that diligence, you still have to decide.
I take reasonable bets on my own blog all the time. Writing about work means deciding how much truth is useful, how much is gratuitous, what I can substantiate, and whether the point is worth the potential discomfort.
The instinct to protect your brand from unnecessary damage is reasonable. So is the instinct to publish work people might actually notice and remember.
Taking calculated content marketing risks means respecting both.
