Managing Reputation Amid theRage-Bait Economy

The internet doesn't need your organization to do anything wrong before it can damage your reputation.

I've been thinking about that while watching the Lindsay Clancy murder trial unfold here in Massachusetts. The underlying case is horrifying. The legal question is also specific: whether Clancy, who killed her three children in 2023, was criminally responsible at the time. The Associated Press reported this week that prosecutors argue she acted intentionally, while her defense says severe mental illness left her unable to bear criminal responsibility.

Online, that bounded question has become something much bigger. My feeds have been full of amateur analysis and clipped testimony, plus medical theories and confident declarations from people with no access to the full evidentiary record. Comsint Communications' internal research counted more than 18,000 Reddit posts and comments tied to the case this week alone.

Then I heard people discussing the same theories at the beach.

That jump from digital to analog matters. Something doesn't have to become credible before it becomes familiar. And familiarity has consequences.

Lisa Fazio and her colleagues have spent years studying what psychologists call the illusory truth effect. In a 2015 study published in the Journal of Experimental Psychology: General, they found that repetition can make false statements seem more truthful even when the person hearing them already knows the correct answer. Fazio revisited the methodology in a 2020 paper in Collabra: Psychology and got the same basic result after measuring participants' prior knowledge before exposure. Knowing better helps less than most of us would like to think.

For corporate leaders, that's the part worth sitting with. Your board members have good judgment. So do your employees. The customer who has known your company for 20 years may be perfectly capable of separating fact from nonsense. Repetition still changes the conditions under which all of them are making that judgment.

The economics are stacked against the boring truth

There's also a business model sitting underneath all of this.

In March, the SIMODS research project led by Science Feedback published its second measurement of misinformation across six major social platforms in four European countries. On YouTube, low-credibility accounts received roughly 11 times more interactions per post per 1,000 followers than high-credibility accounts. On X, the gap was about tenfold. The researchers found the same general "misinformation premium" in two separate measurement periods.

Money follows attention. SIMODS found that 81 percent of eligible low-credibility YouTube channels appeared to be monetized. Motive varies by creator. The system can still pay people who repeatedly produce low-credibility material, while the engagement numbers tell them exactly what works.

Anger works particularly well.

A 2021 study led by Cambridge researcher Steve Rathje analyzed more than 2.7 million social media posts from U.S. politicians and news organizations. Published in the Proceedings of the National Academy of Sciences, it found that each term referring to a political out-group increased the odds that a post would be shared by 67 percent. Content about the other side was especially good at traveling.

Companies get pulled into that machinery even when the original dispute has little to do with them. A hospital can become evidence in somebody's argument about government. A university can become a proxy fight over class or politics. A food company can find itself drafted into a cultural dispute because one ingredient or supplier makes a useful villain.

Once that happens, the old crisis instinct becomes dangerous: answer everything.

The instinct is understandable. Someone says something false about you, so you correct it. Another account repeats a mutated version, so you correct that one too. A creator with 70,000 followers asks for comment on a claim that started with an anonymous account, and suddenly the communications team is spending its day chasing a story that had no fixed source in the first place.

Worse, every denial can create another exposure to the accusation.

Correction still has a place. The decision to respond should be based on who may believe the claim and what that person can do next. Those are very different questions from how many people saw it.

The 2018 Science study led by Soroush Vosoughi is useful here. Looking at 126,000 stories distributed on Twitter from 2006 through 2017, the researchers found false news traveled farther and faster than true news. The problem predates generative AI. AI can increase production, but the human appetite for novelty was already doing plenty of work.

Trying to beat a sensational falsehood at its own distribution game is usually a bad bet.

What communicators should do

Most crisis dashboards are built for the wrong contest.

They show mentions and sentiment. They also show reach. Those numbers can tell you whether something is spreading. They can't tell you whether the people with power over your organization believe it. A rumor viewed by two million strangers may matter less than the same rumor landing with six board members or the regulator who signs your license.

Reputation management in the rage-bait economy must get narrower. The objective is to preserve informed belief among the people whose decisions have consequences for you. That is a very different management problem.

That changes what preparation looks like.

Write the do-not-respond rule before you need it. Decide what level of spread alone will never trigger a statement. Then define the conditions that will, such as evidence that customers are changing behavior or that a regulator is asking questions. Legal counsel and the CEO should agree on those rules while nobody is angry.

Pre-delegate authority too. If a false claim is simple and materially harmful, somebody should already have permission to publish the verified fact without waiting for an emergency meeting. I've seen response plans fail because seven people had to approve twelve words.

Build one place where the truth lives. Stakeholders should know where the company puts material statements and primary documents. During a fast-moving problem, verification should be a lookup rather than an interpretive exercise.

The direct relationships matter even more. You can't build trust with the chair of your board or a major referral source at 9:15 on the night they're forwarding you a viral post. If those relationships already exist, you can tell ten consequential people what happened without issuing a public statement that introduces the accusation to 100,000 more.

The language of the response needs to change too.

In the corporate statements I see, the allegation often gets repeated in the first sentence so it can be denied in the second. That's tidy from a legal drafting standpoint. Psychologically, it may be doing the allegation a favor. Fazio's work is a good reason to write statements around what is true: "Our product contains ingredients A and B." "The employee remains with the company." "No customer records were accessed." State the verified fact. Don't turn your own channels into free distribution for somebody else's wording.

Silence can also be an active decision, but unexplained silence creates its own vacuum. A short line that the company is aware of online claims and will address material factual issues through its established channel can be enough. You don't owe every account a debate.

Boards need to understand this before the bad numbers arrive. If directors have been trained to interpret a spike in negative mentions as evidence that the communications response is failing, management will be pushed toward more public engagement precisely when restraint may be protecting the organization. The board should be asking a harder question: did any person with the power to affect us make a different decision because of what they saw?

That's the metric I would want at 8 a.m. the morning after a viral attack.

Ten million views can be ugly on a dashboard. One confused regulator or a spooked donor can be expensive.