Deny, Justify, Deflect: How Companies Communicate in a Crisis

Why companies make the same mistakes in a crisis so often

An order is late, a product is faulty, an employee gave a customer the wrong advice. The first reflex is rarely openness. People look for reasons outside themselves, stress what went well anyway, or hope nobody asks. That is human, and large companies behave no differently in public crises.

This article therefore asks: Why do companies resort to denial, justification, and deflection in a crisis, and does openness actually work better? To answer it, we first clarify the thinking model of impression management, then look at a study of audit firms and several experiments with two readings, and finally derive how an owner should speak in an emergency.

What is impression management?

Impression management refers to all attempts to steer the impression others have of oneself. The sociologist Erving Goffman described it in 1959 with an image from the theater. On the front stage, you appear before an audience and play a role; on the back stage, you prepare, speak openly, and drop the role. Anyone who has noticed how differently a team talks about a customer than with the customer knows the difference.

In itself this is nothing reprehensible. Every business shows its best side in its proposals, on its website, and in conversations. It becomes a problem when the front stage tells a different story from what is known on the back stage. In a crisis, this gap typically widens.

How audit firms presented themselves before an inquiry

Dunne, Brennan, and Kirwan (2021) applied Goffman's model to a case in which professionals had to speak under public pressure. They analyzed the content of the evidence given by the four large audit firms to the Irish parliamentary inquiry into the banking crisis. For the auditors, the inquiry was a front stage in the literal sense: public, on the record, with questions from members of parliament.

From the literature, the authors developed a framework for impression management and found four impressions the auditors repeatedly conveyed:

1. Our hands are clean: we are not to blame for the failure.
2. Our hands were tied: we had no way to prevent the failure.
3. Our work was good: we audited according to the applicable rules.
4. Our intentions are good: we act in the public interest.

The study describes patterns of self-presentation. It does not judge whether the individual statements were accurate. What is instructive is how familiar the four messages sound. They appear in almost every crisis, at a corporation just as at a trade business explaining to an angry customer why the delay was out of its hands.

If, in a crisis, you mainly explain why it was not your fault, you are answering a question the customer never asked.

What experiments show about denial and apology: two readings

Whether openness works can be tested with experiments in which participants assess invented crisis reports. The common framework is Coombs's Situational Crisis Communication Theory. It distinguishes crises by how much responsibility the public attributes to the company: victim crises such as natural disasters, accidental crises such as technical defects, and preventable crises caused by misconduct or negligence.

First reading: apologizing works better than denying

Ndone (2025) showed 686 participants in the US, recruited through an online platform, reports about an invented factory where workers had been injured. Depending on the group, the company responded with an apology or with denial. Those who read the apology rated the company's reputation better, were more willing to forgive and to buy, and were less inclined to speak badly of the company. The effect ran through acceptance of the account: what mattered was whether readers found the response appropriate and sincere.

In three experiments with a total of 1,718 population-representative participants, Hegner, Schilling, Durmaz, and colleagues (2025) found that the fact of issuing a statement at all had more influence than the strategy chosen. This fits a finding described in Reputation in a crisis: companies that disclose bad news themselves fare better than when others uncover it (Tachkova and Fediuk 2025).

Second reading: those wrongly accused may contradict

In two experiments, Fannes and Claeys (2022) studied companies that were merely accused of wrongdoing. There, an apology led readers to attribute more responsibility to the company than a denial did, which harmed its reputation. An apology is therefore also an admission. If you apologize for something you did not do, you make yourself guilty without being so.

The two readings do not contradict each other. They show that the choice of response must depend on the facts and not on what looks best. The problem is not denial as such, but denial against better knowledge. All the studies cited work with invented cases and respondents who suffered no harm themselves. Real customers may react more harshly.

What a documented case teaches about attacking critics

The collapse of the payment services company Wirecard shows what the attack pattern looks like on a large scale. From 2015, the Financial Times reported on inconsistencies in its accounts. The company rejected the reports. In 2019, Germany's financial regulator BaFin temporarily banned new short positions in the share and filed a criminal complaint against journalists. In June 2020, the company admitted that trust balances of 1.9 billion euros most likely did not exist and filed for insolvency. The article Chronicle of an accounting scandal recounts the sequence of events.

The presumption of innocence applies to those involved; the criminal proceedings have not been concluded. Regardless of that, the case shows the limit of impression management: a performance on the front stage lasts only as long as the facts on the back stage support it. The longer the gap grows, the greater the damage when it becomes visible.

What defensiveness can cost a small business

In small and mid-sized businesses, a crisis is rarely public. It plays out in phone calls, emails, and reviews. The cost of defensiveness is therefore hard to see but easy to calculate. A worked example: 20 customers are affected by a delivery error, each bringing in 25,000 euros in annual revenue.

The rates are assumed, not measured. Their direction, however, matches the experiments: a response perceived as sincere keeps more customers. The guide Delivering bad news describes how to address delivery delays, price changes, and your own mistakes in conversation. The same applies to public criticism online, as Online reviews shows.

The front-stage test: three questions before every crisis statement

From Goffman's model and the experiments, I derive a simple test that I call the front-stage test. Every external statement, whether a call, an email, or a reply to a review, goes through three questions.

1. Is it true, and do we know it?

   Separate what is established, what is suspected, and what is open. State only the first as fact. For what is open, name the time when you will know more.

2. Would we say it the same way back stage?

   If the team talks about the incident differently from the statement, the statement is not honest. The gap between the two is the risk.

3. Does it answer the affected person's question?

   The customer wants to know what happened, what it means for them, and what you are doing. Explanations of why it was not your fault belong at the end, if anywhere.

The test does not replace a legal review. Acknowledging what went wrong does not mean prematurely admitting liability. For personal injury, data breaches, or product defects, get qualified advice early.

What follows in general

Impression management cannot be abolished, and it need not be. All communication selects. The line runs where the selection deceives. For a business, this implies an attitude more than a technique: the front stage should be so close to the back stage that a customer who saw both would not be surprised.

This also applies internally. A team that punishes mistakes hears the same excuses back stage that customers do not believe on the front stage. How a business learns from mistakes without looking for culprits is described in Learning from bad decisions. The role the owner plays as the face of the company is covered in The owner as the face of the company.

Conclusion: openness works when it follows the facts

To the guiding question: companies deny, justify, and deflect because they want to protect the impression others have of them. Research on impression management shows that these patterns recur even among professionals under public scrutiny. Experiments show that in crises a company caused itself, a sincere apology works better than denial, and that silence helps least. A company wrongly accused may contradict, but with evidence and without attacking.

Openness is therefore not a matter of style but of consistency between what you say and what you know.

• Note the three most likely crises for your business and decide for each who will speak.
• Write a short template for one of your own mistakes: what happened, what it means for the customer, what you are doing.
• Run the front-stage test with your team on a past incident.
• Decide that criticism is always answered by addressing the factual question first.