Seven years ago, Krista Berlincourt stepped out of the office, walked to her car, sat down and hoped for divine inspiration. She had just gotten off the phone with a New York Times reporter, one of 87 calls she’d received from every major news outlet over a 48-hour window. Her company was in the midst of a firestorm, as the personal data of an estimated 800,000 customers had vanished on their smartphones. At the time, it was described as “the biggest disaster yet for the whole concept of cloud computing” and is still one of the largest mobile data outages in history. She was the last to leave that day, as the media waited for her response.
The incident forever shaped Berlincourt’s philosophy in responding to what she refers to as “brand threats,” the moments of alarm that have the potential to throw a company into a tailspin. She’s now the Head of Communications at fintech company Simple. Before that, she drove Developer and App communications for Microsoft’s Windows Phone after a stint supporting the company’s Startup Business Group. Today, Berlincourt actively advises startups with their communications strategy, most recently the Jaguar Land Rover Tech Incubator and previously companies participating in the Nike+ TechStars Accelerator.
In this exclusive interview, Berlincourt shares the principles and practices she’s honed over the years to help her navigate painful pinches in even the most highly regulated consumer industries. Here, she not only draws on her own career experience, but the National Forestry Service’s framework to triage and tackle forest fires, a methodology Simple actively employs. Whether you’re an early-stage company just beginning to build its brand or an incumbent under the watchful eye of billions, Berlincourt’s tips on how to proactively prepare and reactively respond to incidents will help you consistently manage and expertly douse any fires.
When things are spiraling, don’t spin with them. Settle into an intense sense of calm, like the eye of a hurricane.
Company crises — like wildfires — are complex developments that can progress in unexpected ways. In the heat of the moment, missteps in management can quickly result in a sullied brand and customer churn.
“When it comes to managing a company ‘crisis’ it’s important to remember that your brand is the gut feeling your customers have when they think of you. And because we’re emotional, intuitive beings, many of the feelings we use to describe our human relationships are also used for brands: love, trust, passion, hate. They're actually not all that different. So from your customer’s standpoint, any misstep is a mismanagement or misinterpretation of those feelings,” says Berlincourt. “You’ll notice that the same responses that offend customers are those that would also leave a close friend or family member in a lurch. You’d never just walk away or hang up the phone on a family member who needed your help, right? Treating each of your customers as generously as your inner circle may seem unscalable, but the truth is the more severe the incident, the more personal impact your customers will feel.”
There’s a framework to combat crises, but first let’s look into a few behaviours and reactions to avoid. Here are the most common mistakes Berlincourt has observed over her career:
Disappearance. “Much as you wouldn’t vanish on a friend in need, brands should avoid going radio silent. This is one of the most common, yet destructive, responses. This avoidance is built around fear, which breeds distrust and immediately puts you on the defensive even if you’ve done nothing wrong. When something's not right, the last thing a customer wants to be met with is defense. Then, more than ever, your brand needs to show up.”
Concealment. “When a brand disappears, it can be unclear whether it’s indecisive or intentional. But when a company knowingly hides an issue, there’s no doubt of the motivation. I’ve been shocked by the number of communications agencies who’ve suggested we simply ‘sweep it under the rug.’ Even the most buried crisis surfaces in the end. By lying to your customers, you’re saying ‘We don’t care about how you feel.’ Indifference is more lethal than incompetence.”
Inertia. “When something’s awry, speed and timing are of the utmost importance. But there’s a difference between acting expediently and hastily. It’s better to use the information available than to wait for additional insight before responding. Again, consider human relationships. Imagine that you’re in the middle of a particularly high-stakes text conversation, share some big news, then for the next thirty minutes you watch the dreaded text bubble ellipses bounce back and forth. You wait. You’re irritated. Your anxiety increases. You feel ignored. That ‘where the hell are you?!’ feeling is the same one your community has while waiting 45 minutes for you to get a statement out.”
Indirectness. “Skirting around an issue only prolongs the suspense. Failing to address the question continues the conversation. Reframing and deflection increase commotion and confusion. It’s more frustrating to your customers — and damaging to your brand — if they have to do the untangling.”
Automation. “Dehumanizing the conversation is a classic mistake. I call it ‘robot speak’ when there’s a scripted stiffness to conversations. Are you talking to humans? Then sound like a human. Robot speak is an acute stress reaction. As is the case in human psychology, brands often respond with emotional detachment and depersonalization. ‘We are deeply saddened by the fact that this incident has been brought to our attention…’ is the brand equivalent of ‘Well, it’s fine, but if you can make time, I was hoping we could possibly…’ Cut out the prepositions and get to the point.”
Assumption. “Take a minute to sit with the situation then use your energy to strip away the surface emotion and ego as much as possible. Rationally frame what you know. One of the errors we often see is when a company responds to a crisis by extrapolating beyond the information available. Then, once things shake out, they backpedal and it’s unclear if that’s intentional or not. Just as you’d do in your personal life, if you’re asked a question, and you only know 10% of the story, don’t offer a response that attempts to bridge the other 90%. It's okay to say, ‘I only have 10% of the information, and I'm working to find the other 90, but right now this is what I know.’ Be honest. As a mentor of mine says: there’s truth and there's truthiness. Offer the truth.”
Winging it. “If luck is what happens when opportunity meets preparation, a crisis is difficulty meeting unreadiness. In business, ‘incidents’ can spin up just about anywhere: a leadership change, a shift in the market, product error, employee churn or data breach. So on a certain level, you will always be reactive. Consider scaling for growth. As the law of truly large numbers says, at a certain size, any outrageous thing is likely to happen. You can’t account for the unknown unknowns, but that doesn’t mean that you should fly by the seat of your pants. Plan — it’s the only way you’ll have some level of comfort with the ‘what ifs’ should they happen. You can’t avoid every iceberg, but you can add a few more safety boats.”
Before you find yourself behind the firehose, prepare your company to firefight with you. Here are the principles Berlincourt has used to create the right culture for her team and tools to work most effectively during a crisis:
Seek to inform, not alarm.
Calling an incident a ‘crisis’ doesn’t benefit anyone — not the communications staffer on point nor the rest of the team. “For most people within an organization, when something goes very wrong, they’re acutely aware and may feel helpless in finding a solution. Over time, this feeling can morph into a belief that you’re also unable to accurately identify a risk,” says Berlincourt. “There’s a system to help with that, but the first step is to dull the visceral response to the word. Use the term ‘incident’ rather than ‘crisis.’ This is less semantics, and more a nuance that can have a real impact on your ability to handle challenging situations. Don’t yell fire if there’s simply smoke.”
At Simple, Berlincourt’s use of “incident” has a widespread effect. “I try to delineate between crisis and incident. The word crisis is scary. If I texted you and said, ‘Hey, I'm having a crisis,’ you would drop everything and come running. At Simple, we seek to inform, not alarm each other,” says Berlincourt. “When something unforeseen happens — whether it’s a hiccup or a disaster —- the panic threshold is different for every team member based on their past experiences. The last thing we want to do is activate undue panic.”
They are incidents. The word "crisis" summons panic and paralysis, not clear, measured action.
Flag early, flag often.
Only you can prevent forest fires. The same goes for brands. Build a culture where every contributor is not only empowered but encouraged to raise their hand when something seems to be going awry. “If you smell smoke, say something. We ask every member of the Simple team to flag incidents, and describe, not analyze them,” says Berlincourt. “For example, the best outcome is that they raise their hand and say, ‘Hey, I noticed this customer issue,’ or ‘We seem to be outpacing forecasted onboarding volume and applications are taking longer to process.’ We do this in a dedicated Slack channel. Then, a segment of the team will field, classify, and triage each issue with support from a technical team, as needed. More often than not, the smoke is eliminated before fire catches.”
Simple’s engineers introduced the U.S. Forest Service's model for incident classification. “It’s National Fire Danger Rating System is accessible and straightforward. The alerts begin with green, but any color beyond green is an incident. If there’s even an inkling from a ranger or the public that something’s afoot, it’ll raise the alert, even if there’s not an immediate action to take. There’s a sensitive trigger with information, but a more deliberate one with action,” she says. “The beauty of this color warning chart is that its basic categories are commonly-known and established, but are retrofitted for specific regions. For example, the Forest Service worked with a local agency in Southern California to develop an additional color warning to better assess the risks specific to that region. In the same way, companies should adjust their rating systems to best suit their risk tolerance given factors such as company-stage, industry and customer base. For Simple, which deals with banking and personally identifiable information, it’s flag early and flag often.”
Flag early, flag often. Otherwise, what a bucket of water would’ve doused now takes 100 firefighters to fix.
Thanks is the only answer — and last word.
What most crisis communication systems neglect is closing the loop with colleagues in the same careful manner as they do with their customers. “To build a culture that supports an effective crisis communications system, you need to thank each and every person who flags an issue,” says Berlincourt. “Even if it’s a false alarm or a blip, it's really important to meet them with compassion, and say, ‘Hey, thank you so much for flagging this, I think we're going to be okay, but we'll keep an eye on it’ as opposed to ’Why would you think this was an issue?’”
This tenet builds on a culture that would rather respond to every shout from the boy who cries wolf than get devoured unaware. “It's a cultural issue. If you have an environment where you even implicitly require 100% success, raising your hand becomes high-risk. In those cases, you’re going to feel the flames before you see the fire, because no one will tell you,” says Krista. “It’s the people on the frontline who see the most, right? For consumer tech companies, the front lines are fueled by customer service reps who may be new to the business. If they’re reluctant to raise their hand because bureaucracy will smack them down, don’t be surprised if an incident develops before you’re aware of it.”
Berlincourt believes that brands, like people, are measured most during bad times. It’s during these depths when values are tested. The result is integrity. However, for many companies, those intangibles dissolve when disaster strikes. To make sure that doesn’t happen at Simple, Berlincourt and her team turn to Simple’s company values and incident management system (IMS) system to ensure their message remains on brand.
The first tool in action is Simple’s IMS. “For any issue flagged by an internal system or someone at Simple, an alert is sent to a specific Slack channel. There are two classifications of alerts: technical and non-technical,” says Berlincourt.
Non-technical incidents have to do with people or relationships. “That could include a customer issue, a partner or vendor issue, or a social firestorm. These types of incidents vary, but generally don’t involve in-house code, so engineers aren’t automatically alerted. The operations or integration team is pinged and will decide if engineering needs to be pulled online to assist.”
Technical incidents have to do with product error. “At Simple, anything other than optimal product operation will immediately trigger PagerDuty and simultaneously engineering, communications, integration and customer relations are alerted. This removes the barrier of playing telephone and creates an instant response team who can immediately dig into an incident. So, even if something happens at 2am, all the right players are in place.”
Simple developed these tracks to empower different groups to tackle unique situations. “The two each require a different team with specific firefighting specialties, so we’ve separated them. You probably don't need an engineer working on the non-technical issues or a business operations specialist to help you debug infrastructure. I’ve found that trying to create one catch-all process only leads to confusion,” says Berlincourt. “The real time-saving tool is activated alongside PagerDuty: our status page.”
Automate a status page.
When technical alerts funnel through IMS, not only are the relevant teams paged, but Simple's status page is then updated to alert customers of any product error, slowdown or planned maintenance. “This transparency may seem too vulnerable to some companies but we return the trust our customers have placed in us to safeguard their financial lives. We have a clear mission to help our customers feel confident with their money,” says Berlincourt. “This means giving them the option to self-serve the question: ‘Is my bank account normal?’ Instead of waiting or debating on when to update the status page or waiting for clarity on a technical issue, we can just switch the color and focus on fixing things.”
Similar to the National Fire Danger Rating System, here’s what the status page looks like in a few different modes:
The status page also serves as a running log of past incidents. They are not swept under the carpet, but shown on a feed. Any scheduled maintenance also appears here to alert Simple’s customers of any interruptions in service in advance.
To gauge transparency, measure how able your customers are to self-serve during a crisis.
A large part of why the status page works is the integration with Simple’s incident management system. “Our old status blog was a Wordpress site managed by Engineering. Incidents were described in technical terms, and while accurate, this wasn’t always understood by your average customer,” says Berlincourt. “Plus there wasn’t a way to stay current until the next update was posted. That sufficed for a young company, because our customers were fellow nerdy early adopters.”
Now when an incident is flagged due to interrupted service — say a slowdown with instant transfers or 10-minute glitch in photo check deposits — the feature’s indicator is changed to yellow or red, depending on severity. More than creating an efficient system, the status page elevates Simple’s quest to strengthen and return trust. “On a grander scale, our vision is to create a self-driving bank account. So fundamentally, if a customer can't answer her own question, then we have a problem. “Being able to see when a service was last offline and check its current function builds trust. We believe your money is yours, you shouldn’t need to call us to answer basic questions. We’ll never stop working to continue earning our customers’ trust.”
Triage the incident with four questions and a framework.
When the alert simultaneously convenes the relevant cohort and updates the status page, the team is afforded a moment to step back and triage the incident. It can now be timely and thoughtful, by methodically working through the following questions to develop messaging while the issue is being resolved. They should be addressed, in this order:
What do you know? “Think about your business, then walk through each group and consider the incident's impact — technical, brand, timing/duration, risk, compliance, and so on. Clarify what you know.”
What don't you know? “If you've uncovered gaps, can you answer them before responding to your community? If not, be clear that you simply don't have all the information rather than extrapolate on what might be or could be. Often, this evolves into what could be a lie once you fill in the holes later.”
What can't you say? “For risk, compliance, privacy (personally identifiable customer information) and other legal reasons, there are things companies cannot say, even if you really really want to. This will help you you create boundaries for messaging.”
What do you want to say? “Now that you've corralled all the information that exists, identified gaps, assessed how long it might be until you know more, and the things you can't say, what do you want to say? Start with bullets, then go from there. One team agreement we have at Simple, no lies — lying is lame.”
Simple also uses a risk classification framework to diagnosis the severity of an issue and the resources it should supply to fix it. It supplements the four questions by helping to measure the potential damage after an incident occurs. “Any incident can be translated into five tiers of severity. It's all about a company knowing and understanding their priorities and what matters in times of urgency,” says Berlincourt. “For Simple, we consider how many internal teams are impacted and customer impact, above all else. We also consider internal impacts and if our external partners need to be kept in the loop.”
Issue a statement based on what you know now.
Once an incident’s been assessed and the internal team’s mobilized to resolve, it’s time to translate this for your customers and broader community. While Berlincourt has developed guidelines for how to respond, she shies away from specific tactics on what to say. “One of the major sins in handling incidents is becoming robotic, so it’s important that your statement be specific to the incident, not ‘canned’,” says Berlincourt. “Tips and tools that we’ve developed to help guide — but not script — our statements include a three-point message, no-timeline default and a no-fly-zone list.”
Three-point statement. “This approach is particularly valuable for especially challenging, entangled incidents. It’s a brief update that follows this format: ‘We are [insert feeling]. We are looking into it. We are here.’ That first sentence should reflect how the company’s feeling on the broader whole. The second sentence confirms you are in pursuit of a solution, and the third statement indicates that you are accessible and available. Many companies could avoid PR fiascos, customer revolts and fiery headlines if they execute their incident management systems and issued a similar response immediately while working on a solution. Lastly, it’s as critical not to over-promise. ”
No default for a timeline. “If you've tapped everyone in the company and still have only 10% of the story, then that is what you know. Expanding your response beyond that 10% would be a lie, even if unintentionally. You're extrapolating with good intent, but you’re trading future pain for temporary relief. I see this a lot with timelines. Your company may say, ‘We'll have more to share in an hour.’ Well, you’ve then committed to an hour. Now are your engineers on that same timeline? What if you don’t have more to share in an hour? You don’t want to blindly and publicly impose a deadline on your broader team. Don’t be the comms lead who speaks for the company but is completely detached from the work that actually needs to be done. Assess what you know and align internally.”
No fly zone list. “A no fly zone is like a safety-belt for empathy. Create a list of words during the best of times that you’d hate to see in a statement, then reference what’s off-limits to keep you on track when an incident hits. These generally fall into three categories — alarmist, careless and robotic — that, together, will help maintain an informed, friendly brand voice. For example, you’ll sound robotic if you’re hyper-formal, but could come off as careless or flippant if you lean toward too casual. One of the words I most hate is ‘oops.’ It’s becoming common for modern, conversational brands that try to sound playful and take the sting out of an incident. We make it a best practice to review incident messaging periodically and red line where we have room to grow.
Close the loop with a learning review.
Every incident needs closure. “We all share a belief in the importance of a growth mindset. We can always improve. That's not to say that you need to get better, but you can. We reinforce this concept in the naming of our process for closure. Rather than ‘post-mortem’ — which literally means an examination of a dead body to determine the cause of death — we use the term ‘learning review,’” says Berlincourt. “Just as we flag incidents, we encourage anyone in the organization to raise their hand and suggest a learning review. We’ll assemble a small team and run through the typical questions: What worked here? Where did we get caught up? Why did it take us so long? What would we do differently next time? How does our language stack up without the chaos of a time-crunch?"
One more unusual addition is how Berlincourt’s team brings checks and balances into closure. “We also have a brand voice lead, who monitors and evaluates brand tone and language across the company. He’s not on the team that responds to incidents, so he offers an objective, outside perspective on our copy,” says Berlincourt. “In your closure or feedback loops, I’d recommend integrating people who aren’t intimately involved in incidents to join your learning reviews after the storm has cleared. It gets you out of your echo chamber and is another chance to learn.”
Wildfires, like company crises, are powerful but manageable with the right fire and escape routes.
When Berlincourt joined Simple, the team was just over 20. Today, the company is over 300 employees. Customer account growth has increased by more than 40% over the last two quarters. The stakes are high and incidents must be met with empathy and exactitude — you can’t just throw up a fail whale in banking.
To prepare for your company’s own inevitable “oh no” moments, learn to sidestep the seven sins of crisis communications. Create a culture where your colleagues can habitually flag incidents early and often. Seek to inform, not alarm. Once that culture is alive, overlay the tools to enable it. An honest status page, incident management system, risk classification framework, no fly zone lists and learning reviews are key tools — and all the more potent when synchronized as a system. This plan can help a lone firefighter build a process that empowers a team, even at scale.
“Fires can move from sparks to an all-out blaze very quickly. In fact, when the conditions align—temperature, humidity and wind—fire whirls have been clocked at more than 50 miles per hour. Compare that to the fastest human foot speed of 28 miles per hour and you begin to understand why you need a plan, a team, and focused response to have a fighting chance against the elements.” says Berlincourt. “The point is that, whether it’s a sprint or a long haul, even the most skilled individual can’t outpace a fire alone. And news travels even faster. If your company has the tools and a system to collectively and methodically handle incidents, you’ll prevent a scorched brand.”
Photography by Ryan Donaldson.