How to detect a crisis in your data

October 10 2016
Published in Tips and Guides

Would you be able to spot the first signs of a looming crisis in your data? Brands are increasingly turning to the use of monitoring tools to pick up on any budding PR issues before they have time to escalate - but with the hundreds of millions of news stories, blog articles, social media posts, photos and videos being published every day, the warning signs of a crisis are often difficult to detect amongst all the noise.

Visibrain has come up with a comprehensive guide explaining how - and where - to spot the first moments of a crisis in your monitoring data. Read on for an excerpt, or download the full guide to discover the four types of data that will help you to pick up on negative sentiment quickly, along with case studies illustrating just how an effective media monitoring strategy can make all the difference.


It may seem obvious, but an increase in the number of times your company is being mentioned anywhere online is one of the easiest ways to anticipate a potentially damaging situation. Even a subtle rise in volumes above their usual levels can provide you with an advance warning on an issue about to escalate.

In terms of volumes, the majority of crisis cases follow a distinct pattern, following three stages:

  • Phase 1: Weak signals
  • Phase 2: Media frenzy
  • Phase 3: Aftermath

The graph below shows a typical example of how mention volumes change as a crisis unfolds. As we can see, the full-blown “media frenzy” phase of the crisis, in which the brand is heavily targeted by the press and social media, is preceded by small bumps in volumes.

The different stages of a crisis in mention volumes

These brief spikes are an invaluable early warning sign, showing that the crisis has either broken on social media but has not yet been picked up by any key influencers - or it has broken in the minor press, such as a small news site or blog, and has not yet attracted too much attention.

From the start of the “Weak signals” phase, brands have a period of time in which to evaluate the situation and prepare their response. This window can last minutes, hours - or in some rare cases even days. If the cause of the problem is an unanswered complaint or a misconception, a fast, effective response can dampen the spread of a crisis, or even stop it completely. All the more reason to keep an eye on volume spikes.

Case study: Red Lobster

Red Lobster’s experience at the Super Bowl in 2016 is the perfect example of an opportunity turned crisis that could have easily been avoided. The restaurant chain experienced major embarrassment on Twitter after pop superstar Beyoncé released a new single - Formation - just before she performed at the Super Bowl half time show. The song’s lyrics mention Red Lobster restaurants - a celebrity shoutout that represented a golden marketing opportunity for the brand.

Fans held their breath, awaiting a response from Red Lobster…and waited, and waited. The brand took 8 hours to react to the endorsement, attracting widespread ridicule on social media.

So what happened? Normal levels of activity for Red Lobster average at around 70 mentions per hour. At 8.52 pm (GMT), mention levels started to rise, quickly going far beyond what was usual for the brand - by 8.57 pm, tweet levels had already gone over 200 mentions per minute, which should most certainly have set alarm bells ringing.

A mention volume graph for Red Lobster, just after news of Beyoncé's endorsement broke online

In just one hour after news of Beyoncé’s Red Lobster shoutout broke on Twitter:

  • The story had been shared by several celebrities and influencers, such as model Kendall Jenner and singer John Legend
  • There were 27,456 tweets about the brand sent by 19,205 individual users
  • The @redlobster Twitter handle was mentioned 2,344 times

As the tweets continued to pour in, Twitter became increasingly impatient at the lack of response from Red Lobster:

The brand finally responded 275,963 tweets later:

Understandably, Twitter was not particularly impressed with Red Lobster’s tardy arrival to the party:

Although the crisis didn’t do any lasting damage to the brand’s reputation and they still benefitted from the celebrity endorsement, the whole embarrassing episode could have been avoided entirely with an effective media monitoring strategy.

To avoid getting caught out by volumes

Red Lobster’s woes could have been avoided with the help of an alert system, set to notify members of staff when the number of mentions went beyond a certain number per minute or hour.

To determine how high to set your own alert threshold, it pays to know what a “normal” number of mentions is for your brand on a day-to-day basis. Once your alert is programmed, ensure that all relevant members of staff will receive it to maximise your chances of it being spotted quickly.

For more tips and case studies, download the full guide

To receive the full Visibrain guide, simply enter your email below.

This field is not valid

Thank you! Your download will begin shortyly...

Published in Tips and Guides