Social Listening in Practice: Reputation and Crisis Monitoring

White Paper

The rapid growth and use of social networks and online media in the past two decades has given us unprecedented access to information. This has had a huge impact on brand and organizational communications. Social media has led to a total transformation of the processes required to respond in real time to breaking situations. Organizations need to be able to differentiate facts from rumors quickly. They must be organized, understand the situation, and then respond in real time.

Communicating during a crisis has never been easy. This paper is an introductory overview of best practice for crisis and issues communications in a social media age.

Get the download

Below is an excerpt of "Social Listening in Practice: Reputation and Crisis Monitoring". To get your free download, and unlimited access to the whole of bizibl.com, simply log in or join free.

download

Reputation Management - Best practice processes

Establishing effective procedures to prepare and rehearse for crisis and issues scenarios is standard practice in organizations worldwide. In every crisis and issues situation there is a four-stage process:

  1. Identification
  2. Evaluation
  3. Escalation
  4. Response

Let’s look at each of these stages in detail.

1) Identification

Whoever said “prevention is better than cure” must have been the first communications professional. As with medical issues, the best way to deal with reputational issues is to do everything in your power to prevent them from occurring in the first place.

There are events that will have an impact on your organization’s reputation that can be prepared for – such as service reduction due to bad weather, product recalls, data leakage, etc. Most organizations are aware of their ‘smoking’ or ‘smoldering’ reputational issues. These issues could be areas of operational or technical weaknesses, legacy sector/industry issues, political or activist challenges, etc. These problems are often being addressed by the organization, but could flare up quickly.

But many of the worst issues simply can’t be predicted and appear out of the blue. In these situations, quickly spotting the cues and keywords associated with sensitive events that are shared on social media is critical. This provides a significant head start when it comes to handling issues, and in many cases can prevent a relatively minor issue from developing or escalating into a larger one.

Most consumer-facing organizations now find about issues and crises through social media before they find about them from their own employees and associates. So establishing an effective social media listening process with social media monitoring tools is the first step any organization should take to help with crisis and issues communications.

It’s uncomfortable, but you should think of the very worst things that could happen to your organization. What are the most likely events that would cause an issue or crisis if they came to light?

Start with the scenarios that would stop your ability to function as a business (crises) then move through to less serious challenges (issues).

Next, establish a series of keywords that relate to how those issues would be described online and through social media. These keywords should kick-start how you track potential issues on a regular basis through monitoring tools.

With any monitoring tool, it is critical to establish a wide range of keywords that you don’t want to see associated with your brand or organization. These should include words or phrases that consumers, campaigners or critics might use.

Large businesses (especially customerfacing businesses with multiple locations like hotels, restaurants, and retailers) are likely to adopt a safety-first approach. Many encourage one person at each location to be responsible for some form of social media monitoring, alongside a central social media monitoring team.

A hotel chain may search for events like:

Fire AND hotelname, terrible AND hotelname threat AND hotelname stolen AND hotelname death OR die OR killed OR die AND hotelname strike AND hotelname An international retailer might also look for other sensitive issues such as: ‘child labor’ AND brandname(s) ‘illegal’ AND brandname

Small business owners may not have the same challenges as global organizations, but it is important to set up a targeted set of keyword searches that are closely related to potential issues as well as their trademarks, senior staff members, brands or products.

It’s worth noting that many social media monitoring tools have thresholds on the number of monitoring query results that are returned. Smaller businesses may need to prioritize how they’ll monitor issues-related phrases that are most likely to occur. The keywords they track will entirely depend on what business they are in and where they think the biggest risks are. Regardless of how small your brand is, someone should be responsible for checking all online media and social channels on a daily basis at the very least.

2) Evaluation

Regardless of whether it is a command center or an individual who first identifies a sensitive issue breaking online, the next step is to evaluate how serious it is.

Most organizations establish some form of ‘triage’ system. This means that when issues are identified, the comments are categorized and ranked before a decision is made of who needs to be informed and who needs to act upon the information.

Often communications and public relations specialists are involved in creating ‘triage’ systems - that’s if they haven’t already been included in predicting the possible issues and crisis events at the keyword planning stage.

There are always a number of factors to consider when ranking the seriousness of something on social media, including:

  • What is being said - the seriousness of the issue
  • Who is saying it - the influence or authority of the person who posts it
  • How often is it being said - is the issue getting more visible or less?

Some social media monitoring software and tools can help automate this process, but it always requires someone who understands the organization and the issues to decide exactly what happens next.

Techniques for evaluating social media mentions

Every organization will have its own models for assessing importance, but there are some fairly well-established guidelines worth considering if you need to develop yours.

What is being said? (The seriousness of the issue)

Social media software should be set up to categorize certain comments, statements or keywords into issues, which can then be ranked on a score of 1-10 according to what effect this might have on the organization. Alternatively, the tracked social media content should be ‘eyeballed’ by a skilled team member and labelled the same way.

Issues may be ‘severity scored’ along these lines:

1-3: Social media posts that are complaints, but appear to be one-off or manageable customer service issues.

4-7: Posts that are serious and are likely to have broader implications for the organization, especially if they escalate further.

For example: threats of a boycott, accusations of discrimination, regulatory or health and safety issues. These issues require very well tuned ‘antenna’ to foresee the potential dangers – see the Marriott Hotels case study below.

8-10: Posts that are likely to be a crisis, by pretty much any definition. For example, criminal or civil charges, fire, terrorism, death or serious injury, personal data-loss and anything else likely to cause an immediate negative shift in perceptions and/or affect share price.

Social media comments relating to a specific crisis will almost certainly be extremely negative at the start. It is worthwhile to establish processes to track and measure sentiment over time (i.e., are people feeling positively, negatively or neutrally?).

Changes in sentiment can prove an effective proxy to help decide whether your resulting actions and communications have a positive or negative effect on what people think of the issue and of your organization.

The importance of ‘fact checking’

During a major crisis, particularly one affecting a lot of people, it can sometimes be hard to separate fact from fiction. Unfortunately some people do try to deliberately mislead. Winston Churchill may well have been paraphrasing Mark Twain when he said,

A lie can get half way round the world before the truth has put his pants on

during the Second World War.

That was way before social media was invented, but never a truer word was said. During Hurricane Sandy, @comfortablysmug’s tweet about the flooding of the NYSE was retweeted more than 600 times – often by people concerned about friends or colleagues working there.

At one point during the day Piers Morgan was even interviewing experts about the impact of the NYSE being flooded.10 minutes later CNN said that the story may have been unconfirmed.

Who is saying it?

The ‘influence’ of an individual is not really important when someone posts irrefutable evidence of a crisis occurring (e.g., a plane crash), but in other issue scenarios, the more influential or authoritative the person who comments, the more likely they are to affect others’ perceptions of the issue.

Influencers can have a significant effect on the nature and outcome of an issue. It therefore helps to be able to understand how to spot ‘influential’ people or social media accounts.

However, influence is a nebulous concept. We’ve written a paper on this– Influencer Marketing, and the following case study explains the role of influence in escalating and deepening an issue.

Case Study 2 - Beverly Hills Hotel and The Dorchester Group

By the spring of 2014, Lesbian, Gay, Bisexual and Transgender groups had established a local campaign to boycott the Beverly Hills Hotel, on the basis that it was owned by the Sultan of Brunei. In Brunei, simply being gay could be punishable by death. The issue, from the hotel’s perspective, was relatively contained.

But global entertainment and business celebrities picked up on the boycott. They started using their social media accounts to highlight the issue and add their support. These ‘influencers’ brought the issue to the attention of international media organizations and millions of people worldwide. This rapid weighting of influence and exposure turned a contained event into a public relations crisis for the Beverly Hills Hotel and owners The Dorchester Group.

The hotel responded by establishing a ‘rebuttal’ website – The Beverly Hills Hotel Facts – but the damage was done. This case study highlights how issues that are believed to be contained can escalate quickly. With the right combination of influencers commenting, a seemingly localized issue can spread globally. Influence is impossible to truly measure, but there are some helpful methods that can be used to decide how much attention to give to online/social content or individuals.

The best social media software will make it relatively easy to receive threshold alerts, based on any of the criteria above. For example, it’s possible to automate a process such that if everyone with more than 1000 followers mentions your product, your CEO, or indeed any relevant keyword, the PR team automatically receives an alert.

How often is it being said?

One of the other things to track during and after a crisis is the volume of social conversation. You can very quickly evaluate how big an issue is by looking at the volume of online mentions you’re seeing. For some organizations, 10 mentions a day might imply a crisis, for others, it might be 10,000. The crucial aspect is to compare the spike in mentions to a typical day, or other issues you’re familiar with.

Volume is also important because it can help you work out whether the issue is growing, peaking or disappearing. The rate of change of the number of mentions – i.e., the velocity – is a key factor in helping organizations work out how fundamental or fleeting a particular issue might be.

For example, during the 2012 Summer Olympics in London, the social media monitoring team was monitoring, evaluating and ‘triaging’ over 2.5 million tweets per day. Peaking of issues and velocity of tweets had to be evaluated on a minute-by-minute basis and in the context of the sheer daily volume of social media mentions. Some crises, particularly those that seem to be initiated by traditional media coverage, can peak very quickly and then decline, while others may grow slowly over time.

So, to summarize, these three factors:

  • What is being said - the seriousness of the issue
  • Who is saying it - the influence/authority of the person who posts it
  • How often is it being said - is the issue getting more or less visible?

can be weighted in relative importance to help decide:

  • Who within the organization needs to be informed
  • Who needs to actually do something as a result
  • How quickly the organization needs to do something, and say something

Case Study 3 - United Breaks Guitars

United Breaks Guitars is one of the oldest examples of a social media crisis, but it illustrates the difficulties associated with evaluating social media activity. The crisis appeared to come from almost nowhere - but rapidly spread globally.

In 2009, Canadian Musician Dave Carroll claimed that United Airlines broke his $3,500 guitar while he was flying with them. He didn’t follow their complaints procedures to the letter, so United refused him compensation. He responded by writing a song called ‘United Breaks Guitars’ and uploaded it to YouTube. It very quickly became a viral hit, garnering 150,000 views within a day.

United Airlines was relatively quick to understand the significance of what was happening – interest in the video was growing exponentially. United’s Managing Director of Customer Solutions called Dave Carroll within a couple of days of the song going live to apologize and ask if they could use the video as part of their internal staff training program. But by then, the story had already taken on a life of its own – with five million views of the YouTube video within a month, Dave Carroll v United Airlines is now recognized as a watershed moment in the importance of online reputation monitoring.

Case Study 4 - Marriott Hotels

Marriott Hotels suffered a PR backlash in 2013, when a genuine attempt at a ‘good deed’ went badly wrong. This example illustrates just how quickly unpredictable, employee-led issues can emerge. On September 11, 2013, a staff member at the San Diego Marriott Mission Valley Hotel decided to offer complimentary coffee and mini muffins to guests in honor of those who lost their lives on 9/11.

One guest tweeted an image of the poster questioning the ‘promotion.’ This image was quickly shared on social media as others reacted angrily. The overwhelming sentiment was negative and many accused the hotel of misunderstanding the mood of the nation. It was picked up globally across news media – damaging Marriott’s reputation way beyond San Diego.

Marriott Hotels issued a response to the Huffington Post and other media, which ran the story:

We are aware of the picture that was tweeted. It shows an offer that was made independently by the hotel and not the Marriott Hotels brand. As far as we know, it was limited to one property. While the hotel was making a sympathetic gesture to its guests in remembrance of 9/11, we apologize and understand why some people may have misunderstood the intent of the offer. We are reminding our hotels to use discretion and be sensitive when remembering major events such as 9/11.

This case study again illustrates the importance of spotting potential crises quickly so that they can be dealt with as soon as possible. As Marriott stated, they followed up the issue with rapid internal communication.

3. Escalation

Social media monitoring can help spot potential issues and rank their relative importance, but another critical element of crisis and issues management is to decide exactly who needs to know what is happening and how quickly they need to know. This is called an ‘escalation process’ and organizations structure these very differently.

Let’s revisit the typical methods of scoring issues:

1-3: Social media posts that are complaints, but appear to be one-off or manageable customer service issues.

4-7: Posts that are serious and are likely to have broader implications for the organization, especially if they escalate further. For example: threats of a boycott, accusations of discrimination, regulatory or health and safety issues. These issues require very well tuned ‘antenna’ to foresee the potential dangers – such as the Marriott Hotels case study.

8-10: Posts that are likely to be a crisis, by pretty much any definition. For example, criminal or civil charges, fire, terrorism, death or serious injury, personal data-loss and anything else likely to cause an immediate negative shift in perceptions and/or affect share price.

In some organizations, where the systems and processes are well established, the social media monitoring team is given the authority to ‘skip’ three or four lines of management. They may be empowered to alert the Chief Executive’s office immediately if they believe there is an issue brewing that could materially damage the business.

In most organizations, level 1-3 would be handled by customer service teams, product development teams or sales departments. PR or communications teams would normally expect to be alerted immediately to any level 4-10 issues, and the C-Suite would be made aware of any level 8-10 issues. They will also have processes in place to automatically alert key crisis team staff members to crises and issues, irrespective of how they’re identified.

The crisis team mobilizes and will typically be comprised of senior PR and communications, customer service, HR, legal, operations, technical, product or marketing experts. Dependent on the issue, people best placed to monitor and respond to social media and traditional media commentary will be included (e.g., technical or IT experts for data or security breaches) – the team will remain mobilized until the crisis calms down.

In practical terms, during a crisis it’s very important not to rely on emails. While email groups can be effective for sharing information, they can often be one of the slowest ways to agree action. Best practice in most serious crises would be to physically convene a core team in one location where they make quick decisions and implement them, but failing that, open lines on conference calls can be effective.

4. Response

Should an organization respond? If so, how?

Once the issue has been escalated and the crisis team is convened, they will need to make two immediate decisions based on all of the information on hand:

  • Do we need to respond?
  • Do we need to change anything as a result of what has happened/is happening?

It takes a brave team to sit tight and say or do nothing when there is a significant spike in online criticism, but sometimes, that can be the best course of action. Some situations calm down of their own accord very quickly. In others, the organization may simply need/want to ‘take it on the chin’ and be prepared for criticism.

Regularly checking the velocity of social coverage is one way of assessing whether an issue is likely to go away quickly or is on the rise. Regularly checking the authority of the critics - and their reaction compared to your customers or other stakeholders - is one way of assessing the importance of the criticism. This is particularly true when an issue generates significant negative ‘traditional’ media coverage, which then generates social media criticism.

Case Study 5 - Walmart’s Asda Christmas Mom ad backlash

In December 2012, Walmart’s UK supermarket Asda launched a TV advertisement that showed the nation’s moms preparing for Christmas using a range of Asda/Walmart products, with no visible support from other family members. Almost as soon as the advertisement launched, much of the British media vigorously criticized it. They claimed it was sexist to depict British women carrying out all the Christmas preparations.

Faced with this barrage of media criticism, Asda/Walmart wanted to know what their actual customers thought about the advertisement. Using social media monitoring of their Facebook Page and gauging sentiment on Twitter, Asda/Walmart customers – typically Moms juggling family and domestic life – were equally vocal about the advertisement. They didn’t find it offensive at all. In fact many empathized with the characters and found it funny.

Based on what their existing customers were saying about the advertisement on social media, Asda/Walmart decided to ignore the media criticism, which soon died down. They did issue a statement at the peak of the crisis, which highlighted the fact that they had listened to their customers in light of criticism from the media:

To any mums and dads who have been upset by our Christmas TV ad – we’d like to offer our sincere apologies. It wasn’t our intention to offend anyone. Our ad depicts what many of the 16m mums who shop in Asda tell us they feel. It is intended to be light-hearted and fun and in the main that’s how it’s been received. We respect all hard-working parents and know just how tough it is managing a family – particularly at Christmas.

Asda/Walmart did not change any aspect of their business as a result of this ‘crisis’ and did not pull a costly advertising campaign. They dealt with it calmly, and won plaudits for their approach. In situations that are more clear-cut (e.g., involving loss of life or serious injury), organizations are duty-bound to respond in a timely fashion, to account for what has happened and clarify what they will do about it.

Where should the organization respond?

Once all the facts are known, if the crisis team decides that a statement is required, either on social or traditional media, the organization’s PR and communications teams are likely to take the lead in preparing it and decide how best to distribute it across social and traditional media channels. The best crisis response statements will be created considering:

  1. The facts and the severity of the crisis
  2. The mood of their audience (see the Asda/Walmart example above)
  3. The channels on which the statement will be used
  4. It will be crafted in an appropriate tone of voice and issued on selected channels.

Every situation is unique, but in general:

  • Organizations’ own websites and corporate/CEO blogs are the most ‘formal’ place for a response, particularly if the crisis will play out over several days. Many organizations prepare ‘dark sites’ using a blog platform that can be switched on instantly in the event of a crisis, in some cases to replace the existing home page.
  • YouTube is the perfect channel on which to rebut some concerns about products (particularly on YouTube), or to give the CEO of an organization an opportunity to respond in a personable way that may well get embedded on websites/news channels, to give more context than simply one quote or soundbite.

During the iPhone 6 ‘Bendgate’ issue following the launch of the iPhone 6+, Apple famously invited influential tech blog The Verge to film the stress tests that the devices went through before being launched, and to share them across the social web.

Facebook Notes on Pages can be good places for longer-form responses that can be linked to from everywhere else, particularly for consumer-facing organizations.

How should the organization respond?

Many organizations rehearse any number of crisis scenarios and prepare how they’d respond publicly in each situation. Most crisis responses tend to follow a fairly standard pattern in the first hour or two after an event.

This pattern played out as textbook crisis communications by Virgin Galactic in October/November 2014 when their test flight exploded. Their Twitter stream summed up what happened, who they were thinking about and how they would act, immediately after the crisis occurred. Their statements on Twitter were echoed across their other social channels, on the news section of their website, and then across virgin.com.

Of course, Virgin has a charismatic CEO who visibly and calmly took control at the scene, and with the media. He even managed to get across the ‘vision’ and ‘we will carry on’ messages in his communications, echoing sentiments from the team on the ground. Getting to the stage of issuing a holding statement, or in Virgin’s case, a series of holding statements, and then issuing other statements requires three distinct but interrelated processes.

  • Social media monitoring and external information gathering – what is the outside world saying about what has happened
  • Internal information-gathering and response creation – what do we know, what are the gaps and how can we best represent our knowledge of the situation
  • Approval and distribution of the response – who ‘signs off’ what we’re saying, and how do we ensure our responses are visible

This model has been formalized by social media crisis communications experts Restless Communications.

The monitoring team’s updates provide context and should include briefings on:

  • Changes in velocity, volume, sentiment and tone of comments and coverage
  • Any new ‘operational’ developments as reported online
  • Any comments from specific stakeholders, e.g., crash investigators, police or fire chiefs, politicians, union representatives or families of those involved
  • Relative volumes of conversations tagged with relevant hashtags

These updates should be ‘played in’ to the crisis team, along with all of the internal information available, so that the central team knows as much as possible about what is happening.

How often should an organization respond to people during/after a crisis?

There are no hard and fast rules about how often you should respond, as each crisis has a different DNA. Each organization has its own policy about responding to people on social media and different resources available to do so. But it is absolutely important to keep monitoring, to be able to:

1. Respond directly to anyone directly affected by the crisis immediately.

Social media monitoring teams should be briefed to monitor relation keywords (e.g., brother, sister, mom, college friend, colleague, etc.) to protect people affected from providing sensitive information in a time of crisis. They should try to take conversations offline by asking people to Direct Message or Private Message (DM or PM) through social channels.

2. Respond generically to others who are interested as resources allow.

This depends on what team is available, and based on strict ‘scripts’ approved centrally. The response would typically point people to where they can find more ‘long-form’ information – e.g., on the relevant corporate web page or YouTube.

3. Post new content on social channels when there is something new to say.

For example, linking to a statement that has been posted online or to live tweet a press conference. Organizations should evaluate and quash any rumors that the social media team spots that are untrue and also let people know when to expect another statement.

How can an organization tell when a crisis has passed?

Many crises disappear almost immediately, but others are likely to have a ‘long tail’ and will affect reputation and visibility through online media, search and social media for a considerable time. The best way to understand where an organization is on the return to normality is to establish effective monitoring mechanisms that identify and report back velocity, sentiment and authority of social commentary.

If velocity is going down, sentiment is improving, and authority is declining, then the crisis is probably passing. But beware, it could return very quickly – particularly if influencers comment, the tone of any public response misjudges the mood, or a perception takes hold that the organization is not doing enough to address the issue. One of the best ways to test the mood after a crisis is to publish a series of ‘vanilla’ statements which, under ‘steady state’ operation, would not generate significant responses or sentiment. If people are critical of them, then it’s too soon to return to normal service on social channels.

Summary

Reputational issues and crises tend to be uncomfortable, unique and unpredictable. That is evidenced through the case studies introduced here, experiencing them ourselves as consumers and by looking at news media and social media daily.

As our use of social media evolves and fragments, organizations know that they will have to expect and deal with issues in different ways on a regular basis - and that it is critical to prepare for the very bad and the very worst events. Regardless of your business sector or the size of your organization, doing the following will make a considerable difference to how well you deal with issues and crises:

  • Choose the right listening tools and set them up properly
  • Be prepared - understanding your smoking, smoldering issues and create crisis scenarios and responses
  • Create issues/crisis management processes
  • Rehearse scenarios
  • Train your staff
  • Consider using specialist support to prepare, identify and deal with the most sensitive situations

History tells us that organizations that recover from crises with their reputations most intact are the ones who are most prepared. We hope that this paper has helped you identify where you should focus any improvements.

Want more like this?

Want more like this?

Insight delivered to your inbox

Keep up to date with our free email. Hand picked whitepapers and posts from our blog, as well as exclusive videos and webinar invitations keep our Users one step ahead.

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy

side image splash

By clicking 'SIGN UP', you agree to our Terms of Use and Privacy Policy