A couple of months ago a client asked us how his front line customer service agents would be able to spot an issue or a crisis brewing on social media so they could escalate to the comms team appropriately.
Not too difficult I thought, we’ve done similar things in the past to help agents identify potential wildfires and we’ve run a whole bunch of training sessions on how to handle social media issues. We’ve even run simulations to help organisations practice their crisis process and response on social media.
So we created a simple check-list for agents to sit alongside their standard response triage and handling processes. This checklist asked agents to look for the following when they saw a negative comment of query:
- Is the customer a celebrity and/or journalist?
- Does the customer have more than 50,000 followers?
- Does their bio or social profile indicate that they are influential in your sector?
- Is the issue something unexpected that would likely affect more than 50 people?
- Are more than 5 people suddenly talking about the same issue in a 2 hour period?
- Has the message received more than 20 retweets in the last hour?
- Has the customer paid to promote their complaint
- Is the customer threatening the organisation or its staff?
- Is the issue alerting you to a non-BAU situation?
- Does it relate to a ‘hot issue’ that you have been briefed on by the comms team?
- Does it make your organisation look unethical?
- Does it ring alarm bells for any other reason?
This allowed the agents to quickly escalate any worrying posts to the comms team for review.
Then the client threw the curve ball. He asked us to provide a crisis assessment model to guide the comms and customer teams in evaluating the impact of the issue quickly and efficiently to guide the level and tone of response.
Interesting! I had certainly never seen an assessment model for crises and I the more I thought about it the more I recognised that this was a huge gap for businesses, as effective customer management in a crisis depends entirely on a joined up response from comms and customer teams.
So we built one.
We started by thinking about what constitutes a reputational issue or crisis. What characteristics do crises have?
Borrowing from Situational Theory, we came up with three core characteristics that we called ‘Red Flag’ characteristics.
- Impact – Impact is based on severity, volume to complexity, e.g. those issues where the impact on the customer is severe, where the issue is hard to resolve or where large numbers or people are affected).
- Reach – Reach is based on the level of dissemination of the issue, e.g. where it has been picked up by a celebrity or influencer or has a large amount of groundswell or public pressure behind it
- Depth – Depth is more nebulous but probably the most important characteristic of a crisis in terms of a severity indicator. We defined depth characteristics as those that exposed the organisation negatively, e.g. where the issue indicates some sort of systemic failure or wrongdoing in the business or implicates senior staff
Of course, not every issue is a crisis. An event can impact many people but may not be serious or show any sort of failure in the organisation. Alternatively, an influencer may be talking about your brand but it might not be any cause for concern.
For this reason we decided that as a rule of thumb a single red flag was most likely to be a simple PR issue but if an event had 2 or more red flags it was likely to represent a more complex issue, with a greater potential to become a crisis.
For example, if you have an issue that is impacting a lot of customers AND an industry influencer is talking about it, you need to manage it much more closely. Or, if you have a systems outage that you should have prevented AND Stephen Fry is talking about it, you probably need to do some damage limitation work.
So if you have a 2+ red flag event on your hands, how do you assess it and begin to manage it?
We approached this by creating a balanced scorecard for each characteristic.
Here we set definitions, relevant for the client that allowed us to give the event a score (ranging from 1 (incidental) to 5 (severe)) for impact and reach.
Plotting these scores allowed us to get a sense of the reputational risk posed by the event and its potential impact on the organisation if not handled correctly.
Example balanced scorecard for Impact (generic organisation)
Depth we treated slightly differently. Here we rated the event as either Business as Usual (BAU) (implying no organisational fault), Isolated (preventable but not malicious) or Systemic (attributable to organisational mismanagement or bad practice).
Depth thus became a weighting mechanism in the assessment. If two issues have the same impact and reach scores but one is systemic and one is BAU, the systemic issue has far more potential to damage reputation and also affect the bottom line of the business. It therefore requires a more robust and thorough crisis response from the organisation.
Moreover, the depth score provides a steer on the crisis management strategy and tone – BAU events are likely to be non-fault events and can be managed pragmatically by providing solutions and fixes. Systemic events will require much more contrition and evidence of genuine internal change in the organisation not only to resolve the event but also ensure it cannot happen again.
So how does this model work in practice? Pretty well I think.
Lets take three famous social media crises.
- Burger King – Twitter account hacked and altered to say that they had been bought out by McDonalds.
- HSBC – Suffered a major systems outage in the UK that took a few days to fix
- Starbucks – A social media campaign that Starbucks ran was hijacked by users protesting about their alleged failure to pay tax
- First Great Western – When the UK was hit by massive floods, the rail company had to manage huge disruption on social media
Using our model this is how we would score these:
These map to our scorecard as:
On this basis, we can make an assessment on how we would have ideally handled these crises and how the brands actually did.
Burger King – an embarrassing issue with quite a large reach but with low impact. The fact that Burger King were the victim in the event means that a correction / explanation would probably suffice. We would also advise some review of security on the account though!
HSBC – moderate impact and a very wide reach, this crisis was also very complex and, whilst not systemic, it was preventable, highlighting some failings in the organisation in terms of technical infrastructure. This crisis should have been prioritized and given significant management resource to keep the public informed. In reality this issue was well mitigated by HSBC’s social media response (although this was based on a crisis plan which was developed after a previous outage which was not handled well)
Starbucks – Ouch! This issue should have really rung alarm bells and Starbucks ideally should have rolled out the big guns internally to show how they were going to effect change as a result of this. This issue has the potential to flare again and will continue to impact their marketing efforts
First Great Western – Well this was truly severe impact and severe reach. However, this was ultimately a BAU issue – the weather was to blame here not FGW. It is easy to see how public sentiment could turn though, so information is king in handling this crisis – timely, relevant and supportive news on events can mitigate further impact. In reality, this is exactly what FGW did and as a result, they provided a great level of service in what were very difficult circumstances, not only mitigating a reputational crisis but also reducing the impact and the reach of the event!
I hope our crisis model will be useful for your planning. We’d love to hear your feedback on this. If you want to discuss any elements of our approach or see how this could be adapted for your organisation, please give me a call on 07824 557814 or email firstname.lastname@example.org
Image credit: Brian Moran