Why External Variables Matter in Incident Management

When it comes to incident management, it’s easy to fall into an insular mindset. We spend months planning and configuring systems that alert us of any issues within the system, and to cover our bases, we establish traditional customer support channels to identify issues we don’t catch ourselves. While this train of thought isn’t wrong, this approach has led to the rise of users reporting issues they are experiencing in public forums like Twitter and Facebook.

Social Media Paving a New Path

Social media has become a great way for organizations to connect directly with their users in a casual setting, which has opened up the door for closer two-way communication. Beyond simply being a more personal way to communicate with an organization, social media has also become a more effective way to get help.

Because of the public nature of the medium, users have found that when they aren’t able to get help via traditional routes, such as a phone call or support ticket, sending a frustrated tweet often can yield much quicker results. I can personally attest to the effectiveness of using social media when all other avenues have failed. It’s not just about fielding complaints, though. Many users prefer to use social media to communicate with organizations because it makes them feel like their input is heard.

I’ve personally shot a few courtesy tweets to companies when something looks wrong.

• Is their site throwing 500 errors? Send a tweet.
• Is there a wonky JavaScript interaction that makes the experience less than stellar? Send a tweet.

Sure, it’s probable that they are aware of the issue, but as a developer, I know that I appreciate receiving a heads-up on problems I might not be aware of. “If you see something, say something,” applies perfectly to this type of situation.

Synchronous vs. Asynchronous

While the “reporting” aspect of social media support is incredibly powerful, there’s more to it than ease of use. Social media platforms — Twitter in particular — have a “real-time” feel to them.

Email, for example, is an asynchronous communication medium which, in the context of customer support, means that your issues are added to a queue. There’s no sense of urgency, and the non-disruptive experience can leave you feeling like you’re waiting in line at the DMV. Twitter, on the other hand, has a more synchronous flow to it. While it’s not exactly a real-time support channel like a phone call or live chat, there’s more of a sense of urgency associated with it than email because the company’s time to response is recorded publicly. Feedback is more immediate and personable, which makes your issue feel just as important as anyone else’s.

Separating the Needles from the Haystack

So how do we stay on top of external variables without getting buried in the non-incident management-related input? At a high level, the solution to this problem is the same as any support channel: triage and delegate.

Triage

Because social media isn’t intended to be used solely for customer service, accounts can get jammed with non-support chatter. It is important to physically isolate the bug reports from the rest of the feed in order to effectively respond to them. Beyond simply isolating the bugs from the non-bugs, identifying commonalities between the reports is also key. This will allow you and your team to detect patterns and escalate issues that affect multiple users before they get out of hand. The Operations Command Console, for instance, can correlate data sources such as tweets to a specific event in your infrastructure as well as visualize the blast radius. This way, you can understand if a failed deploy or outage directly led to a customer reaction on social, and if so, the extent of the customer impact.

Delegate

Once an issue has been identified and isolated from the rest of the group, it needs to be assigned to the appropriate person or team. This can be accomplished in a myriad of ways, either by simply forwarding the report to your existing help desk application, or using a more dedicated social media support application.

No longer solely owned by marketing departments, social media platforms are an excellent way to get real-time information about trouble within your application infrastructure, and an absolutely essential source of data in today’s digital world to provide complete visibility. By giving these platforms the same level of process and dedication as more traditional support channels, you can respond to incidents as the initial reports come in and maximize customer happiness, rather than after a problem has grown into something more damaging.

While the call center isn’t necessarily dead, in order to ensure happy customers, organizations must be proactive instead of reactive in leveraging the wealth of customer data that is now available through a public medium such as social. Those who include social media as a critical component of their monitoring and incident management strategy will have deeper visibility into the quality of their users’ experience and reap the benefits of improving customer loyalty.