Prison Break: The Truth About Crisis Communications in a Social World

October 13, 2015

[mark lenniham/ap/corbis]
[mark lenniham/ap/corbis]

Ferguson. Baltimore. New York City. Across the nation, public sentiment has piled up against law enforcement.   

But not in upstate New York’s North Country, where signs, banners and blue ribbons dot the landscape, proclaiming residents’ gratitude. A local auto dealership planned a law-enforcement party. A bowling alley offered troopers free games.

All of this in the wake of a 23-day manhunt for two convicts who escaped from the high-security Clinton Correctional Facility in Dannemora, N.Y.

In the public’s eyes, authorities obviously did some things right: The manhunt ended in the death of convict Richard Matt and the shooting and eventual capture of the other, David Sweat. But there was also room for improvement, especially on the social media front.

Public information officers and social media

When I was the director of public relations for the State University of New York at Plattsburgh, I decided to test our social media prowess during a full-scale emergency drill.

Partnering with Homeland Security, local fire and police crews and other emergency officials, we simulated what would happen if a bomb went off in one of our labs. Students in makeup played the dead and injured. Others portrayed reporters and parents.

Because public information officers (PIOs) from other local agencies wanted to experience the drill, I invited them to help provide support for our communications needs. I created private groups on Facebook, one designed to simulate our main Facebook page, and the other our Twitter feed. (I didn’t want to use the real social media feeds, lest someone mistake our posts for a real emergency.) Then I invited our students to test us by posting to these sites.

The result? Even with the extra supports in place, we were not able to focus enough on social media. Our posts were far from optimized. Comments came in so fast that the old computers in our emergency operations center couldn’t refresh quickly enough to keep up with them. We also discovered broad gaps in our social media policy — loopholes we have since closed.

I mention this because I believe at most agencies, public information offices are staffed with, at most, one or two people assigned to handle social media, and handling those feeds is often only a small part of their job. As a result, when the media and the public descend with demands that eat time, Facebook, Twitter and other platforms get neglected. But if we don’t handle social media, others will — perhaps to the detriment of our organizations and public safety.

After all, the number of Americans getting news from Twitter and Facebook continues to rise. A recent Pew Research Center survey found that 63 percent of Twitter users and 63 percent of Facebook users get news from these platforms — and those numbers have gone up more than 10 percent in just the past few years.

Knowing this, PIOs — especially those who deal in the realm of public health and safety — need to change their priorities and find ways to move social media to the fore.

The tweets

At 9:04 a.m. on June 6, the police tweeted that they had received a call from the Clinton Correctional Facility, with a link to the full message on the police's Facebook page.

Twenty-nine people did what the State Police asked and dutifully retweeted the post. However, the Facebook link in it didn’t work.

Perhaps there was a typo in the link, or maybe the original Facebook post was taken down for editing. Either way, the tweet itself should have been removed so as not to promote an ineffective link.

Twenty minutes later, a second tweet went up, with the same information.

In this tweet, the link works, but there are still problems.

Readers cannot immediately tell what this (or the previous post, for that matter) is about. The note about sharing, and the fact that it comes from the New York State Police are the only indicators of its importance.

Tweeter Paul Schreiber advised in his response: “@nyspolice @NYGovCuomo don’t bury the lede! Put key details in the tweet.”

Schreiber was right. In our posts, especially during times of emergency, we should place critical information front and center; people should never have to hunt for it.

The Facebook link

When readers clicked on the working link, they landed on the New York State Police's Facebook post, with descriptions and photos of the escaped inmates.

The police repurposed the press release from their website, which was smart. Doing so saves time, when time is of the essence. However, some tweaks could optimize this post:

1. Start with the lede. The public should not have to read through lines of text to get to the part where escaped convicts are on the loose. The Facebook post and the press release should — like the tweet — get to the point quickly. Tell us what’s important. Fill in the details later.

2. Use the headline. The news release on the website uses the headline “State Police Search for Escaped Prisoners from Clinton Correctional Facility.” This drives home the key points and could have been used in the social media posts.

3. Pay attention to readability. Research has shown that scanning plays an important role in online reading. Readers seldom tolerate large blocks of text. This is why online posts tend to use a lot of headlines, subheads and bullet points.

4. Cut and paste. Finally, most of these goals could have been achieved by simply cutting and pasting the headline into a Facebook status, along with a link to the online press release.

Authoritative information sources

When you need to post emergency content, where should it reside?

Twitter doesn’t allow for enough information, so for most agencies, that generally leaves either a website or Facebook.

In a case like this, where false information could cost a life, the PIO needs to control the information as much as possible.

In general, anyone can add comments on Facebook. This makes the platform a great place for conversation and feedback. But it also makes it a perfect place for misinformation that can draw attention away from important messages.

Posting the main article directly on your website and linking to it ensures that you will have one undiluted, authoritative source for information. Furthermore, since you have only linked to it and not copied the entire article, should details change, you will only need to make edits to one site.

Use Twitter and Facebook to push out small but important bits of information, drawing people through links to the authoritative site. Then monitor. Here’s why:

Day-one responses

The Facebook post had 2,687 responses. Many of these contained political rantings. Some were “go get ’em” messages. Others let the police know that they had shared the post with friends.

But embedded in these were thoughts that the police may have wanted to consider:

1. There was a post about possibly searching for the men in the seasonal camps in North Country, which are often used for hunting. (The escapees did, in fact, hide out in some of these camps.) Now, police had likely already thought of searching these, but you never know when someone might have an idea that you missed.

2. Combining the thoughts about the camps with the numerous sentiments about guns, police might have been led to consider putting out a statement asking people to make sure their guns were secured. (The men ended up taking weapons from cabins.)

3. Misinformation was already being shared. One comment said that it was confirmed that the police were in a high-speed chase with the fugitives. Another said that the men were driving a blue Hyundai. These posts sat uncorrected on the police Facebook page.

4. Finally, the combination of this rumor with a post elsewhere about the jumpiness of the population may have led to other worries, especially for people driving a blue Hyundai.

Media rumors

A few days later, on June 10, an even more problematic rumor evolved — more problematic because it made its way into the mainstream media.

A sighting of two unidentified men resulted in the manhunt moving to the small town of Willsboro, N.Y. Soon, Facebook and Twitter were filled with reports that police had the escapees surrounded. The New York Post mentioned this online, and the New York Daily News posted a story. A local news station, WPTZ, tweeted about it. Then a radio station picked it up. Even the Nigeria News Feed was posting the story.

The same thing happened on June 26, when officers shot Richard Matt. Twitter was flooded with false reports of a shootout between police and the other fugitive, David Sweat. Once again, these rumors found their way into the mainstream media.

These traditional news channels would have done well to be more cautious with their posts, double-checking their sources. But their efforts were, undoubtedly, made much more difficult by the crush of communication hindering access to the state police’s PIO. Every reporter I have discussed #PrisonBreakNY with said they were only able to get to voicemail, with an inbox that was so full they couldn’t leave a message. One reported having some luck by email, but others did not.

When hundreds of reporters converge upon a communications department that was never set up to handle this kind of flow, bottlenecks can be expected. But something has to be done to relieve this pressure and help those reporters have quick access to accurate information.

Social media could have been the solution. Two simple posts could have cleared up much of the confusion: “Prison escapees remain at large. #PrisonBreakNY” on June 10 and “Sweat remains at large. #PrisonBreakNY” on June 26.

All told, during the 23-day search, the New York State Police posted 33 tweets related to the prison break. June 10 had the most  — the day of the Willsboro rumors. However, not one of those posts addressed the rumors. The police posted twice on June 26, the day Matt was shot and the day of the Sweat rumors, but neither post addressed these events.

Furthermore, in their 33 tweets, the police did not use a single hashtag, and nearly 40 percent did not contain any keywords. People searching for news on the prison break would likely not find anything unless they knew to look directly at the @NYSPolice feed.

Had the police been monitoring, known that the hashtag #PrisonBreakNY was trending and updated the status of the search using it, they might have quickly quelled rumors and helped reporters.

Virtual operations support teams

So how can an agency effectively manage all of this when chaos strikes and resources are in short supply? They can do it with a little help from virtual operations support teams (VOST) — groups of trained citizen volunteers or, as VOST founders call them, “trusted agents.”

Often activated by emergency management officials during fires out west, these teams monitor a situation by looking for public sentiment, misinformation or confusion, and then refer issues back to the PIO. Or they may be asked to amplify communication efforts or help the officer spread official news to the public.

VOST founders molded their system on the Incident Command System (ICS), a hierarchical format the government designed to help various organizations work together in an emergency. The system emphasizes the use of common terms, structures and reporting systems so as to cut down on confusion among agencies. VOST adopts those structures, working through the PIO and using a version of ICS forms. In one recent fire, emergency officials said that VOST reduced calls to emergency centers by 70 percent.

In the prison break response, VOST could have helped the police see inefficiencies in the social media response early on, while the situation was still unfolding. A team could have found troublesome tweets and encouraged action.

More information on VOST, including a list of current teams, is available on

Calm has returned to upstate New York. Road blocks and searches through forests and swamps have ended, and North Country counts itself grateful to law enforcement.

But one can envision an ending that might have gone differently, especially with chaos reigning in the social spheres.

And those of us who are charged with helping to ensure public safety through our communications must take note of those alternate endings and find ways to tame our portions of this new virtual world.

Michelle Marasch Ouellette
Michelle Marasch Ouellette is an assistant professor of public relations at the State University of New York College at Plattsburgh. Email: Twitter: @M_2_Oh.


Mary Dorsey says:

Very timeline information as we were just strategizing this yesterday! thanks

Oct. 14, 2015

Michelle Ouellette says:

Thanks, Mary. Glad you found it useful.

Oct. 19, 2015

Sue Spissinger says:

Well done and terrific analysis Michelle!!!

Oct. 19, 2015

Kathy Ehlers says:

As someone who writes the emergency plans for my school district, this article is very helpful as we review and update our plans. The crisis communication area needs to be looked into deeper. Thanks!

Oct. 20, 2015

Arsenio Hall Franklin says:

There is a lot of talk about how to use social media, but I'm sure not many companies test it during drills or hold it to the same standards as their other communication channels. I'll remember this.

Nov. 28, 2015

Carla Beecher says:

Nice article, Michelle.

Dec. 10, 2015

Post a Comment

Editor’s Note: Please limit your comments to the specific post. We reserve the right to omit any response that is not related to the article or that may be considered objectionable.


To help us ensure that you are a real human, please type the total number of circles that appear in the following images in the box below.

(image of four circles) + (image of three circles) + (image of three circles) =



Digital Edition