To Follow Up or Not? What’s Best Practice in Emergency Alerts?

Here’s a story about a missing person incident in Needham, MA.  It seems that an autistic young woman went missing and two emergency alert calls were sent to the entire town (one by mistake.)  The story says that some residents were upset that they didn’t get a follow up call to tell them the woman was found and returned home safely six hours later.

It seems that Needham uses an emergency alert provider that charges them every time they use the system.  So a follow up call would cost the town hundreds of dollars, according to our information.  And that mistaken second call probably cost hundreds as well.  (They should switch to one of the unlimited use plans Hyper-Reach offers and both save money and avoid this issue.)   

But regardless of cost, what’s the right protocol for follow up information for alerts?  We don’t think there’s one right answer, but here are some thoughts:

  1. More information is usually better than less.  In general, a follow up call is a good idea.
  2. Context is important.  Needham seems like a close-knit community.  In contrast, we’ve seen anecdotal evidence that “missing person” alert calls in big cities are sometimes unwelcome.  And if the call is unwelcome, a follow up call won’t be welcome either.
  3. Time of day is important.  We wouldn’t recommend calling after 9PM for a non-urgent call.  (The woman was found around 10PM.)
  4. There are other ways to deliver follow up information.  A news release could let at least some of the community know.

Hard research is lacking in this area, however.  And it’s something we think is needed.  In the meantime, good judgment is the best alternative.  And it’s best if that judgment isn’t affected by unnecessary financial constraints.

Hazards of Misfires with Wireless Emergency Alerts

This story from illustrates the danger of mistakes in the use of IPAWS and the Wireless Emergency Alert system.

In late September, the LAPD sent out a WEA alert warning about about a dock fire at the port of LA (see photo from Reddit).  But the WEA alert went out to residents as far as 40 miles away.

To make matters worse – according to the article – the alert went out at around midnight, three hours after the fire was mostly contained.  Which means it wasn’t relevant, wasn’t useful and was seriously irritating to many of the people who received it.

We have a number of concerns and some thoughts about how to repair the damage:

  1. Mistakes like this are going to cause people to turn off WEA alerts on their phone.  It’s a simple process, if you know how, to turn off all alerts except those from the President.
  2. Confusion with other alert systems will result in resistance to enrolling in those.  If you read a story like this, how likely are you to want to sign up for other public safety alerts?
  3. The resulting PR blow back may inhibit future emergency alerts from being sent.  Which could endanger the public further.

Does it have to be this way?  Here are some thoughts about making things more effective:

  1. Avoid the cryptic messaging.  Instead of “in this area”, use “within x miles of the Port of LA”.  Replace “Take Shelter Now” with “stay indoors to avoid toxic fumes”.
  2. Use two messages.  WEA messaging is limited to 90 characters, which helps explain the cryptic text.  But authorities can send multiple messages that provide the needed detail.  And we’d bet that two messages with meaningful information would be better received than one message that’s indecipherable.
  3. Devise “best practices” for WEA messaging in advance.  Writing a message on the fly at midnight risks an ill-conceived text.
  4. Use a good interface.  The LAist article suggests that the group selected was chosen deliberately.  Maybe…  But we’d bet that a good interface that made it easy to select a smaller area might have helped limit who got woken up at midnight.

We’re glad that local officials are using the WEA system for alerting, since it gets around the problem that emergency alert systems are confronting due to lack of wireline listings to call.  But it’s important to get this right.


LA Fire mapLAPD WEA fire warning


Are You Ready for Zombie Preparedness Month?

Throughout the year, states across the country organize disaster preparedness campaigns to educate residents about best practices before, during, and after a natural disaster. Gulf states try to get citizens ready for hurricane season; northeastern states prep their residents for severe winter storms; west coast folks talk up preparation for wildfires and earthquakes; and the Midwest plans for tornadoes.  Etc.

Nothing wrong with these efforts, but perhaps a new angle could get a little more attention.

In a bold stroke, Kansas Governor Sam Brownback has recently declared that October is “Zombie Preparedness Month in Kansas.”  The idea is to find a creative approach to grabbing residents’ attention: preparing for the zombie apocalypse is not just for ‘preppers’ any more.  Which means that zombie video game buffs and ‘Walking Dead’ enthusiasts can now apply their survival skills to real life preparation!

State and local authorities hope this initiative will encourage people to actually think about disasters and get them to translate from a flesh eating attack to making it through a tornado strike.

But there’s one major problem with this campaign, as any real prepper could tell the governor.  While the proclamation calls for Kansas residents to have a three-day stockpile of emergency supplies, real zombie fanatics know that a zombie apocalypse would not be a short term disaster. ‘Preppers’ want to be ready for years of isolation and anarchy, while most weather-related disaster plans call for just a few short days of ‘surviving’ on their own.

Still, a few days’ survival is better than none, and this campaign has the advantage of appealing to a younger demographic than traditional campaigns would.

An idea like this is too good not to steal and we hope that other states pick up on it as well.  And why stop with zombies?  With Halloween just a few weeks away, preparing for monsters of all sorts is certainly warranted!

The good news is that everyone, regardless of their disaster risk, has the opportunity to register for emergency notifications at to receive up-to-date local alerts.

The Future of Social Media and What that Means for Emergency Alerts

This article is a list of predictions from “experts” on where social media will be in 25 years.

While we agree with a lot of the comments that this list is superficial, it’s a interesting prompt to think about how the use of social media will change for emergency managers. 

 by [2039], use of social media will be ubiquitous and integrated into our daily lives in a multitude of ways. It is expected social media will be integrated into wearables that track our habits, and virtual experiences will be part of the package. The challenge will be coping with the massive amounts of data that will deluge the masses. [emphasis added].

Here’s a prediction that we’ll hazard to make.  Emergency alerts will be much more effective once emergency alert systems get “connectors” signed up and enable them to easily re-post alert information.

Our thinking goes a bit like this:  most folks are not attuned to emergency warnings, but some people are.  And among the people who are, some of those are the social connectors that Malcolm Gladwell describes in his book The Tipping Point.  These are the same people who are the “active nodes” that were used in the study we referenced a few days ago.  They’re the folks that spread the word among their network of contacts.  So getting an alert to these people is especially helpful for making sure that the news is disseminated quickly and thoroughly.

And – once an alert is out – interest (assuming that the issue is sufficiently compelling to enough people) will build on its own.  Which leads to people following specific hashtags and folks who make themselves leaders on an issue.

An emergency alert is not necessary for this process to get started.  Remember  Michelle Sollicito’s “SnowedOutAtlanta” Facebook page during Winter Storm “Leon” in January?  She got over 50,000 followers in very short order.  Ms. Sollicito didn’t need an emergency alert to tell her it was snowing.  But we’d bet that getting the Michelle Sollicitos of this world signed up for alerts would make them much more effective – especially for events like tornado warnings, wildfires and hurricanes.

Relying on social media alone to spread emergency alert information seems very haphazard.  We know many people who check their Facebook and Twitter feeds only  intermittently.  And many emergency events require quick response.  (One study we’ve seen says a 15 minute warning can reduce tornado casualties by up to 45%.)

But while social media alone may not be effective, imagine a future that grabs the attention of the most prolific and effective communicators and enables them to get the word out ASAP.

Emergency Management’s Tips on Getting Folks Signed Up for Emergency Alerts

We like Emergency Management magazine, and we were glad to see this article on getting the public signed up for emergency alerts.

And we agree with the suggestions in the article.  For example, the suggestion of offering incentives parallels our experience at the Burke County fair, which our unscientific assessment suggests at least tripled our success.

We also love the suggestion to use an opt-out process, rather than opt-in, but wonder how often it’s a real option.  You need a list of mobile numbers and addresses to make that work.  Public utilities could be a good source, if you can get their cooperation.  We’ve talked to one agency that had access to their water utility’s records and used those on an opt-out basis.

The other suggestions:

  1. Base the messaging on empowerment, rather than fear;
  2. Share success stories about how alerts have helped people;
  3. Get a trusted partner to help out.  For example, using a senior agency to reach the elderly.

These are all great ideas.  Let us know if you have more.


Can Wireless Emergency Amber Alerts Be Made More Effective?

Wireless Emergency Alerts are the alerts that use the IPAWS system and are broadcast to all newer mobile phones in a target area.  They’re intended for imminent emergencies with a high probability of loss of life or property, but they are also specifically enabled for Amber Alerts, which notify the public of an abducted child.

The concern we have with WEA-based Amber Alerts is how cryptic they are.  Here’s a recent one:

Emergency alert Sampson County, NC AMBER Alert: LIC/KHX-728 (SC) 2001 Brown Buick Century Type:Amber Alert

Last year, California residents experienced this kind of frustration deciphering an AMBER Alert WEA message.  As reported by the LA Times, residents across the state received a confusing alert in the middle of the night, containing only the ambiguous message: Boulevard, CA AMBER Alert. UPDATE: LIC/6WCU986 (CA) Blue Nissan Versa 4 door.

We asked a few friends and none of them were 100% sure what an AMBER Alert was exactly.  Most thought it was about a missing child, rather than an abducted one.  So a cryptic description of the vehicle in which the child is suspected of being abducted may not be enough information.

WEA messages are limited to 90 characters and so are limited in what they can say.  But – as a Sprint engineer once pointed out to us – there’s nothing that prevents sending a second message.  And since the message goes out only to a specified geographical area, we wonder if the text would be re-written.

For example this re-write uses one less character and doesn’t depend on people understanding what an Amber Alert is:

Emergency alert Abducted child: Report this car: 2001 Brown Buick Century SC Lic/KHX-728 Type:Amber Alert

And here’s a version that uses a 2-part message:

  1. Emergency alert Child abducted in Sampson County, NC. Look for older Brown Buick Century. Type:Amber Alert
  2. Emergency alert cont’d: SC lic plate: KHX-728  If seen, call 911. Type:Amber Alert

And another, using information about the child:

  1. Emergency alert Child abducted in Sampson County, NC. Boy, age 2, white (Jeremiah Hughes) Type:Amber Alert
  2. Emergency alert Call 911 if you see older Brown Buick Century. SC lic plate: KHX-728   Type:Amber Alert

We have no idea which of these is more effective or if there’s a better re-write possible.  But as professional communicators, it seems to us that there should be research to determine best practices in this area.

Can Phone Call/Text Traffic Help Identify Emergencies?

This article describes an Israeli study which tried to find the existence of emergencies based on spikes in telephone and text traffic.  The idea goes like this:

  • Identify high traffic nodes, essentially people who make lots of calls and text messaging (the data was all anonymized to avoid identifying actual people);
  • Take non-random samples of call and text traffic at these high traffic nodes;
  • Spikes in traffic, amplified by a fancy algorithm, suggest the presence of a big event.

To quote the article:

The researchers applied an algorithm they call the Social Amplifier to this nonrandom sample to tease out where and when emergency events could be found. Those uncovered included eight real emergencies—three storms, a bombing, an earthquake, a blackout, and two airplane-related events—as well as eight other big local happenings, such as concerts and festivals, that were not emergencies.

Can such an approach be done here in the US?  We wonder if the public would believe that the use of cell phone data – especially data based on identifying heavy callers – was really anonymous.  And since the approach appears to identify many events with big social impact – not just emergencies – it’s really only valuable as a supplemental data source.

Still, in a world that is being rapidly shaped by the use of “big data”, it’s an interesting and potentially useful idea.

Why So Few Downloads for Emergency Alert Apps?

Hyper-Reach offers an emergency alert app for smartphones, as do several of our competitors. We think ours is pretty cool and has some neat features you won’t see on other apps.

But we need to be humble.  As does everyone who’s created an emergency alert app for smartphones.  Because the simple truth is that these things are not getting much attention from the public.  As best we can tell, none of the apps provided by any mass emergency notification provider have reached even 1% of their target audience.

When we search for emergency alert apps on the iPhone and Google Play stores, a few potential reasons become obvious:

  1. Fragmentation.  Search for “emergency alerts” and hundreds of apps come up on the Google Play site with no good way to filter through them.  They include apps for specific places (Alaska, Alberta, etc.) and apps that are to sound the alert instead of receiving one.  For many folks, that’s enough to stop a search altogether.
  2. Confusion.  Hand-in-hand with fragmentation is confusion.  Why install an app for emergency weather alerts if I can get weather alerts for free from another, non-emergency app?  Where’s the value?  It’s just not clear.
  3. More confusion.  One of our biggest competitors shares its name with a lot of unrelated apps.  Type “codered” into Google Play and you’ll see apps to “insert code into your brain” and track female menstrual cycles, among other things.
  4. Poor reviews.  Some of the apps we consider most relevant to mass emergency notification have pretty weak ratings.  3+ out of 5 is not strong and will turn off many potential users. (Our rating is 4+, BTW.)
  5. Cost.  Some apps charge money, while others are free.  Why pay for weather alerts if I can get them free from or my local news station?

We’ve got ideas about how to fix this, or at least make the situation better.  But it’s going to take work.  And we’d love to share your thoughts.

What Role for Emergency Alerts with Ebola?

Now that we have the first case of a patient being diagnosed with Ebola in the US, it’s worth asking whether emergency alerts can play a role in helping deal with this disease, either in Dallas, where this particular patient is located, or in some other city, which could have another case.

Dallas is special because one of the two Americans flown back from Liberia with Ebola was from Dallas, so news reports suggest that there’s already a high awareness of the disease in that area.

It seem to us that emergency alerts can potentially help in a few different ways:

  1. Calm fears with education.  An epidemic of Ebola is highly unlikely in the US.   But that doesn’t mean the public understands this.  One poll found that one in four Americans are worried they or a family member will get the disease.  Alerts could help direct folks to information that could allay their concerns.
  2. Direct citizens to health care.  Getting someone with Ebola into the health care system is critical in stopping its spread.  This is one of the reasons the disease has gotten out of hand in African countries, where people have avoided health care workers.  Telling folks to report their symptoms – especially if they’ve been exposed to someone who’s been to Africa – might be helpful in getting them into the health care system sooner.
  3. Mobilize public safety and health care workers.  One of the surprising things about the Dallas case is that the patient went to the hospital and was sent home and was not then diagnosed until going back to the hospital a few days later.  It turns out that the nurse asked the patient, but somehow the information didn’t get to whoever made the discharge decision.  Asking someone with a fever if they’ve been to Africa within the past month or exposed to someone who has seems like a pretty simple question and one that should focus attention well.  Reminding staff to pass the information makes sense. And since the incidence level of the disease is going to be very small, reminders using the alert system could be helpful in making sure that question gets asked and the information shared.  Many hospitals use systems such as Hyper-Reach and many public safety agencies use their alert systems both for staff communication as well as public alerts.

There may be other roles that emergency alert systems like Hyper-Reach can play.  If you have any thoughts, feel free to suggest them.  We’d love to hear from you.


Are Government Mandates the Solution to Alert Registration?

Local officials in Kennett Square, PA have a new idea for how to solve the problem of folks not registered for emergency alerts.  They’re suggesting that registration be mandated by law.

According to this article, the borough council president there  is proposing that “officials mandate all residents of the borough sign up for emergency notifications”.  It’s an interesting idea, but it raises a lot of issues.  For example:

1) Privacy.  Many folks will object to requiring giving up the personal information.  That concern is probably overblown, but it’s a political problem. 

2) Enforcement.  How will officials know if someone is NOT registered?  Homeowners may be easy to find, but most municipalities don’t have a list of non-homeowners they can easily turn to.

3) Garbage data.  If someone wants to stay off the radar, they could register a bogus number.  They’d show up as “registered”, but with no beneficial effect.

4) Language issues.  According to Wikipedia, about 40% of the town is hispanic, but the registration form is only available in English.

This requirement is somewhat similar to that of the requirement for smoke and CO2 detectors, but there are some important differences.  For one, the building inspection process means that homes will be checked when they are sold.  And landlords might be responsible for insuring detectors in apartments.

We think it’s fantastic that a town cares enough that it would want to require folks to sign up for emergency alerts.  But perhaps local officials should start with a registration campaign – in English and in Spanish – and use a government mandate as a last resort.