Logo PublicSonar

Six tips for monitoring social media for public safety and security

Local storms, floods, riots, (traffic) accidents, fires, events, all of these are situations in which you, as a municipality, province or safety region, want to know early on when something is getting out of hand and how the situation is unfolding. Every day billions of messages are shared online on Twitter, Facebook, YouTube, Telegram, blogs and (weather/news) websites. These form a goldmine of real-time information. But how do you make use of it? 

Monitoring public safety and security requires a different approach than social media communication with citizens or marketing-oriented engagement. You want real-time signals, 24/7. In addition, many messages are not directly aimed at your organisation. Also, people express themselves with many different words about the same event. Searching on ‘fire’ is not enough. You also want to search on ‘smoke’, ‘burning’ and ‘smouldering’. 

In short, smart searches are needed to see all the important messages in time. Without noise and irrelevant chatter. But where do you start? Below are 6 tips to get you started, but perhaps a live demonstration with practical examples can also help.

Tip 1: Be concrete about your search for risks

What risks do you want to monitor? For whom? How do you use this to make more informed operational decisions? 

Start with a list of all the types of situations you want to monitor. Think of severe weather, events, traffic accidents or troubled neighbourhoods. What scenarios are in your emergency plan? Does your region have specific risks, such as theme parks, industry, earthquake or flood risks? Also ask your colleagues. Are there any missing scenarios? What do you want to know about them? And how do you and/or your colleagues want to get the information (automatically) to you (see also tip 5 and 6)?

Social media monitoring major events

Tip 2: Define your geographical search area

Some safety risks, such as national riots or extreme weather, are national and may have local effects. But often you are only interested in events in your municipality or region. A fire or accident on the other side of the country is not as interesting as an accident in the heart of your city.

If you are mainly interested in local effects, determine which words/characters mark your region. Think of town and street names, specific buildings or locations of interest, nicknames of locations and outline an area on the map to retrieve all messages with a so-called GEO tag. 

With a solution specifically for public safety and security, artificial intelligence can give you additional suggestions or even automatically complete all place names, such as street names, for you.

Want to know more?

Feel free to contact us at info@publicsonar.com, follow us on LinkedIn or attend one of our free webinars.  

Tip 3: Determine who the authorities and influencers are

Besides searching by geographic features, you can also search by websites or accounts. Think about:

  • Regional news websites
  • Individuals (such as press photographers) who are always on top of the news
  • Associations
  • Sports clubs
  • Public figures
  • Specific sources about (regional) weather or traffic 

With intelligent software solutions, you can easily turn these authorities and influencers into groups. An additional advantage is that a smart solution supports you in complying with (privacy) laws and regulations.

Tip 4: Filter messages per risk subject

By following tip 2 and tip 3 you increase the chance of being complete and collecting all relevant messages. But at the same time, there are probably too many to handle! That’s why it’s important to create different searches per type of risk so you can filter. So think about what words people use per risk. 

For example, suppose you want to filter on “overcrowding” within your geographical area. You can then filter on different words, for example: “cramped”, “congested” and “packed”. You also want to exclude some words. To make this process easier, you can use intelligent solutions to select a (supported) building block that selects hundreds of words at once and excludes the irrelevant ones. In this way, you can filter the large amount of messages more easily. 

Tip 5: Set up alerts per topic

When there are safety incidents, you don’t want to miss a single signal. You can prevent this by setting up alerts in advance. Suppose within a few seconds/minutes there are several messages about “fallen tree”, “collapsed”, “injured”, “havoc” and “destroyed” that all correspond to your geographical search area, then you want to know about it immediately. An SMS, sent to a predetermined group of colleagues, saves time. This way, you quickly know how citizens experience the consequences.

Tip 6: Use dashboards & reports

Information has no value if it does not contribute to better operational decisions. Therefore, distribute the information via quickly created reports by emailing them directly to stakeholders. Or by using a dashboard in which several individuals or departments can watch developments live. For example, by projecting a dashboard as a real-time screen in a team room. 

This way you can immediately use valuable insights from public sources to take informed action and prevent escalations. 

Want to know more? Save your spot for a free webinar!

Crowd safety and major events: How to harness social media?

Our webinars cover:


The webinar is hosted on 16 December.

"*" indicates required fields