I was happy to be invited to present on analytics at the emergency services edition of the 140 Conference (aka #140confNW) sponsored by Clark Regional Emergency Services Agency. The video of the presentation covering how to get started thinking about, understanding and using web analytics is below.
Here are my notes.
Thinking about web analytics
If you want to get better at doing stuff, one thing you can try is measuring and using web analytics. Here’s a simple overview of how you can get started.
- Figure out what your goals are. Be really clear about your goals, don’t just use tools because you feel like you have to use new tools.
- Determine what pages or online events signal that someone has reached the goal. For example, filled out a “volunteer” form and gets the “Thank you for volunteering” page.
- Look at all online traffic profile of the people who reach the goal in aggregate and see if you notice any patterns.
- Use those patterns to inform future improvements.
Web analytics and the real time web
For emergency services, decisions often have to be made very quickly. Some popular web analytics tools (Google Analytics *cough* *cough*) have a 24 hour delay. This might be too long.
The mission objectives that are solved by the emergency services industry are well-aligned with the strengths of the real time web. But not all measurement tools are well adapted to the real time web. This is a challenge.
- Google Analytics has a 24 hour data delay on their standard free package
- Another web analytics tool, Woopra, is a real time web analytics package
For many organizations, a 24 hour delay in data isn’t that big a deal. But for emergency services real time is of critical importance.
One way emergency services web professionals might use real time analytics is in identifying emerging trends or situations on the fly. Using real time web analytics, they can find out what words people were using to find the website in search engines. A sudden jump in traffic for a particular keyword may be an early indicator that something is happening which may require attention.
For example, if there is a sudden spike in traffic from people searching using terms associated with a health problem then the emergency services agency can quickly respond in several ways:
- Begin specific outreach to see if an epidemic is occurring.
- Provide more or more accessible information related to that topic.
- Notify other appropriate organizations (perhaps hospitals/clinics) that there has been a recent spike in traffic related to the health problem.
Using real time web analytics tools, emergency services organizations can begin to use their own websites as sensory organs for identifying trends and concerns of the populations they serve.
Understanding emergency services audiences beyond the internal web site
Many conversations that are relevant to the mission of an emergency services organization occur beyond the organization website. For example, people using Facebook or Twitter to communicate.
The challenge with off-site conversations is that they are often a jumble of chatter on varying topics from all over the world. The web is worldwide, after all.
However, using techniques such as applying a filter to search Twitter by location, an emergency services organization can begin to listen and observe the population it serves.
Some things which might be observable by applying such a filter:
- Social media traffic report: often eyewitnesses who use Twitter will let others know of traffic delays.
- Monitoring location-specific Twitter chatter for keywords which are related to the emergency services mission (“fire” “accident” “tornado” etc)
First steps in listening beyond the emergency services agency website
Sifting through all the qualitative data that populations are offering up on public channels such as Twitter is a challenge. For those who are using these tools, how is all that chatter filtered down into something meaningful?
If every online conversation within a 10km radius is streaming into your computer, how do find the one where someone needs help? How do you figure out if there are indicators of something big happening?
- Set up a location based Twitter search.
- Add a keyword to your search.
I have a reference to help people search Twitter by location with several examples.
The web isn’t just for broadcasting information. You can use it to listen as well.
The web is great for getting information out to the world. But it’s also a great way to get information into your organization.
You can use the web to listen. You can use web analytics to listen. You can use tools like a geocoded twitter filter to find out what’s happening in your community by listening.
The challenge of real time web analytics for emergency services (and everybody else)
The problem that quickly arises is that there’s a lot of data and not enough people to turn that data into information. There are not enough people to compare and contrast information and generate knowledge. As a result the opportunity to turn that knowledge into experience is passed over by those with decision-making authority. This continues a cycle which discourages innovation and continuous improvement.
This is an HR challenge. And it’s an HR challenge that is beyond the traditional not-enough-butts-for-the-number-of-chairs-we-own. It’s an HR challenge in that people at all levels of organizations–from implementers and front-line workers through to management–need a comfortable and safe environment to experiment and learn about data-driven capabilities and opportunities.
Some ways to reverse the cycle of non-innovation include:
- Identify your mission objective.
- Identify the trigger words that are related to your objective.
- Layer technology on so long as it supports your mission objective.
Ending thoughts on analytics, the real time web and emergency services
Things I hope you come away with, after this presentation:
- Know your objectives.
- Use social media and the real time web as an observation platform.
- Find technology and techniques that are well matched with your specific mission.