Hi,
Welcome to Roll Call, a newsletter about the state of policy and technology in public safety, and the people shaping it. If you’d like to sign up, you can do so here. Or just read on…
TL;DR
As human infrastructure, the true value of public safety is difficult to measure quantitatively.
Every technology in law enforcement extends human problem-solving capacity in 1 (or 2) of 6 vectors, which raises the question of what products are competing without seeming like it:
There was a Roll Call piece a few weeks ago about the current tech stack (pt. 1) used by law enforcement agencies. There was one line that I’ve spent a lot of time since thinking about:
“The technology stack augments officers with an improved ability to move, see, categorize, communicate, remember, and deliver force.”
I argued even earlier that one of the reasons public safety will be the most important job in 2050 is immediacy. First responders are the primary applicators of law and medicine:
“Telecommunication officers and dispatchers evaluate calls for service and allocate resources. They determine the urgency of the incident and direct the appropriate units. Police officers are the human extension of the municipal code. They observe, enforce, and represent the city or county in court. They are the law, (usually) without a legal degree. Firefighters and emergency medical services are the first line of medical defense. They are the tourniquet of the city or county that ensures survival until a doctor visit is possible. They are medicine without a medical degree. Dispatch, law enforcement, fire, and emergency medical services (EMS). The human infrastructure of society. The most tangled, spooky, messy parts of civic life."
This theory of first responders as human infrastructure and tech as augmentation combine to give us a far deeper insight into public safety technology and how it differs most significantly from ‘tech’ in its classic Silicon Valley form. Today’s note will take this rough cut of an idea and polish it into a useful framework for thinking about the value of technology in this space. It’s Infrastructure Week, baby! We’re going to make it happen, even if the federal government can’t/won’t.
Infrastructure
A product-focused director at a public safety company recently told me that it’s extremely difficult to learn how to be an effective product manager at a public safety company. If not impossible, it’s ill-advised. The reason, he said with a shrug that I could feel through the phone, was that there is no one data point that you can improve and thus prove the product’s value. In B2B software, an enterprising individual can look for some process/data point/metric that, if truly improved, sells the solution on its own. Say you built a system for online retailers that moves customers from an 85% shopping cart abandonment rate to a 70% rate. That sells. The metric clearly improves upon the status quo, and there is a dollar amount attached to that rate. The price you charge will be less than the value added to your customers, and everyone makes more money. The pie grows.
In public safety, what is the one metric that matters over everything else? “Wow, Mitch, you couldn’t figure that out? It’s lives saved, you dunce.” Okay, point taken. Lives matter, and public safety is designed to ensure a citizenry that is safe and alive. However, there is a difference between identifying the metric and addressing it. Addressing it is made trickier by the fact that it’s defined by its negative: deaths are easier to tally than lives saved. How does you measure lives saved? Who is accountable for that metric? What is the baseline? How much variance across cities is permissible? Can you isolate the success of any one variable? Does anyone care?
Again, of course people care about saving lives. People care a lot about the story of the firefighters saving the family and the dog. They care about the anecdote of EMS performing 8 minutes of chest compressions to save the elderly neighbor. We also care about lives unnecessarily lost, like when telecommunicators aren’t able to locate a 911 caller in time. A great deal of gray area exists between anecdotal extremes of lives unexpectedly saved and lost. No one, from residents to first responders to city officials, has a good grasp on what the right metrics are. We know RapidSOS helps locate callers faster. We know Tasers are less lethal than firearms. We’re not exactly sure how much faster, how much less lethal, or if they’re deployed optimally every time. So we rely on anecdotes. Sometimes the anecdotes add up enough to where a community begins to assume a narrative. Based on the narrative, public officials react. They’ll buy products or get rid of processes, lose or keep their jobs through elections and appointments. But we never really know if it’s working, we only know status quo vs bad enough that we need to fix it.
This is the same way that we treat infrastructure: good is normal, normal is normal, bad is tolerated right up until it isn’t. We take a spectrum and interpret it as discrete buckets of quality. The point is that public safety operations deal with differences between reality and perception, which places it closer to traditional infrastructure than modern business.
Let’s add one final layer to this cake that will help us understand the essential purposes of public safety technologies. Bear with me, I want to be careful to not come across as a pop psychologist here. Consider that the problems public safety addresses are almost purely human problems. They are not supply chain or marketing or customer experience problems. I think it’s safe to say that issues of emotion, anger, and poor decision-making are best addressed with equal and opposite measure of rationality, empathy, and good decision-making. A good law enforcement call for service involves an officer finding a tense, highly charged scene and defusing it. Take it from a 10 to a 2. The best tool for responding to someone in medical distress is a paramedic who knows how to react in specific circumstances. Firefighting is a little different, but the main method of extinguishing most fires is a human with a hose. As a baseline, the applicator of public safety solutions is a (trained) human being. Mobile, verbal, dexterous, adaptable. What has changed over the last 50 years are the tools that extend the human capacity to solve these problems.
Augmentation
So what does it mean to say that a tech stack augments officers? If human problems demand human solutions, then the tools need to accentuate and extend human abilities. In a law enforcement context, I argue that there are only six things that matter, six abilities to be augmented:
Sight - We only have two eyes, and we have to blink. There’s significant value in increasing the raw number of eyes, the width of angles, and the points of view during a particular incident. (e.g. CCTV, BWC)
Memory - Adults have a short term memory capacity of seven, plus or minus two. Software has eaten the world because of the ability to store vast sums of ‘memories’ cheaply, and public safety is no exception. Criminal law rests on this human capacity, and the databases within law enforcement are vital for solving and preventing crime. Selectively applied, it is dangerous. Humans have a strong innate ability to categorize things quickly; I’m including categorization as a subset of Memory (categorizing categories, whew!). Most of the most potent and convoluted questions in public safety today are around data retention policies. How long should memories be stored? (e.g. RMS, case management)
Communication - Public safety is a team sport, and the responsive nature of the work demands tight communication loops. After incidents, many teams and localities need to share and filter information. (e.g. radio, broadband)
Movement - Fast response is the name of the game. Different resources need to be deployed where they are needed, in different terrains or structures. (e.g. vehicles, weapons)
Identification - Public safety revolves around recognizing unique entities: people, locations, vehicles, property. The power of Identification is most like Memory, but requires finding and pulling a specific entity from the pile. (e.g. fingerprint scanner, geographic information system)
Force - This is why the job is difficult and fraught with controversy. These tools increase the range and power (and safety) when incapacitating suspects. (e.g. Tasers, sidearms)
A reminder of the CFS lifecycle and those involved:
Now we bring back our modern duty belt diagram, with software on top and hardware below. Augmented with flair:
A few takeaways. First, hardware usually extends one human ability. Camera-based systems are just extra eyes, until connected software makes that data accessible and valuable. LTE Devices are counted here as simply Memory augmenters, but because of their connectivity can pull on all of the value in the software above the CFS line. Furthermore, because they share the same use case time slot as push-to-talk devices, it is likely that attempts to merge LTE and Radio hardware will be frequent. We saw this at IACP 2019 with Motorola’s APX NEXT. Second, I’m not sure what to think of the long term value of purely Memory-focused software products. The workflows and categorizations are complex in case, physical property, and digital evidence management systems, but they seem to have the shallowest moats. Third, the power of location data is astonishing. Nailing the combination of increasing vision for multiple types of users and positively identifying unique locations of interest is powerful. What’s to stop ESRI (or other GIS player) from building a CAD, RMS, or investigation management tool? Would be very interested to see how they’d approach it, leveraging their unique depth in geographic modeling.
I’m going to continue playing around with this framework. There’s more valuable insights buried here.
Thanks for reading. If you enjoy this newsletter, please share it, forward it to your friends, or just sign up here.
-MA