Abandoned vehicles and data silos

This blog was previously posted on Medium. 

We had a car towed today. It was a cause for celebration, since this car had been parked on the street in front of our house for months. And, hopefully, it will be cause for celebration for the owner, who reported it stolen months ago. However, we were left feeling frustrated with how difficult it was to get this done. Surely there was a better way? And why is dealing with almost every organization an exercise in how persistent you can possibly be until you can finally get a basic thing done?

This car appeared in front of our house three months ago. That’s no surprise — there are lots of cars that park on our street, and we certainly don’t have ownership over the area in front of our house. As the weeks went on, however, we started to wonder what was going on with it. We kind of assumed it belonged to a neighbor, and it wasn’t really hurting anything. One day, my wife started asking around and learned that it didn’t belong to any of the neighbors immediately around us. From then on, it took on the appearance of abandoned trash instead of somebody’s car.

Eventually, we got fed up with it, and did what a few Google searches told us to do: report it to the city’s abandoned vehicle’s department. They told us that they’d send an inspector out.

Weeks passed. Nothing happened, but it was the holidays, so we figured that might have something to do with it (and it really wasn’t top of mind).

Then, on New Year’s Eve, somebody parked pretty close behind our car, and left. After a few days, it started to feel like our car was sandwiched between two abandoned vehicles. That started to get frustrating.

Again, this is a public street, so we’re used to not having an ideal parking spot. However, there’s a big difference between getting what spot you can in an ever-changing system of parking and having to deal with two unchanging scrap heaps.

My wife called the city again, and learned that they had sent out an inspector weeks ago and closed the case. Since the vehicle had air in its tires and didn’t have expired plates, the abandoned vehicle department couldn’t do anything. Not having any kind of feedback loop was frustrating, and we didn’t know what to do after that. The non-emergency police line specifically says to not call about potentially stolen vehicles, and the abandoned vehicles department won’t do anything if the vehicle isn’t blatantly abandoned. What now? A clearly abandoned car was left on our street, for months, and we couldn’t do anything about it.

At this point, tons of people started telling us to just let the air out of the tires and call the abandoned vehicles department back. We were convinced that there had to be a better way, though. Our city can’t run that poorly or have that many inefficiencies, surely.

A spate of digging and asking around got us a lead: buried deep in the Next Door archives, somebody suggested calling the city’s auto-records department and inquire about a potentially stolen vehicle. Note that the website for this number doesn’t say anything about that, and everything else that we could find was talking about the abandoned vehicles department.

However, we called this morning, and the auto-records department confirmed that the vehicle had been reported stolen. They told us to then call the non-emergency police line, and tell them that they had confirmed the stolen status.

Within 20 minutes of doing that, there was an officer at the door, and a tow truck shortly after, and the car was hopefully off to be reunited with its owner.

Once we discovered the appropriate incantation, the city worked great. It blew us away how incredibly inefficient everything was to get to that point, however. Surely the abandoned vehicles inspector would be able to run a check on the license plate to see if that were the case? What about a mechanism for us to check on our own, prominently shown, without having to discover the local lore that told us about a particular number to call? It was incredibly frustrating how none of these systems worked together, and how much the burden was on us to push through bureaucratic challenges to get this resolution. And now, hopefully, we’ve helped reunite a grateful owner and their stolen car. Think about what the end result would have been without us being persistent, though — the owner never would have gotten the car back.

I think we all have stories like this, about various organizations or companies. The villain becomes the organization, instead of the thief or the cause of the problem. These are stories of personal inconvenience and frustration, but ultimately they’re stories of silo’d information and systemic failure to make data available in an easy-to-access and clean method. It’s another sign of how early we are as an industry in really making a difference with technology. So much effort is wasted on trying to get at data, and trying to turn vague signals into person-level information.

One approach to solving these issues is to consolidate everything into one single database. That’s led to the rise of the integrated marketing hubs, and the astonishingly complex systems recommended by IBM and the like. It requires a huge investment, a company-wide investment, and a massive retooling for uncertain gains. Ultimately, these gains also become increasingly inflexible. Technology, and the expectations of consumers, is moving so incredibly fast that what might work for one use case can quickly feel stale and brittle and fail to solve an end-user’s problem.

At work, we’re focusing on an open omnichannel approach. Rather than try and pull in everything from every other system, we want to play nicely with as many things as we can. Products like Urban Airship Connect provide an outbound stream of everything that we have, which means that developers can build directly on top of it and let us do the hard work and the heavy lifting. In addition to that, we’re building lots of connectors with other systems (many bidrectional) so that it becomes really easy to, for instance, trigger events in SendGrid Threads from things happening in your mobile app, or analyze engagement information alongside other events inAmplitude.

This approach of getting excellent point solutions to work well together, and fight against data silos, is gaining momentum and we’re going to see a lot more of it. A single solution just can’t keep up, and point solutions are getting better and better at solving customer needs. The next step is for them to work together. Let’s break down these silos.