Location Intelligence, Spatial Data Analysis | Pitney Bowes
The State of Data; The Strata of Commerce
It took geologists over 250 years to go from the simple idea that the study of strata – layer cake geology – could map the very history of the world, to the complex reality of being able to read and interpret that history with scientific accuracy and precision. It wasn’t until the discovery of the real time process of radioactive decay and the radioactive isotopes in those rock layers that we humans were finally able to divine their true meaning, and to see in their arrangement the slow unfolding of the history of the world and our place within it.
In business, data is everywhere. It is the “sediment” that contributes to the formation of our modern digital age. As with geological strata, layers of data can tell a more comprehensive story and provide more substantive insight together than can any single dimension, or even a set of dimensions that are kept apart. Once data is combined, each set positioned relative to and juxtaposed with each other, data becomes more than just a resource; it becomes a value, an asset, a fuel that can power insight, drive revenue and refine a customer’s experience.
We know this. We’ve known this for a long time. But like geologists, it takes the right identification of real time elements, the right tools, and the right perspective to make good on the promise.
Why Layers Matter
For nearly 100 years, Pitney Bowes has worked to develop those tools and perspectives, to organize the data around us to understand how addresses connect and bind people, places, and things together. With digital and behavioral data, happening in real time, we now have the ability to treat the address as a pivot or fulcrum for layers of data, converting the data record into a real time history of customers and their moments of need, as well as the impact of a massive array of variables on how the places those people inhabit and pass through are used.
Fundamentally, the difference between vast piles of data and useful knowledge is the degree, complexity, and value of the layers of data that can be integrated. Data, in other words, is a layering problem. Every divination of an address involves a host of variables: physical locations, structures and the materials of which they’re made, the people associated with it in the past and present, tax codes, school districts, zoning, weather, topography, and many, many more. If these various elements and attributes aren’t arranged effectively, if the property boundaries are off a few feet from the zoning, if the school district runs one block short of the tax codes, then the data is out of sync, and the errors in judgment that arise from that divergence might not be realized until they’re already negatively affecting the bottom line.
That’s why when Pitney Bowes looked to convert their century of knowledge about how addresses worked as a fulcrum for layering data, we did so by carefully combining the data and insights we’ve collected, along with data we’ve purchased through third-parties, and publicly available data. We then invested in aggregating, validating, building, and packaging this data with the goal of identifying and capturing real world change; simplifying the process for our customers to acquire and ingest the data; and then building the right fit-for-purpose products as quickly as possible.
This isn’t something we did to stand out from other data vendors. We did it because customers both demand and deserve that level of fidelity. Today, whether someone satisfies their data needs from us, or from another vendor, we encourage them to ask about the process of assembling and processing those layers of data, so that they can better determine if the data they’re buying will produce real results.
What’s in a Layer?
When we talk about layers, we’re talking about attributes with some sort of geographic reference.
Looking at a single layer of topographical data, for instance, will help you plan properly for a hike. The mountain is this tall, the elevation gain is this many feet. A map of homeownership in a neighborhood is a different layer of data. The streets in that neighborhood; the highways leading in and out of town; the interstates that connect our cities—layers of data tethered to location.
But a single layer of data can reveal far deeper insights when it is combined with insights gathered from the other layers that relate to it.
In one state, for example, local government planners worked to optimize the best route for a new interstate. Constructing the Interstate would require insights garnered from layers of information, including the environment (the highway should avoid remediation, and wetland areas), geologic (the highway must be on solid ground), hydrology (where the highway will cross water and necessitate bridges) and existing infrastructure (the highway should be routed around historically significant locations/buildings, like cemeteries or landmarks). Alone, each of these layers provide an insight. Together, these layers provide the blueprint from which we can construct a multi-million-dollar highway.
“The more complicated the question; the more data layers will be required to answer it,” says Chris Walls, Chief Operating Officer for 39 DEGREES NORTH LLC. This is a reality that is likely to grow more pronounced with time, as today companies are looking beyond aesthetics, and toward unforeseen truths that can reshape how they engage customers or how they develop and execute on strategy. They’ll only make discoveries of value if the data can supply them, and only then if the layering of the data is done accurately and interpreted correctly.
The easy metaphor is that of cartography or the production of maps from available land mass data. “Today’s technology has allowed us to construct maps in a fashion limited only by the amount of data you’re willing to ingest, says Joe Francica, Managing Director of Location Intelligence for Pitney Bowes. “We’re taking the construct of a layerable map to develop more complex models that reveal new insights we might not have been able to previously see.”
Insurance companies—property and casualty insurance, in particular, start with a map of a neighborhood, or a layer of addresses. Then they begin to work with Pitney Bowes to weave more data into this map, layering on, for example, environmental and weather patterns, to see which areas are prone to flood or wildfire. This allows them to create what they call rating territories, or boundaries. But the data doesn’t stop there. Insurers need to understand areas that are susceptible to manmade hazards, like crime, to determine coverage. So, Pitney Bowes helps them layer on historical, demographic, civic and municipal data, and a richer, more valuable picture begins to take shape.
The Future of Layers
Indeed, the future of data and analytics in business isn’t just accumulating more and more data. The future is the way we take high quality data, optimize it for access over storage, and combine it with other data to provide the insights companies need to make decisions. “Individual features within layers includes attributes,” says Walls, and “attributes can be unlimited.” If a picture is worth a thousand words, in other words, a map is worth a thousand tables.
Insurers spend a lot of time examining and analyzing very specific types of risk, so they can effectively save money through more efficient underwriting. With Pitney Bowes data and software, they can speed up their assessments and can come to informed decisions faster. There are a vast number of details that determine the level of coverage that a household will require; often the decisions and the models are decided by minute details, like what kind of roof a household has, and what percentage of the house sits on a flood plain.
Layers of data are valuable to other industries. Telecommunications companies are always seeking to provide better coverage, especially when planning a rollout of a new product or service. To determine which market to situate the launch, telecom companies are looking for population centers with multiple transmitters. The more transmitters there are in a given area, the shorter the wavelength, and higher the frequency—and the better the signal. But in addition to transmitters, the telecom companies need to decide what market they’re targeting.
Layer on the demographic data and suddenly these telecom companies can see not only the population center with multiple viable transmitters, but also the number of commercial businesses and residents’ age groups, income levels, and more in a single area. “Big data solutions allow us to ask questions we didn’t think were possible. We thought they would be too time-consuming, even on a computer. But now we have the capability to think larger because we can ingest more data and iterate on different possible scenarios. We are limited only by our imaginations,” says Francica.
“Everything today has a geographic component,” he continues. “But how do we make it relevant to somebody (or company) that is looking for an answer?”
There is raw data everywhere, after all, from traffic lights to cell phone transmissions. All of these attributes have value if layered properly, but often times companies don’t have enough information to know what layers to stack, and what algorithms to deploy to yield real insight and knowledge.
Companies need to take advantage of the years of investments data scientists, analysts, and strategists have made in performing critical functions: address verification, validation, and standardization, to enable proper layer alignment and empower decision makers with actionable information. We know, for example, that much of the data we collect is malleable. Street names change as populations grow and shrink—Walnut Street may become West Walnut Street, First Avenue named after a city’s retired mayor. Other locations can sit at multiple addresses—buildings themselves can and do often have multiple addresses—even their own zip codes. And meanwhile multiple buildings may share one address. If the data vendor isn’t tracking and coordinating these changes, then too much of the logistical work falls on the company trying to benefit from the information, and interferes with an accurate understanding of the commercial strata.
But when this is done properly, the result is magical, even revolutionary. As Mike Ashmore, the Director of Global Geocoding Management at Pitney Bowes notes, “We are taking what has historically been a specialist function, which is to interpret spatial data, and democratizing it. Our objective is to take this spatial data to the masses. That means organizations can build a business case to do things they’ve never done before because it’s too difficult or cost-prohibitive.”
Visit Pitney Bowes for more information.
You may also like
By uncovering ambiguous relationships between data sets, location intelligence coupled with geospatial analysis can help organizations find business value in big data.Read more