If you are old enough to remember when geography was taught in elementary school you undoubtedly recall the memorization of capitals, countries and cultures. To the curious-minded, geography uncovered the world, and maps displayed proximity relationships, offering a perspective not revealed by merely reading the textbook. Some thematic maps depicted the size of economies relative to GDP, for example. A simple but effective geographic depiction of statistics.
Fast forward several years to the late 1980’s when desktop mapping systems were available to PC users. Thematic maps introduced a new way of visualizing demographic and business data. But integrating customer information was still a challenging process and employing the most current demographic data was always the next decennial census away.
Then in 2005 came the great disintermediation of geography: Google Maps. Maps were no longer trapped in desktop mapping systems. Geographic information about countries and their economies and satellite images of the terrain were just an Internet click away. But Google Maps’ greatest impact brought location-based data to users easily. Businesses found a very simple and inexpensive way of mapping sales and marketing information. Still, not everyone recognized the benefits.
From 2004 to 2014, I founded and chaired the Location Intelligence Conference. The objective of this event was to place geographic information system (GIS) solution providers in the same room as those offering business intelligence (BI) software. Companies like MapInfo and Group 1 Software (both acquired by Pitney Bowes, the company for whom I work), SRC (now Alteryx), Esri and Tactician mixed it up with PeopleSoft and Seibel (both acquired by Oracle), TIBCO, Information Builders and Oracle. Google and Microsoft participated, as did GPS chip manufacturers. The result: the location intelligence ecosystem coalesced among some of the major IT companies and many startups. However, most BI vendors just didn’t see the value of location-based data and later stopped participating altogether. To too many analysts (1, 2) watching the field, spatial was not special; spatial data was just another data type.
Today, all BI vendors, including enterprise players such as Microsoft, IBM and SAP, are grasping for more improved geospatial technology integration and functionality. From better styled maps to more accurate geocoding, business intelligence solutions crave superior geospatial processing tools. Nearly every BI software product provides rudimentary mapping capabilities out of the box, but the vendors are scrambling to meet competitive threats. Maps and location technology, according to a senior sales director at Pitney Bowes, are now “table stakes.”
Why the change?
While it’s true that spatial data is just another data type, more of it is swamping CIO’s and CDO’s, every day. The key, a latitude and longitude coordinate pair, is attached to nearly every transaction, both virtual and physical. From tweets to check-ins, from fob payments to beacons, data with location is ubiquitous. The need, and the demand, to understand spatial extent, proximity and relevance is important to every business.
This still doesn’t explain the entire transformation. BI solutions are splintering into specializations to leverage unique visualization and analytical capabilities. Maps occupy a significant portion of real estate on BI dashboards because maps portray multivariate information relative to place, something with which most people can more readily identify. But, BI solution providers cannot simply duplicate basic GIS functionality, like putting pie charts on thematic maps, to be competitive.
Three significant industry trends are driving immense change today and will into the future: mobile payments, drones and autonomous vehicles. While seemingly unrelated, it’s all about data.
Mobile payments drive the need for identity resolution to detect fraud and provide retailers with improved customer segmentation. Payment data – where, when and how much – is flooding into data warehouses. Integration with data streams, aka “big data,” from points of sale (POS) and smart sensors puts increasing emphasis on processing data faster to reveal meaningful, actionable location intelligence in near real-time. The faster you can recognize a transaction as either valid or fraudulent, the faster you can flag that customer as loyal or corrupt, respectively. The next action, then, is to send them a coupon or lock the credit card. Spatial processing of data streams will be required and will be fully integrated with BI solutions.
Drones are driving two complementary phenomena: precision delivery (as proposed by Amazon) and the possibility of capturing remotely-sensed data from image sensors mounted on drones. Drone delivery may precipitate the need for global geocoding beyond streets, parcel boundaries or even building locations (e.g. what3words). Drones will reach any place on Earth anytime, anywhere. Need an auto part while stuck in the outback of Australia? No problem. As intriguing will be the use of drones to detect the composition of rooftops, land development changes, emergency events or any other geographic attributes that can be observed remotely. Privacy trade-offs will abound. But collecting ancillary data about the ambient environment will become a byproduct of drones or any unmanned autonomous vehicle (UAV) delivery system.
The evolution of autonomous vehicles is accelerating. The prospect of driverless cars, for example, demands highly precise digital street data plus image recognition of street furniture, signage and other obstacles. Surely, Uber purchased deCarta, a geospatial solution provider, because it believed that capturing an accurate street network for its vehicles would also require better geoprocessing. Its competitor, Alphabet (Google’s parent) is investing in autonomous vehicles too, and leverages its own authoritative source, Google Maps, for an accurate street network. In addition, a consortium of German auto manufacturers purchased Nokia’s HERE business unit that supplies navigable street centerline data. Realizing the potential of platooning vehicles to reduce highway congestion will require better digital road maps integrated with advanced sensor technology. The ability to quickly analyze streaming telematics data is essential to effective and safe operation of vehicles. For property and casualty insurers, employing telematics for usage-based insurance (UBI) is driving policy underwriting to capture individual driving behavior netted from sensors. Such sensors measure speed, braking frequency, cornering and other driver propensities. Multiply this by the millions of drivers on the road and you quickly see that UBI is one of the drivers of big data BI analytics, but may also accelerate the change to autonomous vehicles. Car ownership may drop in favor of ride-sharing and concierge services from Uber or Lyft, for example.
The use cases mentioned above are creating vast data lakes of digital geographic information. Other sources of geospatial information, such Earth observing “smallsats,” will add to the data deluge as well as many other sensors placed in buildings like visual light communication (VLC). Indeed, BI vendors will be pressured to embed advanced capabilities such as spatial querying and modeling. The “table stakes” will be raised and only the “location intelligent” company will reap the rewards.