Shawn Rogers, Senior Director of Analytic Strategy at TIBCO Software
You don’t have to be a Silicon Valley insider to grasp the central role technology plays in driving disruptive economics. The rise of the internet introduced the digital economy, which has leveraged cloud power and big data to transform pretty much every industry with systemic changes in the delivery of information. The current hubbub over artificial intelligence (AI), machine learning, and advanced analytics signifies the new big technological and business cultural disruptor: the algorithm economy.
Opportunity no longer lies in simply amassing piles of information in various places throughout the enterprise, but in automating what kind of action and what kind of insight can be derived from it all. This is the world of analytics, where algorithms define action to manifest value.
Analytics now pervade almost every aspect of our lives, from the way we work to the way we communicate to the way we pay our bills, watch TV, book our travel, ad infinitum. All that big data equates to automated application everywhere we go. To draw just one example, how many people shop online today? And how many of those people wholly expect that, at the bottom of the page on their shopping site, they're going to get a recommendation that says something akin to, “Based on the fact that you just bought a golf shirt, you might be interested in these matching khakis” or “Your friends who also shop here purchased the following types of items.”
This kind of recommendation engine is a very simple but familiar indicator, since it shows how algorithms underpin modern digital experience and deliver functions that people have grown to expect. The value inherent in this expectation is what fuels the algorithm economy.
Broadly, there are three important paths for successful participation in the algorithm economy:
Enhancement: Embedding and enhancing the products and services that a company already has by using algorithms to infuse them with more insight, better service, and/or smarter ways of doing business with advanced analytics.
Creation: Creating brand new product lines and new businesses based entirely on data and algorithmic insight.
Amplification: Scaling data-science investment for productive outcomes.
Examples of enterprise forging these algorithm economy paths stretch beyond what we think of as typical “tech” companies and are found in the most unusual and unexpected places.
Producing value” is a great summation of what drives the burgeoning algorithm economy
For an enhancement example, take the case of garbage collection reinvention embodied in Bigbelly. While waste management isn’t something usually equated with big data and analytics, Bigbelly enhanced the level of trash service available to cities and municipalities via embedding data science in an existing product. Their municipal trash cans are equipped with sensors that communicate usage data to a home base about whether cans are empty or full, use that information to streamline truck routes for pickup, and often come outfitted with automated internal compactors to further increase efficiency. This business shows how an enterprise can add value to an existing product algorithmically, as Bigbelly allows cities to reduce waste and recycling collections by 70-80 percent on average.
An example of the second path, actually creating new lines of business in the algorithm economy, can be witnessed in a European mobile phone company’s establishment of a new kind of financial service catering to the “unbanked.” The company had long been serving customers in emerging economies, where populations often function via cash or barter systems and don’t possess bank accounts typical in developed countries. Those populations also tend to be heavily reliant on mobile phones and represent an enormous market. The mobile provider figured out a way to bypass western-style banking’s income verification and credit scores by mining their own troves of mobile customer data to check call histories, payment records, and geospatial graphing and determine credit viability. Phone usage data and analytics supplied a new method to assess risk and provide credit models for financial transactions, leveraging algorithms to create a previously nonexistent mobile-based financial product and unlock a new line of business for the provider built on existing digital relationships with their customers.
The third path to success in the algorithm economy, amplification, stems from the fact that data science requires a highly specialized skill set. Data scientists are often called “unicorns” in enterprise recruiting circles because they’re impossible to find, they're very expensive, and even lucky companies can only afford to keep a few of them around. That’s changing as analytics software becomes easier to use, but there’s still a long way to go. To move an algorithm-based strategy forward right now, you have to find a way to scale that unicorn expertise. One way is to tap into the wide variety of analytics marketplaces that have emerged. Just as we utilize the grocery store because we don't always want to make our own butter at home, there are analytics stores and marketplaces to speed data science work. So instead of reinventing the wheel, analytics teams can reach out to these marketplaces and with a click of the mouse, they're able to bring hundreds of applications into their environment without actually having to author all of them. There's Apervita for health analytics and data, Algo Market for agriculture and energy, Quantiacs for the financial sector, PrecisionHawk for the drone industry. Algorithmia is a marketplace with over 800 algorithms in their library for things such as natural language processing and sentiment analysis and a whole host of different classifiers. Many of them are free, some can be used for an incredibly small charge, and you can often drag and drop from a marketplace directly into the heart of your environment and into your work applications. These types of algorithm exchanges and marketplaces enable greater speed in “unicorn” work. They help inject a level of standardization and repeatability necessary to scale enterprise data science investments to produce the greatest value.
“Producing value” is a great summation of what drives the burgeoning algorithm economy. It is this quest to wrest value from information that impels businesses to embed algorithmic enhancements into existing products and services, to create new lines of business derived from data, and endeavor to speed and amplify their data science investments.