Turning 'Data Exhaust' Into a Competitive Advantage

As the volume of data grows astronomically, decreasing amounts of it are being saved for analysis and monetization. But there may be considerable value in all that discarded 'data exhaust.' JL

Scott Snyder and Alex Castrounis report in Knowledge@Wharton, photo by Ruben de Rijke:

IoT will generate  400 zettabytes (or 400 trillion gigabytes) of data a year by 2018. Only a small amount (8.6 Zettabytes) will be sent to data centers for storage and analysis. With improvements in longer range, lower power IoT connectivity, smaller, lower cost sensors, and cloud computing with AI, more data exhaust can be analyzed for insights (and) turned into real-time actions. By turning data streams into insights and prescriptive analytics, value can be unlocked and provide competitive advantage through analytics and monetization.
A vast amount of data that is discarded — the so-called ‘data exhaust’ — actually hold a lot of value and could be tapped to create new competitive advantages.
Instead of the Internet of Things (IoT), perhaps we should call it the data of things or the internet of data?
IoT will generate a staggering 400 zettabytes (or 400 trillion gigabytes) of data a year by 2018, according to the 2016 Cisco Visual Networking Index. This is being driven by everything from wearables and smart home devices to high-end connected platforms like the Boeing 787, which generates 40 terabytes per hour of flight, or a Rio Tinto mining operation that can generate up to 2.4 terabytes of data a minute (more than 20 times what Twitter generates in a day). Despite this huge growth in data from IoT devices, only a small amount (8.6 Zettabytes) will actually be sent to data centers for storage and subsequent analysis — the ‘data exhaust’ is much bigger than what’s actually being analyzed for insights. But with the rapid improvements in longer range, lower power IoT connectivity, smaller and lower cost sensors, and cloud and edge computing with artificial intelligence (AI), more of this data exhaust not only can be analyzed for new insights, but also turned into real-time actions. Self-driving cars are a great example of harvesting large amounts of sensor data to learn safe and efficient driving behaviors based on each new scenario and environment.
According to BCG, 80% of the most innovative companies leverage data to drive advantages in their business. This is why companies like Progressive, Nike and John Deere continue to invest in sensorizing their product offerings and to reap the benefit of this in terms of customer insights, economics and business model innovation. Yet for most companies, the value of data is an afterthought in developing new products or service offerings, let alone improving their business operations. Instead of starting with how valuable the resulting data may be, they focus on just making something “connected” as the value.
Maybe some things should never be connected– like the i.Con smart condom that can tally the calories burned during sex, or the Juicera connected juicer, which no one cared about despite attracting $120 million in investments.
What if data was the driving force from the beginning, and the entire solution was built starting with the question of how and why one should leverage the generated data? Despite common misperceptions about IoT — that the value is in the hardware and connectivity — these are just mediums to collect data more efficiently and seamlessly, and perhaps even access data we could not get to before. For example, by embedding a sensor in an asthma or COPD inhaler to collect real-time data on when and where people use their inhalers, startup Propeller Health has reinvented how people manage respiratory diseases. After integrating detailed usage data across patient populations with external data like weather and air quality to provide real-time, personalized coaching, Propeller has shown a 50% reduction in unplanned asthma attacks with the potential to save billions in health care costs.

“The ‘data exhaust’ is much bigger than what’s actually being analyzed for insights.”

Likewise, Progressive Insurance is turning data on driving behaviors from their Snapshot vehicle sensor into improved risk models and lower premiums for participating drivers that exhibit good driving behaviors. It now represents over 20% of their direct channel revenues. In both cases, these companies are using IoT to make data collection seamless and invisible to the end-user while creating value for them and their customers.
On the B2B side, companies like John Deere have used IoT data to shift their business model. The average farm went from generating 190,000 data points per day in 2014 to a projected 4.1 million data points in 2020 fueled by the significant growth in sensorization of fields and equipment. By turning these data streams into insights and prescriptive analytics, or automated  decisions based on data, Deere moved from selling farm equipment to delivering ‘Precision Farming’ services, guided by their data advantage. Like the John Deere example, IoT data is enabling “as-a-service” (XaaS) business models in many other asset-intensive industries like aviation, mining, transportation and construction by being able to monitor and optimize the usage and reliability of each product or platform.
Data almost always contains significant value and information if one knows where and how to look for it. Data not only represents value in the information that it contains, but in its value as currency for companies to leverage with potential partners. Data can be sold, traded and so on. The methods by which value can be unlocked from data, and ultimately provide competitive advantage, is through various forms of analytics and monetization.

Racing to Win with Data
Professional motorsports is a form of IoT, with the car acting as the “thing.” On-car sensor data is transmitted in real-time to engineers for analysis. This data is transmitted to engineers in the pits via trackside telemetry based on radio frequency (RF) technology, and also perhaps over a network (e.g. WAN or cellular) to engineers in a remote location such as a race team’s headquarters.
The typical IndyCar has up to 80 sensors. Assuming a sampling rate of 1000 kHz, that’s 80,000 data samples per second, or 640 billion data samples during the Indy 500. Given the massive amounts of data generated every time the car is on the track, a primary source of competitive advantage is the ability to turn reams of data into actionable value in near real time. Enter analytics.
It is not uncommon for a two-car IndyCar racing team’s annual budget to be about $15 million, and some Formula 1 teams are up to $500 million. For existing teams, the ability of race teams to acquire sponsorship dollars, and the amount that they can get, are entirely determined by the team’s historical performance and competitive advantage, both actual and perceived. At a high-level, the two ways that race teams can increase performance and competitive advantage are either by “making the car go faster” (a common phrase of most race team members), or by gaining positions and possibly winning a race through superior race strategy or happenstance.
Making the car go faster is made possible through analytics of data and data alone. This data can come from on-car sensors and data acquisition, driver feedback, wind tunnels, seven-post shaker rigs, computer simulation, on-track testing, tire testing, fluid dynamics simulation and testing, engine dynamometer testing and so on. The ability to outsmart competitors through better race strategy during the actual race itself is heavily driven by both historical and real-time race data.

“Don’t let your IoT data exhaust go to waste. It just might contain the insights to fuel competitive advantage.”

In both cases, leveraging advanced analytics allows race teams to convert the data into increased performance (faster cars and lower laptimes) and also perhaps maximized race results through strategy. Some of the analytics is manually performed by data scientists and engineers, and other analytics is automated. Analytics in this case can be descriptive analytics, data visualization, statistical analysis and advanced analytics, which can include the application of artificial intelligence and machine learning techniques and algorithms.
The faster the car and the better the race results on average, the easier it is for race teams to secure money — and more of it. This is exactly the same in business. The more competitive advantages a business generates, the better would be its financial performance as it becomes easier to acquire, retain and increase the number of customers.
Turn ‘Data Exhaust’ into Value
Having a competitive advantage means that you or your business can provide some type of value that your competitors cannot. This can be through innovation, superior execution, better user experiences or a greater ability to obtain and convert data into value.
Data transforms the act of decision-making from one based on gut feel or educated guesses culled from history and experience into objective decisions based on patterns and predictions. Data also enables us to quantify the underlying assumptions upon which our decisions are based, and assign a likelihood and order of magnitude. Indeed, data lets us know by how much we can improve or optimize something.
IoT not only represents an opportunity to collect significantly more data on products, services, and operations, but it also provides an opportunity to drive a competitive advantage for your business, just like with an IndyCar race team. But in order to unlock this advantage, you need to lay the groundwork prior to “race day” in order to win. This includes the following”
— Instrumenting for Outcomes. This starts with identifying the potential opportunities for deriving value from IoT data and determining the most seamless and economical means to collect the data to enable these. For example, if an athlete is already wearing cleats or a jersey, perhaps we can just embed the sensor there instead of having to create a new device and calculate the cost.
Nike has made significant investments to extend the Nike+ platform (connected sneakers, wearables and apps) to collect more granular data on more than 7 million runners and their behaviors to increase sales through smart replacement (knowing when the treads are worn out) and more personalized shoes and apparel.
Under Armour has spent nearly $500 million to access 150 million digital fitness and health users, and launched HealthBox to unlock the same data collection opportunity and push even deeper into health and wellness data insights. In addition to identifying the most efficient means of collecting data (sensors and connectivity), companies also need to implement a big data environment to ingest, manage, and analyze the different IoT data streams they expect to handle.
— Tuning to Optimize. Once the instrumentation is in place to collect and store the data, you will need to have the brainpower and tools to turn this IoT data into insights and decisions that deliver business value. This means having a team of data scientists, data engineers, and architects that can both answer known questions to optimize performance, and answer new questions by discovering hidden patterns in the data. In smart city applications, understanding the flow of vehicles and determining alternatives to optimize traffic is a known problem.
Hypothesizing that ride sharing has a negative impact on traffic in certain areas due to a reduction on public transit use would be a new question. This means having a data and analytics team capable of quickly proposing and testing hypotheses using the latest predictive analytics and AI platforms (such as TensorFlow, Torch, or H20.ai) to generate recommended strategies and actions for creating the desired outcome.
Finding this breed of critical thinking data scientists is not easy and may require integrating employees with different skills (Python programmers, modelers, statisticians and business analysts) while building up a pipeline through other channels (like open data competitions on Kaggle or sponsoring student projects to identify new talent).
— Racing to Win. Having the data, insights, and recommended actions will allow you to further optimize your current business or platform. But fully unlocking the value of data often requires rethinking your overall offerings and business model with data and the ability to leverage it as a core advantage.
Given the ability of IoT to scale to incredible volumes of data around your current business operations and customer experience, companies have a unique opportunity to shift to a data-centric business model and offer their products as an on-going service such as a pharmaceutical company moving from selling drugs to delivering disease management (via patient monitoring) or a car manufacturer shifting from selling cars to transportation miles. In addition, companies can develop brand new connected products and offerings by integrating IoT data assets and data science capabilities upfront in their innovation process instead of treating them as an afterthought.

“Fully unlocking the value of data often requires rethinking your overall offerings and business model with data and the ability to leverage it as a core advantage.”

In summary, if IoT is the new internet and data is the new oil, don’t let your IoT data exhaust go to waste. It just might contain the insights to fuel the competitive advantage that helps you win the future race in your market. But first it requires building the capabilities to efficiently capture and analyze this data, and revamping your innovation process that leverage the power of IoT-generated data into new products, services, and business models that separate you from the pack. Get ready to start your engines.