As the Internet of Things (IoT) revolution works its way through marketing hype and seeks its place of valuable contribution within companies and industries, you might pause to wonder how IoT can create opportunities for your company. Yet that assessment is difficult in part because the buzz does not always align with reality. In short, it’s no simple task to discern the true potential of IoT today, leaving one to wonder: What is realistic, what difference could IoT make in my company, and how mature are other companies in embracing IoT potential?
These are just some of the questions explored in the IoT market survey recently conducted jointly between Radiant Advisors and Unisphere Research, a division of Information Today, Inc. The survey primarily polled North American business and technical people from various industries, including finance, government, and healthcare, and revealed the real-world progress organizations are making with IoT as well as the critical factors contributing to IoT success.
In some cases, the advent of devices and sensors enables enterprises to collect new types of data about their processes, products, and consumers’ behavior for insights and opportunities. In other cases, sensors and innovative devices generate streams of IoT data that are being processed and analyzed to create new forms of business or consumer value, sometimes disrupting industries with hyper-personalized or on-demand business models (among others). Regardless, an overall positive sentiment is driving organizations’ desire to make IoT a reality, and early adopters are tackling the challenges of data security, infrastructure, and skill sets necessary for this new data.
Enterprise IoT Adoption
A significant portion of IoT adopters are involved with their own internal research and education, and 82% of respondents are currently seeking education. Notably, 54% of the respondents state research/education as their highest level of IoT adoption. These companies are looking for applicable examples, devices, and practical scenarios that will spark the inspiration for a business use case to test. In their search, people consistently report looking to online publications and resources, webinars, white papers, and industry events for case studies as their primary educational tools.
Of organizations that are advancing beyond the initial research phase, 33% of respondents plan to start exploring or working with IoT in the next year, another 33% are already in the first year of their initiative, and 33% have more than one year’s experience.
The primary drivers for research and adoption are to begin or increase new business lines (30%), increase product or customer knowledge (21%), and gain operational benefits such as reducing costs (19%). Here, enterprises seek to create new business lines and revenue opportunities where data is their product, enabling new services not possible before. Increased knowledge about customer behavior and product usage improves internal efficiencies around product development.
It’s also interesting to examine the motivation and champions for enterprise adoption, as well as the challenges that surface along the way. Champions driving IoT initiatives are mostly CIOs, IT leadership, and architects, according to 49% of buyer-side survey respondents (non-IoT vendors). In most cases, these leaders represent the strategy and vision for an enterprise IoT adoption program with their experience and expertise to ensure that technology investments are made within the context of the broader enterprise technical architecture. The next largest group of IoT champions consists of analytics and data science teams, as reported by 36% of buyer-side respondents. This segment represents people that understand the business use cases, and the value that can be exploited with IoT data. The analytics and data science teams are likely to drive evaluations and proofs of concept in smaller, controllable experiments. Likewise, IT leadership is likely to take the broader strategic approach to the feasibility of data volumes and technology requirements that many data science teams will have for an IoT data environments.
Business Disruption With IoT Adoption
With new lines of business/revenue identified as the No. 1 motivation for IoT adoption, it’s no surprise that survey respondents seek more out-of-the-box thinking from executive leadership to drive the initiatives and provide a vision for the company’s IoT strategy. This may be a struggle for entrenched enterprises that tend to prioritize within the confines of existing business models, revenue generation, and traditional organizational infrastructure investments.
IoT adoption for the greater efficiency of key processes is valuable—but it is not necessarily disruptive. Product innovation can be derived from data science on IoT data in the context of the existing business paradigm, but this is still different from business-model innovations and industry disruption. It is a difficult balance and challenge to seek disruption while continuing to leverage a company’s core competency and brand.
When we consider true disruption, we can look back to how internet ecommerce upended the retail business with its interconnectivity, automation, and efficiency. Alongside these technology trends, businesses began to give away free valuable services just to collect and analyze data anywhere they could get it—including internet, mobile, and now IoT sources. More recently, we’re witnessing the asset-oriented business shifting to a borrow-when-needed consumption culture (or sharing economy) with companies such as Uber and Airbnb. We see more cases of big data disrupting business, demonstrating that data itself is the product, and insights and algorithms derived from big data can be more valuable than traditional business products. Similar to the internet, ecommerce programs, and early smartphones, IoT is a technology that collects a gold mine of data, just waiting for the right creativity to devise the next killer app or transformative business model.
In the pursuit of disruption, enterprises are initially implementing IoT infrastructure and devices for operational efficiency and then conducting proofs of concept in concert with data science teams to explore new opportunities. These organizations will begin with a vision for how to deliver value with individual IoT devices, how to deliver value with the aggregate of all devices, and then how to monetize the data that’s generated in new, original ways.
In the meantime, IoT devices and sensors will quickly answer the consumer or business question for “How do I know when to …” This is possible through continuous data stream monitoring and event detection. Analysis of this data can permit organizations to better understand when to replenish products in stores (based on shelf sensors rather than POS data) or to avoid machine failures and extend the working life of equipment. Data science takes analytic capabilities beyond event detection to probabilistic analytics in prediction and communicating data visually in order to take action.
The consumer IoT world is very imaginative, and the marketed messages that promise to simplify and add efficiencies to our lives often inspire people to think outside traditional convention. We’re familiar with mobile and wearable technology that captures GPS locations, acceleration, and heart rates to allow users to visualize and interact with data directly to better understand themselves and their environments. A smartphone can become a mobile hub for connected devices within a smart home to control heating, lighting, front door video cameras, window sensors, or a garage door. Similarly, a mobile health app can track individuals’ activity, weight, and blood pressure and can scan foods for nutritional data. For the most part, IoT devices can be programmed to do pretty much anything; that said, businesses should not get caught up in what they can do but should rather strive to create functionality that will be relevant and provide true value.
Technology Adoption and IoT Analytics
While nifty consumer devices and disruptive businesses get the lion’s share of the media attention, none of that would be possible without the data and technology infrastructure behind the scenes. The IoT market survey explored how companies handle the technology considerations and requirements for adopting IoT, including how much data is involved, where to store the data and for how long, and how the data will be leveraged—as well as important concerns for handling data securely.
The top three technologies that buyer-side respondents plan to add for IoT initiatives relate to properly supporting data science. Analytics or data science platforms led the technology wish list for 48% of respondents, and was followed by cloud-based big data platforms for data acquisition (40%) and technologies for data security, encryption, and masking (33%). These prioritized technologies are consistent with companies building their enterprise architecture for data centricity and analytic excellence overall. Only 13% of buyer-side respondents indicate that they currently have all the technologies, platforms, or services required to support IoT initiatives.
Beyond technology needs, the survey also looked at intent and goals for IoT data. Advanced analytics and data science algorithms rank as the highest priority for how companies intend to leverage either streaming or repositories of IoT data, according to 62% of respondents. This indicates that it isn’t IoT data alone that’s so valuable, but rather, what can be done with this new data. Basic aggregation of data for insights (descriptive analytics) and visual analytics were selected by 43% of respondents. The hot trend of machine learning and cognitive intelligence apps also ranks highly, at 41%.
A common perception exists that IoT adoption involves receiving and processing mountains of data. While this is a relative term, the IoT survey found that 25% of buyer-side respondents expect to capture less than 100GB of IoT data per day and another 13% expect to capture between 100GB and 500GB per day. Only 15% of buyer-side respondents expect to capture larger daily data volumes above 1TB per day, which could add up to processing and storing hundreds of terabytes or petabytes per year. Notably, 42% do not yet know the volume of data they can expect per day.
As for accumulating volumes of data and storage, 31% of buyer-side respondents are currently planning to store their IoT device data indefinitely. This is presumably to support data science and analytics applications that would leverage detailed historical data for identifying patterns and clusters or enabling regression analysis over time. Interestingly, 8% of buyer-side respondents state no intention to persist their IoT data at all. These analytic applications would likely involve pure real-time decisions, recommendations, and event actions on a stream of data with a limited buffering window of data. In either case, 40% of buyer-side respondents are looking to cloud-based big data platforms or services to acquire this data, which contrasts the 21% planning to add on-premises big data platforms, such as MPP databases, NoSQL, or Hadoop.
Moving Forward
Momentum in the IoT world will be fueled by opportunities around new business revenue and the excitement of being a hot trend. As more people prepare for their company’s initiatives with their internal research and education, they’ll be able to identify tangible business use cases to experiment with and demonstrate value. These limited-scope experiments or proofs of concept are more manageable initiatives for companies to undertake and learn from as opposed to launching full-steam ahead with strategic infrastructure to support IoT. This, in turn, will generate more stories, case studies, and patterns to accelerate confidence, and adoption.
The challenges in the future will be taking IoT out of company labs and putting them into the strategic enterprise architecture and management, where data privacy and security is not to be taken lightly. An acute reminder of the importance of security surfaced in late 2016, when hackers leveraged thousands of internet-connected devices as a platform to launch denial-of-service attacks of internet registry servers, interrupting portions of the public internet and ecommerce. This incident serves as a catalyst for companies to enhance requirements for securing devices and data transport.
A key challenge identified repeatedly in the IoT survey is the issue of how to find and/or educate people with the necessary skill sets for working with IoT data. These individuals often serve as data engineers who can work with flowing data in pipelines, as data scientists skilled in developing algorithms and coding, and as application developers who can embed advanced analytics into applications that interact with users or other systems. Skills gaps transcend technologies, and companies will continue to be challenged to find, develop, and nurture people as they advance their core analytics competency and strategy.
On a closing note, it is important to emphasize the need to manage expectations within a company. The IoT survey identified perceptions about when companies could expect to see returns on their investment. Transitioning from incremental value to enterprise-platform value and return on investment is difficult to manage as so many factors come into play for each individual organization. Setting the proper expectation for learning along the way, being agile, and building incremental value milestones will help avoid discouragement and setbacks as the imaginative IoT world unfolds.