Stalking the Silent Killer of Industrial IoT
There is a silent killer amongst industrial IoT solutions — the failure of the product business model. Click here to learn more about the silent killer of Industrial IoT.
Join the DZone community and get the full member experience.Join For Free
All enterprise software projects have a notion of costs and expected business impact, but the relationship is nowhere so tightly coupled as with Internet of Things (IoT) initiatives and the connected products and services they seek to deliver. Your IoT system won’t just be a system that your product connects to; your IoT system IS your product. Your customers are buying the business outcomes produced by the overall system — both the physical activities of the equipment and the data produced in the execution of these activities. Your customers are buying the insights and recommended actions produced by the combination of machine data, existing enterprise data (both yours and theirs), and 3rd-party data (weather, fuel prices, etc). The software components of your IoT solution are as much a part of the Cost of Goods Sold (COGS) of your connected product as the gears and bolts inside each physical unit. As such, it’s critical to optimize your software development, maintenance, and operating expenses for the life of your connected products in order to ensure a profitable business model.
Building the IoT Business Model
While the costs of gears and bolts can indeed fluctuate from time to time, your organization probably has several strategies and domain expertise for maximizing cost predictability. This is true for your enterprise IoT software components as well. Your development teams are surely made of smart people who can figure out how to connect your equipment at the edge to AWS or Microsoft Azure and display that data through a variety of mediums. They get a system running, demos go well, and your customers agree to run pilots with your connected equipment installed inside their facilities. These turn into paid subscriptions and more customers are connected to your solution. The business appears to be up and running.
Breaking the IoT Business Model
Then, the feature requests start to come in — some from your external customers, some from your internal sales and service teams. Integration with another CRM, monitoring a new product line, customized notifications — this is where development teams often begin to realize the limitations of their IoT solution architecture. Your teams often find ways to make the required modifications and keep the system running. But, as we’ve said in the past and will continue to emphasize, IoT is no longer so much a technology challenge as it is a business challenge. A series of patches, workarounds, and bridges may keep the system running, but both the daily operating costs and risk of ‘surprises’ begin to increase over time.
A connected product ‘surprise’ is like a minor heart attack caused by a lack of best practices in diet or exercise. The system works under normal conditions, but a sudden shock or strain reveals the underlying weakness. An unexpected event unleashes a runaway process or data surge, resulting in huge auto-scaling hosting bills or cellular service charges that consume 6 months of revenue overnight. Alarm bells ring, the team scrambles to fix the immediate problem, and normal service is restored. Such wake-up calls are usually survivable, and appropriate interventions and behavior modifications can result in a more predictable future.
IoT Business Model Failure
What sadly occurs with increasing frequency, however, is the massive coronary failure, resulting from the slow accumulation of inefficiencies and systemic weaknesses over time that goes unnoticed until it is too late. This type of failure can stop digital transformation in its tracks and is fatal to both the connected product business model and the careers of those charged with its success. This silent killer lurks in the shadows of complex enterprise systems, vast streams of dirty data, and a cacophony of external and internal user demands. It is fed by the engineering efforts of hard-working teams who are simply fighting an unfair battle that they cannot win on their own.
While there are cases where the technology itself collapses first under its own weight, the more common scenario is an inexorable constriction of space between revenue and expense for the IoT-enabled business model. Even where revenues generated by new differentiated connected products and services remain on target, the costs of operating sub-optimally architected IoT systems continue to rise and at an increasing rate with each feature addition and business integration.
What Doesn't Kill you Today Only Kills you Tomorrow
Your methods for managing initial volumes of data, machines, and users required a number of public cloud resources that you could justify based on expected revenue. As those volumes grow with the success of your business, however, your original architecture demands a growth in additional cloud resources at a rate outpacing revenue generation. Public cloud vendors have created incredible services. These services are also amazingly scalable. But, the scaling of resources means scaling of costs, and if your technology is scaling faster than your business, then your business is going to die. How well does your team understand (in both theory and execution) the 5 best practices for running IoT solutions at scale on AWS? Are they able to implement the critical portions of the Azure IoT Reference Guide for Industrial Equipment Manufacturers? There are many ways to build a basic IoT system — this can be done by connecting your device to the cloud and seeing data on a dashboard. There are fewer paths for delivering large-scale solutions with multiple product lines, customers, and user roles. While they are more challenging, ultimately, all that matters is the delivery of a system that can scale both at an engineering level and as a profitable business.
Published at DZone with permission of Marc Phillips, DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.