warning, bees ahead sign
PHOTO: Hubert Mousseigne

Just as economists study and worry about national debt, organizations should study and worry about technical debt.

A software development concept, technical debt refers to situations where, for example, a team implements a rapid software fix that solves an immediate problem within a tight budget and deadline, but in turn creates potential complications down the line that increase costs over the life of the software.

The emergence of the internet of things (IoT), and with it, the increase in deployments of systems featuring connected devices creates new opportunities for organizations to interact with consumers and new avenues for product development. But the IoT also introduces a plethora of situations that can result in technical debt. And that technical debt can affect the execution of business strategies.

Quick Fixes Have Lasting Impacts

Technical debt's impact is very real. A company may incur technical debt when it chooses an easy solution over an alternative that may provide better quality or create new opportunities over the long run. Many times, companies make such decisions when schedules are tight and they feel they have no choice but to bypass solutions that might take too long to implement. But they have to live with the consequences of their actions, one of which could be a lack of resources to maintain the chosen option. These effects can come to the foreground when a competitor introduces a great new product while the company is still grappling with the fallout from the earlier decision.

For years, technical debt was a problem mostly associated with developers making decisions about what architecture made sense for developing websites, apps, chatbots and other software-based offerings. However, as software has become more infused into products and services, and as data has become equally widespread, marketers and others have begun to face choices that could lead to technical debt. 

In the case of marketing, when making decisions about how media, analytics and data-driven operations should be executed, they have to weigh immediate results against long-term impacts. The result: Seeking the holy grail of an amazing digital experience for customers has exposed marketers to the trade-offs involved in software development decisions. 

Related Article: This Is How We DevOps

Imagine the Consequences

Organizations as a whole must recognize their decisions in IoT projects will have long-term impacts on the digital experience and put some effort into considering the consequences of their choices.

Imagining the impacts isn't difficult. For example, when a project involves IoT devices that are subject to regulatory mandates, such as medical devices requiring FDA approval, development scheduling conflicts can result. The desire to bring a product into production rapidly may be thwarted by a time-consuming regulatory approval process. And once the system goes into production, subsequent problems may arise involving issues such as the scheduling of software updates or managing recertification processes.

Machine learning can raise technical debt further.

IoT devices can generate large amounts of data suitable for machine learning. For example, a connected car, according to a McKinsey & Company report on IoT devices, can generate 25GB of data per hour. But that stream of information is made up of many different types of data and therefore will require a lot of exploratory work, such as examinations of time stamps and data formats in device APIs. Faced with such a challenge, businesses may feel pressure to make expedient choices — choices that may involve technical costs if they don’t take into account the way machine learning bias can creep into systematic decisions.

Related Article: Investigating the Cost, Integration and Other Realities of Digital Experience Platforms

Reconsider Minimum Viable Products

Businesses can start to overcome technical debt by recognizing the limitations of so-called minimum viable products (MVP). Conducting an overview of an IoT network and data creation in the associated devices can help highlight the proper prioritization of data-related adjustments for exploratory data analysis and encourage discussion regarding technical assumptions made by the team.

A blind relentless zeal for rapidly deploying an MVP can lead teams to incur a lot of technical debt when they roll out smart products or services. An entire organization ultimately pays for the debt in poor quality, poor customer experience, and poor market position.