The promise of the Internet of Things (IoT) extends to almost all industries and markets. Coupled with the power of big data, IoT can offer massive improvements in messaging personalization and product customizations through complete customer empathy and supply chain efficacy. However, starting with IoT and unlocking its value often ends up being a daunting task for managers. This blog proposes a simple exercise to help design your IoT strategy and implement a customer and user centric IoT ecosystem.\nThe typical challenge faced by managers charged with designing and building the IoT strategy and ecosystem for an enterprise is often three-fold. First, how should relevant signals be defined, generated and collected. Second, how should these signals be stored and processed and Third: How should business value be generated from these signals.\nAn IoT ecosystem is not only hard to implement but hard to manage also. These ecosystems are highly prone to problems ranging from malfunctioning sensors that over or under sample, broken or suboptimal data delivery pipelines that lose or corrupt data, inefficient or ineffective processing systems that are not able to derive meaningful insights from the data and incomplete action and feedback loops that make it a challenge to adapt the system based on the newly identified insights. These problems make it harder for enterprises to generate additional business value and increased customer delight from their IoT investments.\nManagers struggling with how to get started with IoT should follow this simple approach. This approach is based on the fact and understanding that the strongest potential of IoT will be in the area of customer satisfaction through deeper and complete customer empathy. Managers looking to make progress should try and fill in the blanks as given below:\nIf I Knew ___ about my customer\/user, I would \u00a0___\nWhat information or insight into your customer or user\u2019s profile would enable you to make your product or service the best available option for your user built in such a way that it has the highest potential to delight them? Sensors can be used to collect this information about your user\u2019s profile and this information can then be used to either design better products and services or offer a better personalized and contextualized experience.\nFor example, Redbox; the movie and game vending machine company could use information about a user\u2019s demographics such as gender, age, occupation, education etc to better design the collection of movies offered at each vending machine. From the recommendations sent to the user offline (email) or shown when the user is at the vending machine to the offers made to the user while they are browsing movies and games can be better customized and optimized for conversion.\nBy placing sensors in these vending machines that can track the temperature, humidity, precipitation, foot traffic patterns and user characteristics, Redbox can instantly customize the look & feel of the vending machine, the messaging displayed on the screen around and inside a user session and the stock mix to better serve the predicted demand at a specific vending machine location.\nIf I Knew My Customer\/User was doing or planning to do ____, I would _____\nWhat information or insight about your customer\/user\u2019s plans or tasks could enable you to either design better products and services or offer your products and services in a more relevant manner? Sensors can be used to collect inputs that signal intent of the user in either near real time. They can also be used to collect information that can predict a future user need or intent.\nFor example, by embedding sensors in shoes, Nike can collect changes over time in how the foot makes contact with the shoe and how the shoe makes contact with the surface below it. These signals can be plotted over time to detect the optimum time for shoe replacement in order to minimize foot damage while maximizing user comfort and derived value from the shoe.\nThe same information collected from the shoe can also be used to model the physical activity and condition of the user and this information can be used for medical, physical and health tracking and improvement programs.\nThe same sensors can also be used to generate social signals indicating the user\u2019s location at any given point in time and this information can be shared with other users nearby who might be socially relevant or interesting.\nIf I Knew ____ about my customer\/user\u2019s habit, I would ____\nWhat information or insight into the customer\/user\u2019s habits and patterns would enable you to make or provide better products and services to them. Sensors can be used to collect data that can reveal patterns or habits in the user\u2019s daily workflows. These patterns can be used to build models of behavior or matched against known patterns to reveal the associated needs and wants of the user to best fulfil their workflow.\nFor example, understanding a user\u2019s travel patterns at the individual level such as between work and home, home and groceries or patterns on the weekends vs. weekdays etc could help Starbucks present better messaging, offers and deals to their customers that is adapted and customized to the user\u2019s presence across location and time. This could mean that Starbucks can identify the go to place for coffee for a user due its proximity to a location frequented by the user.\nIt can also mean that once a user enters a Starbucks store, the service offered to them can be customized to suit their needs and preferences. Starbucks could create a deep understanding of what the user likes and does not like on weekdays vs. weekends, mornings vs. evenings, summer vs. winter, near home vs. near work or at public locations like airports and sporting arenas and utilize it to provide superior customer service and drive delight.\n6 Tips For Getting Started With IoT\nOnce the goals and design of the IoT ecosystem has been defined, managers should keep the following tips and best practices in mind as they implement their ecosystem.\nTypes of Sensors: Sensors can be embedded in things that users interact with (stores, vending machines etc), in things that customers carry on them (clothes and shoes, phones etc) and things that customers use (cars, devices, appliances). Sensors can and should collect information about the user\u2019s environment (weather, location), user (demographics, what they are doing, how they are doing it, when and where they are doing it) and should be enhanced with information from other sensors that have relevant contextual time and location information such as correlation or causal events and locations.\nMobile Apps as Sensors: Mobile apps that run on devices have access to the various sensors embedded in the mobile device such as motion, temperature, ambient light etc. These apps offer a rich target for instrumentation to collect data from these sensors.\nBaseline Sensory Information: The location, time and channel of the sensor should be collected by default and used to tag all information collected about the user and what they are doing.\nBalancing Privacy and Value: Information payloads sent from sensors can contain a lot of information that can enable enterprises to build very comprehensive and detailed profiles of customers and end users. However, the data collected should be determined after due consideration to privacy and laws and regulations. Information should be stored and processed with the intent and goal to protect user privacy\nDesign for Change: Your business priorities, product mix and use cases will change. Designing an IoT system that can deal with new sensor endpoints, new sensor technology, new or updated information payloads and new ways to process the data on the backend will ensure that the IoT ecosystem continues to deliver sustained usage and value.\nDesign for Inconsistency: Sensors and sensory networks tend to be inconsistent due to hardware problems and\/or network problems. The IoT ecosystem needs to be resilient to problems, errors and outages. Design should incorporate the ability to handle and adapt to incomplete data, corrupted data or non existent data.\nConclusion\nEnterprises looking to capitalize on IoT need to think customer and user first. Designing the IoT strategy and ecosystem around the end user\/customer needs in order to deliver on their needs and wants adapted and contextualized to the demand is the key to success in the post IoT world. Success in this goal require that the entire IoT strategy and implementation is purpose driven to drive customer empathy and customer delight. At the same time, enterprises need to ensure that the data they collect and how they collect and process it is non intrusive to the end user, does not go against the expectations of usage of the user and is only used to drive customer satisfaction.