Harvesting Value from IoT: Required Core Capabilities

Harvesting value in IoT is not all that easy. The traditional business models have usually not been equipped to make or monetize the tremendous amounts of data that comes from from a device or a sensor. So when a business wants to get into IoT, it needs to make sure that it has some core capabilities within the organization. It is not just enough to get on to the IoT bandwagon without having the capabilities to actually harvest the value. Depending on the organization, you may need different levels of these capabilities. So what are these core capabilities?

1. Additional Infrastructure: Many organizations definitely see value in IoT but are unsure if their existing infrastructure can support their vision. Their current operating equipment most likely do not have sensors attached to them. In addition to the sensors, they will need to invest in other technology infrastructure that can receive, store and process the data streams.
   
   
2. Human Capital: The current skillsets of the employees (even those in information technology) are probably not geared towards storing/sifting/analyzing the endless stream of data. Some firms are debating whether they should outsource the data analytics to a specialist provider or invest in the skills in-house. Some of the larger cloud providers can today provide sophisticated services that mitigates the need for companies to invest in employee skill development.
   
   
3. Processes: It is one thing for a company to have mature processes and governance in place. It is another for companies to have processes and governance structure tailored to meet the unique challenges from IoT. So far, the industry has not put in enough thought into governance of IoT enabled processes. Before venturing too far, companies ought to think through how IoT can affect their current GRC (governance, risk and compliance) processes.
   
   
4. Additional Capital: Investing into IoT may not be a cheap endeavor. Depending on the current operating environment, significant amounts of capital may need to be infused simply to develop the basic infrastructure, processes and capabilities in place.
   
   
5. Data Handling, Privacy & Security: If you look at the negative climate of data security incidents in the past few years (Home Depot, UPS, Target, and many others), it is clear that companies have just begun to take information/data security seriously. And all of this with data movement as a result of traditional transactional processes. When you couple this with an endless stream of sensor data from all sources, additional challenges certainly arise. Think about the ramifications of leaking PII (personally identifiable information) from a device planted on a consumer device. Think about data leakage of potentially sensitive geospatial information from government devices. The threat possibilities are simply too numerous to fathom.

Advanced interaction patterns in Internet of Things (IoT)

In addition to the six basic patterns described in an earlier post, there are three advanced patterns that are more specialized patterns of the basic ones.

1. Multi-Customer Interaction: In this pattern, multiple users interact with a single thing and business. In this situation, it is important to keep the interaction stateless between the interactions. Examples are parking meter transactions where customers sequentially interact with the kiosk to pay for their parking spot. Another example is that of virtual tool booths where a single sensor (thing) senses multiple cars (customers) and sends the information to the toll authority (business).
   
   
2. Multi-Thing Interaction: Multiple things interact with each other and a single customer and a single business. The things (devices) need to communicate/collaborate with each other to ensure that there is a smooth workflow for the transaction. So, if the workflow transaction occurs out of sequence, it may likely cause a confusion with false states. Examples of this pattern are connected cars which can communicate driving/traffic conditions to each other. Another example of a multi-thing interaction is the crop watering sensors deployed on a farm. The soil moisture sensors interact with each other to create an optimum water spray pattern for the crops.
   
   
   
3. Multi-Business Interaction: Multiple businesses interact with each other and a single set of customer-sensor pair in this pattern. Data sharing protocols between enterprises need to be established for this pattern to be viable. Governance, compliance, data sovereignty, data privacy are all considerations in the development of this pattern. For instance, the fire department, police and the nearest hospital may decide to interact with each other based on a smoke sensor emanating from a citizen’s home. The data vending and ecosystem monetization models use this pattern extensively.

In my next post, I will be talking about how to go about harvesting value from IoT initiatives.

Basic Interaction Patterns of Internet of Things (IoT)

When the industry talks about IoT, it does seem like there is only a singular pattern that is applicable across all industries and consumer segments.  The reality is that there are multiple IoT patterns that need to be considered when devising solutions or creating value.  In fact, this may be one of the main problems why an effective value framework does not exist to date.  In this post, I will be discussing six main interaction patterns in IoT.  In my next post, I will discuss a few advanced interaction patterns as well.

At the essence there are three main actors in a typical IoT interaction. They are:

1. End user or customer
2. Business or Organization
3. Device/Sensor/Appliance (the thing in question)

Kees, Oberlander, Roglinger, Rosemann (2015) in their discussion paper at the European Conference on Information Systems present the interactions between these three actors that yield the following basic set of interaction patterns:

1. Thing to Customer Interaction: In this pattern, the interaction is primarily is between the device and the end customer. Little or no interaction exists between the device and the business organization. Examples of this pattern are fitness wearables (Microsoft Band, FitBit, etc.), remote control (remote enablement of automobiles, etc.).
2. Thing to Business Interaction: In this pattern, the interaction and connectivity is between the device and the business organization. Many times, the end user of the device is probably not aware (or does not care) of all the interactions taking place. Examples are firmware updates of internet provider modems, routers, etc. Another example is a home where a connected refrigerator may be transmitting water filter usage information to the manufacturer.
3. Customer Centric Interaction: The customer is the key focal point in this interaction model. Both the device and the business communicate through the customer. If the customer chooses to enable connectivity between the other actors, then there is no data exchange. For example, smart operation of lights in a home can only be done by the manufacturer of the device with active cooperation by the customer. Another example is remote entry into a house; unless the customer explicitly authorizes each entry, the security device cannot be unlocked by the company.
4. Business Centric Interaction: The business is the key focus in this interaction pattern. The usage of this pattern usually implies that some sort of smart learnings are being applied in the operation of the device. Without the business getting involved in the middle and storing/analyzing the data, no smart learnings can be gleaned. A common usage is the drive monitoring OBD (On Board Device) supplied by insurance companies to ostensibly provide cost effective premiums. “Snapshot” OBD provided by Progressive Insurance is a good example. Another example for this pattern is NEST, the machine learning digital thermostat that monitors temperature conditions to optimize heating/cooling in the house.
5. Thing Centric Interaction: This interaction pattern is probably the least used in monetization models. Most usage scenarios for this pattern reflect emergency situations where the action based on device feedback needs to happen imminently. Consider the case of a sensor that detects abnormal radioactive levels in a nuclear power plant. Obviously taking into account the probability of a false alarm, algorithms are built by the organization that automatically activate safety procedures without waiting for customer intervention. Another example is that of a health bracelet that, upon detecting low pulse rate of the wearer, automatically alerts emergency response team.
6. Full Actor Interaction: The final pattern is that where all the actors (customers, businesses and devices) interact with each other actively. Consider the following situation: A driver allows her location information and current fuel level of her car to be continuously monitored and shared to the business. In addition, the driver has also enabled her calendar information to be shared with the business. The fuel gauge device automatically triggers an event when it detects low fuel levels. Based on this information (and perhaps time of day), the business can recommend various refueling options. Incorporating the calendar information, fuel levels, traffic conditions and time of day, the organization can recommend a course of events where the driver can refuel, eat lunch and then proceed to her destination in the most efficient way possible.

Valufication of Internet of Things: 5 Common Monetization Challenges

Despite all the hoopla about IoT, monetization has not been easy. Companies are still discovering how to make money off of this latest phenomenon. Currently, monetization models exist (as described earlier) and new ones are being developed. However, quite a few challenges need to be overcome before enterprises view this as a viable revenue stream.

Primarily, the issue is of data security. I don’t have to bring up the numerous attacks on data over the last few years. These attacks have primarily occurred on “data at rest” which means databases and other repositories of data (shared and secure drives both on premises and online in the cloud). Now, compound this with constantly streaming analytic data spewed from a multitude of sensors and the issue becomes even more serious. These sensors open up additional points of vulnerability that hitherto did not exist. In addition to the technical challenge of securing sensor data, which by itself is an enormous undertaking, the additional legal ramifications are significant as well. Consider the situation of a connected television where viewing habits are tracked. Even with the data anonymized, viewers may still be uncomfortable with the threat of their viewing profile going viral. Another example that comes to mind is Internet radio channels such as “Spotify” or “Pandora”. Do I really want someone to know the kinds of cheesy songs that I listen to?

The second issue is more technical. There are no unifying standards on how such streaming data can be obtained or shared today. A few protocols exist today where emitters can transmit sensor data to subscribed receivers. Most IoT initiatives are proprietary to the enterprise and therefore not much thought has been put into developing a single standard sensor data sharing protocol. Unless standards of data sharing evolve, enterprises are likely to spend significant amounts on data integration which might make the entire IoT value proposition suspect.

Third, the ecosystem monetization model is still evolving and quite specific to a few industry verticals. There is certainly value to IoT when the sensor data is shared internally in the enterprise whether it is for business enablement, operational efficiency or safety. The value of IoT data is multiple orders of magnitude larger when it is able to be shared throughout the ecosystem of the business both upstream to the suppliers and downstream to the customers. This cannot happen unless there is trust and security throughout the system. Consider the example of a retail giant which shares its consumer trend data to its suppliers. While the sensor data is still anonymized, the data provider needs to have adequate liability protection in ways that has currently not been thought of today.

A fourth issue is that traditional business models are today not equipped to take advantage of this potential value stream. Companies are still trying to figure out how to price information and successfully integrate them with traditional revenue sources. Business models need to change/adapt before IoT monetization attempts are made. I am not sure enough thought leadership has been developed in this area. My future blog post will actually propose a value framework for IoT.

A fifth issue is the need to modify existing product designs. Some manufacturing designs have been developed decades ago with little or no need to change. Obviously, these products cannot stream data or connect to the Internet today. Incorporating additional sensors into these traditional devices means that products will need to be redesigned and tested in the market before any commercial monetization model is adopted or created.

Valufication of Internet of Things - Monetization models

While monetization models are evolving with increased sophistication, today most of them can be grouped into primarily six categories:

Device Operations: One of the older monetization models for IoT is in device operability.  The earliest implementations of IoT were basic in that it mostly provided for remote operation and maintenance of devices. When the industry was (and still is) trying to figure out how to monetize this disruptive innovation, it provided for an initial revenue stream.  Examples of this model are in remote diagnosis of stalled automobiles, remote locking and unlocking of devices, etc.

Premium Service: Premium services is a model where the provider offers connectivity and operations for advanced capabilities.  They may or may not offer free basic services.  Examples are On Star Corporation which provides value add services for a premium.

Data Vending: This revenue model is still a nascent concept at the moment.  It has gained a lot of traction in the consumer goods sector where device usage data is collected (with the permission of the user) and then the data (either raw or analyzed) is then sold to interested parties.

Information Ecosystem: Industries with an established supply chain system are good candidates for the Ecosystem revenue model.  The data from the end user is passed back along the chain to the suppliers upstream in the chain.  An example of this is in the consumer goods sector.  Consumers buying habits and spending profiles have been always collected with point of sale data which was hitherto passed back to the suppliers.  Now, with IoT, even their device operational data can be analysed by the supplier.  For example, if connected refrigerators report consumers’ water drinking habits, the makers of the water filters can use that information to send replacement reminders at the right time.

Freemium: The free/reduced cost model is also one of the older revenue models.  Examples are internet providers who provide the modem for free while charging for the actual internet connection.  Another example is also the tollway devices (iPass, EZPass, etc.) where the device is free but then needs to be periodically replenished based on usage.

Demand Driven:  The market segmentation revenue model is probably the most sophisticated of the lot.  In this category, IoT data is extensively used to analyze usage patterns.  The patterns are then used to tailor is extensively analyzed to identify usage patterns.  It is then used to uniquely price different user segments.  A common example is the internet data plans offered by cell phone providers.to uniquely price different user segments.  A common example are the internet data plan offered by ings available to consumers.

In the next post, I will talk about common monetization challenges faced by the industry.

Valufication of Internet of Things - Changing Mindset of Value

Gordon Hui (HBR, 2014) discusses the shift of the value mindset due to IoT. He makes a distinction between Value Creation and Value Capture. He refers to value creation as those activities that increase the value of a company’s offering and encourage customer’s willingness to pay. Value capture, according to him, is simply the monetization of value.

One of the key distinctions between the traditional mindset and the IoT mindset is in the product offering. The traditional mindset has a concept of a standalone product versus the continuously updated product in the IoT mindset. We have been doing it for a while in the software development industry. To a certain extent, we have also done in the hardware industry as well. Think firmware updates to your hardware devices. But largely, this concept is just beginning to be adopted in the consumer goods industry.

Another area of distinction is in handling customer data. Traditionally, it was one data points during the purchase and perhaps a few more during the maintenance cycles. In today’s environment, cheaper IoT controllers and near-ubiquitous internet connection provides a continuous valuable stream of data. This data stream not only helps companies to solve customer needs proactively (increased service quality) but also provides a recurring revenue stream (sell value adds).

The following table captures his description of shift in value mindset more completely:

		Traditional Mindset	IoT Mindset
Value Creation	Customer Needs	Solve for today’s needs reactively	Proactively address needs through predictive analytics
	Offering	Stand alone product that cannot be updated	Subscription oriented product refreshes
	Role of Data	Single data point during purchase	Continuous information convergence
Value Capture	Profit Path	Sell more of the same	Recurring revenue model
	Control Points	Feature sets, IP & Brand	Personalization, Context
	Capability Development	Build core capabilities and add improve over time	Use eco-system and customer experience to continuously improve capabilities

In my next post, I will talk about monetization models in IoT.

Valufication of Internet of Things - Background & Current Landscape

Background & Current Landscape

Sometime in 1949, the bar code was invented which set the ball rolling for machine reading capabilties. ARPANET was used to send the first electronic message in 1969. In 1973, the patent for RFID was awarded. 1974 marked the beginnnings of what we know is the core backbone of internet communication; TCP/IP. Universal Product Codes (UPC) began to be used commercially in supermarkets. The first use of micro-controllers were actually when Carnegie Mellon researchers used them to monitor stock and dispensation of products in campus vending machines. Domain Name Service (DNS) was introduced in 1984. Five years later, the world wide web (WWW) was introduced but the first web page was only able to be created in 1991. In 1994, the first wearable webcam was created.

1995 saw the true beginnings of E-commerce. This year also saw wearable computing being developed. LG announced plans for an internet connected refrigerator with a purpose of …. In 2002, Forbes published the first article on the Internet of Things. Sometime in 2003, IoT as a term goes mainstream. Walmart starts to use RFID extensively to track its inventory and supplies. IoT gets further impetus with the introduction of cheap mass produced controllers in 2005. The first IoT conference was held in …Between 2008 and 2009, more things than people were connected to the internet. The age of Internet connectivity of things was officially here. In 2008, the US National Intelligence Council determines that IoT is a disruptive influence. Finally, in 2011 research firm Gartner adds IoT to its famous Hype Cycle.

“I could be wrong, but I’m fairly sure the phrase ‘Internet of Things’ started life as the title of a presentation I made at Procter & Gamble (P&G) in 1999. Linking the new idea of RFID in P&G’s supply chain to the then-red-hot topic of the Internet was more than just a good way to get executive attention. It summed up an important insight—one that 10 years later, after the Internet of Things has become the title of everything from an article in Scientific American to the name of a European Union conference, is still often misunderstood.” – Kevin Ashton, RFID Journal 2009

Today, IoT is a term that is much used but its implications are not well understood. Every company wants to get on the IoT bandwagon but really does not know how to commercialize the phenomenon.

Commercial Potential

The internet (an ironic source considering we are talking about IoT) has pundits and research firms galore talking up the potential commercial potential of this disruptive technology. We have estimates of potential value ranging from 2 trillion dollars to 20 trilion dollars by the year 2020. There is a lot of hype and speculation in the industry resulting in such wild estimations most of which are not backed up by empirical evidence.

While the industry estimates look very promising, the reality is that most executives have not started any meaningful initiatives. Most of them are still in pilot stages and have not yet realized the full potential that has been laid out in the business case. The graphic below gives a general idea of how the industry views the potential in this space.

From the Pundits...

Now that I have set some background and context, I will be talking about the shifting mindset of value and the commercialization models often used in IoT. Stay tuned...

Value of Value Conversation

Why do we need to have a value conversation? Is it not enough to produce great deliverables and services and let the receiver decide on how useful they are to him/her? Unfortunately, it is not. Often times, customers are convinced of the usefulness of a service they have received but are not able to fully articulate its value to others. For example, consider the case of a CIO of an insurance company who knew that she needed to do something about Big Data given the numerous requests from her business units. She was pretty aware of the high level capabilities of her IT department but unaware of how her data and information was setup for Big Data. So she brings in an outside consultant who not only provides an assessment of her current capabilities but provides a roadmap for a recommended target state for the insurance company. After the engagement, how does the CIO articulate the value of this assessment and the roadmap to the business? What was the consultant able to do that her own staff were not able to do so? From the consultant’s perspective, she would like to stay engaged at the insurance company for obvious reasons. So she needs to show to the business, through the CIO, the good work that she was able to produce for her. How does she begin to demonstrate the value of her work? These are precisely some of the reasons for the need for a structured value conversation.

You don't get paid for the hour. You get paid for the value you bring to the hour. – Jim Rohn

Consultants and Advice Givers Perspective

Management consultants and enterprise architects have concrete deliverables just like their technical consultants (developers, project managers, etc.) However, while a technical deliverable such as the implementation of a new feature or completion of a development milestone can be easily measured for its value to the business, strategic deliverables cannot be directly measured. When a management consultant provides advice to customers in terms of roadmaps and assessments it cannot be easily converted to quantified value by the customer. A value measurement framework that quantifies improvement in productivity, decrease in risk, etc. would be a useful tool in evaluating the deliverables from a strategic engagement.

The value of an idea lies in the using of it. – Thomas Edison

Executives and Advice Receivers Perspective

Executives struggle with value all the time. This is because of the nature of the ambiguous decisions that they have to make. As their decisions are usually at a strategic level, one normally does not see the result immediately. Often, these decisions manifest into success or failure after months and sometimes years of implementation. Therefore, any tool that can measure potential value, i.e., value of potential choices of action, would be a welcome supplement to their decision making process. Measurement of realized value, i.e., when a decision has been implemented, is also useful in vindicating their stance.

Introduction to Valufication

Introduction

Before we begin to discuss this subject, it is important to agree on a working definition of value. Obviously, various definitions of value exist. In fact, value is such a subjective aspect that more often than not, it is in the eye of the beholder. So what then, is the definition of value?

Merriam-Webster dictionary (Definition of Value, 2014) gives the following definition of value.

¹val·ue

noun \ˈval-(ˌ)yü\

: the amount of money that something is worth : the price or cost of something

: something that can be bought for a low or fair price

: usefulness or importance

For the purposes of our discussion, let us agree that value is the worth of “something”. When quantified, it can be measured in dollars. When it is not able to be quantified, value can be measured in terms of the goodwill it generates. Obviously, we cannot measure goodwill in absolute terms. But we can certainly measure it in relative terms based on a benchmark of sorts. The concept of a goodwill index is being developed in this blog. Analogous to the Dow Jones Index, the goodwill index is a relative measure of goodwill that fluctuates up and down over time. By itself, it has no meaning but when you compare two different activities relative to each other, the goodwill index of one over another helps to compare their values.

Then the next ambiguity that we need to resolve is the “something.” In the context of this discussion, it is an activity or service that is provided to the receiver.

Price is what you pay, value is what you get – Warren Buffet

Accordingly, for this discussion, we will define value as:

Quantifiable: The amount of dollars that can be exchanged in return for an activity or service.
Intangible: The amount of goodwill that is generated as a result of an activity. From a consumerization perspective, it is expected that all goodwill will result in something tangible for the value creator. For instance, better customer relationship will result in increased business in the future.

Implicit in the above definition is the fact that both quantifiable and intangible values needs to be both observable.

So then, let us begin the definition of “Valufication”. I define Valufication simply as follows:


“The act or process of measuring value of a good or service”

• What is a cynic? A man who knows the price of everything and the value of nothing. – Oscar Wilde

Welcome to Valufication of Things

This is a site for a healthy discussion of value measurement for consultants and advice givers in any area. As consultants and advice givers, we often dole out valuable strategic information, roadmaps, guidance and briefings. How do we then measure the value of such advice? Many times, strategy is not tangible but nevertheless valuable. This blog is devoted to valuating those things that are hard to quantify. Oh, by the way, all posts and content on this site/blog is subject to copyright protection and cannot be reproduced without my explicit written permission.