The point of this blog series has been to suggest that the traditional departmental corporate form (i.e. sales, marketing, finance, ops, IT, etc.) remains virtually unchanged since its origin at the dawn of the industrial age. Back then, innovation moved in multi-year cycles and infrastructure and depth were more determinant of success than speed of innovation. If the two decades since the dawn of the internet have taught us anything, it’s that innovation is more determinant of success than any of those other things.
And yet, it is the very way in which corporations solve problems that is killing innovation in many of the world’s greatest traditional enterprises. From the way companies develop strategy, serve customers, and attract employees, to the way they create and manage work product, digital is driving the need for innovation.
In our first post, we noted that work itself is changing. The way work product is created (real-time collaboration), how business issues are communicated and discussed (electronically), and the way consensus is formed (via social media). We instead argue for a new management framework–the Digital Innovation Lifecycle–as an alternative method for driving innovation, speed, and agility out of the organization.
In our second post, we broke down the Digital Innovation Lifecycle into 3 major components:
- Productivity and Collaboration (PAC) – Tools like Google’s G Suite make it possible for the heartbeat and brain power of the organization to become a single force.
- Thinking and Planning (TAP) – The left hand side of the diagram, with its elements arranged like the 5-side of dice, captures how strategy, customer observation, and quantitative insight inform and reframe the process of solving business challenges.
- Doing and Creating (DAC) – The baseball diamond of topics on the right of the Digital Innovation Lifecycle. In this loop, we integrate product management concepts with agility, experience design, and change management.
In our third post, we dove more deeply into the left hand side of the TAP diagram. This upside down triangle on the left reflects the idea that strategy guides user research and that the learning from user research often reframes strategy; while quantitative insight informs strategy and user research. In the classic corporate model, the development of a strategy is divided into pieces by department. Using the Digital Innovation Lifecycle, that same problem set is addressed holistically by a multidisciplinary team.
In this post, we turn to the triangle on the right hand side of the diagram and extend the discussion of the TAP cycle. Quantitative insight is increasingly creating new opportunities to create value. Yet, most enterprises are new to the power of data, data science, and machine learning techniques. While there’s a fever pitch of hype about the potential for machine learning, there is substance behind the hysteria. We are convinced these techniques will increasingly drive innovation, and so enterprises must master them.
The IoT-Location-Insight Loop
For the last decade, industry analysts have sung the clarion call of the power of data. Fifteen years ago, Gartner Group began talking about a concept called BAM (Business Activity Monitoring). The idea was that by hooking up the key points in an organization’s value chain, management could tune the business to optimize performance. One analysis by Gartner called the year 2002 the “calm before the storm”. Now fifteen years later, we all still await the first drops of rain. But rain it will, finally, and soon.
At its core, business improvement is rooted in the idea that key levers of value can be measured and by measuring them they can be improved. But in 2002, the clouds Gartner was talking about didn’t hold data, clouds were what ruined a sunny day at the beach. Today’s concept of cloud, may finally enable Gartner’s vision of the future. But it won’t be the kind of KPI metric reporting of the past, but rather the art of artificial intelligence and machine learning that is just beginning to emerge in today’s cloud-driven enterprise. The inhibition to Gartner’s vision in the past was driven by three factors:
- Data was in silos, locked in proprietary systems, and stored on the petabytes of storage sold to enterprise in huge volumes by EMC. Saved for eternity but entirely inaccessible.
- Compute fabric was expensive and not economically capable of processing the volumes necessary to gain real insight.
- Many of the most vexing problems required judgement, perspective, and human style perception to penetrate beyond the most superficial observations.
Of course, no additional expository material is required to show that the emergence of cloud environments like Google Cloud Platform and Amazon Web Services have blown a gigantic whole it the first two once insurmountable challenge. But, what’s really changed is the speed with which machine learning techniques have become a viable solution set to these problems.
When Google made the images library in July of 2001, then Google President Eric Schmidt joked that Google Images search was created because of the desire to view Jennifer Lopez in her exotic green Versace dress. At the time, Google Images launched with 250 million indexed images. By 2005, the image count crossed the 1 billion threshold, and by 2010 that number had reached over 10 billion. Today, user’s query and select images on the site billions of times per day and every time someone enters a query (why does everyone always use cat examples?), Google’s algorithms for classifying images with artificial intelligence gets a little better and its been become incredibly detailed and accurate. At some level, this is replicating the judgement and intuition of a human being. In short, it isn’t metrics and quantitative insight, it’s a machine emulating human vision and judgement.
Now imagine a widget factory somewhere in the industrial midwest–ABC Widget Inc. A great firm with plants all over the world that cranks out millions of widgets per day and at the end of every line in every factory around the world where the shiny new widgets pop of the assembly, an inspector visually inspects the product for quality. Her practiced eye, traces the length of cylinder looking for flaws among the die cuts and flanges. Training a new guy takes months and even on a good day, sometimes they miss a bad one.
One day, Marty Machine Learning shows up with a camera and a computer and begins shooting pictures of each widget and tracking the decisions a human inspector makes about quality (this one’s fine, here’s a flaw). Now multiply that by every line, by every factory making a similar product or using a similar process. Just like Google trained its machines to recognize cats and red balls, ABC Widget can use the same visual techniques to both eliminate cost (sorry Ms. inspector, but it’s the march of progress) and improve quality (no late night parties, to blur the vision when the shift starts at 7am).
When Gartner began the discussion of BAM in 2002, they may have imagined this scenario, but I doubt it. Today, with the tools and technologies available this solution is entirely possible. But it isn’t just about a whole new way of thinking about quantitative insight, it’s also about what knowing location can give in this sort of IoT context. To quote the old SNL character Rosanne Rosannadanna (played by Gilda Radner in an early and semi-famous skit), “everywhere you go, there you are.” Everything is somewhere, and location in this new cloud-based world can give deep insight and context to the decision making driven by quantitative insight.
Take our ABC Widget Inc. example from above. Every machine in the line that made that widget generates motion data. Clamps open and close, dies turn and cut, and parts are moved down the line. The air in the factory is a certain temperature, the humidity is higher in India, and lower in North China. Just like the analytics infrastructure can learn to tell a siamese from a tabby, it can learn what conditions favor higher quality, what conditions don’t.
The Digital Innovation Lifecycle is a new way of thinking about solving business problems, generating growth, and innovating. It requires that the business, technology, and design of the services and solutions corporations provide customers, employees, and vendors combine all of the disciplines of the corporation in new and innovative ways. The right hand side of the TAP cycle encourages leadership to think creatively about how emerging quantitative insight techniques coupled with location and IoT data can enable new opportunities to create value. Even organizations that make old style industrial products, increasingly have to think of critical business process both in terms of atoms and bits.
This part of the TAP loop in the Digital Innovation Lifecycle emphasizes new techniques that take advantage of the sheer power, scale, and capacity of the cloud to solve human cognition problems previously thought to be the sole province of humankind. But it’s not your mother’s business model anymore and movements like Industry 4.0 are driving the world’s leading companies to embrace new approaches that can be more readily solved using the techniques encouraged by the Digital Innovation Lifecycle.
In our final post in this series on the Digital Innovation Lifecycle we’ll turn to a discussion of the baseball diamond of topics on the right of the Digital Innovation Lifecycle–the “doing and creating” (DAC) domain. In this loop, we have the strategy, we’re armed with the insight, and now we integrate product management concepts with agility, experience design, and change management to make the power of digital innovation apparent to the customer, the market, and investors.