I gave the following presentation at the Open Group Conference in Sydney yesterday afternoon. It is about how Cloud Computing can be used as a tool for innovation in organisations of any size as it is the perfect way to tinker with ideas and create prototypes. All the details are below, and the slides are available on Speakerdeck.
I would like to talk about how organisations of all sizes can use Cloud Computing as a tool for innovation.
I draw from my experience as developer of applications that run on the Cloud, being familiar with the power and agility that this development model has given me. I also draw from my work as a lecturer of innovation and technology management where I taught students some of the tools and theories of innovation.
It is not often that a “game changer” comes along in an industry with the power to shift ingrained practices and assumptions. I believe that Cloud Computing is such a game changer. It is a game changer not only for the IT industry, but one that will influence the way that business works in general.
And one of the specific areas Cloud Computing is changing is that of innovation.
You may have heard of the expression “standing on the shoulders of giants.” Wikipedia interprets it as “One who develops future intellectual pursuits by understanding and building on the research and works created by notable thinkers of the past”.
As a Cloud developer, I really do feel that I am building on the shoulders of giants. These giants are the creators of the infrastructure that I use in my applications to make it possible for me, a one-man operator who can only afford a rather small percentage of his time to programming, to build applications with features and functionality that only a few years ago would require a sizeable developer team, with significant resources (money and time). Cloud Computing makes me innovative as much as it is making me productive. It gives me the tools to build applications that solve real problems quickly.
But big companies use the Cloud too. Cases of Cloud Computing used by pharmaceutical companies doing new drug research or marketing companies doing market research using data mining and big data techniques are plentiful.
But here I want to focus on the individual, or at most the small team of people working on problems.
As Cloud computing has released people from the tyranny of basic infrastructure management and turned many of the basic app functionalities into components, we have been free to invest our time and effort into exploring possibilities. This freedom, as I will explain, sits at the core of the innovation process.
The story of the Post-it Note is well known, so I will not repeat it. I only wish to refer to it in order to ask a, mostly rhetorical, question:
If everyone knows how the post-it note was developed, and admires it, how come so few organisations even try to replicate the ‘method’?
The fact that the adhesive used in Post-it notes was invented largely by accident is a critical part of the method of invention. The method of invention, includes elements of tinkering, serendipity, freedom, imagination, technical capability, and the right tools, in proportions that vary widely. Invention is a messy, hard to manage, business.
Cloud Computing can be used as a tool for, first, invention, and then for innovation, so that companies of any size can innovate like 3M.
Innovation vs Invention
Some definitions first.
Innovation is the market exploitation of an invention. Invention is the activity of solving a real world technical problem. Usually, the deliverable of an “act of invention” is a technical contraption (mechanical, electronic/electrical, software).
As new products often start their life as inventions, it follows that an organisation that does not innovate can only, at best, achieve growth consistent with the growth of the population in its market by continuing to sell existing products. Such organisations rely on solutions to problems of the past that may not be relevant (or as relevant) in the present. Think Polaroid. Think Kodak.
If one considers that the market is fiercely competitive, then even population growth-level business growth is optimistic. Population growth in Australia last year was 1.43% (World Bank data). Not good enough!
So, invention fuels innovation. But what fuels invention?
That would be creativity.
Creativity fuels invention because creativity is the capacity of a person to perceive/understand a problem and propose solutions. These solutions are then explored as part of the invention process.
The good news is that creativity is spread evenly and widely among us. Although there are variances in terms of the so called ability to be creative, research shows that people as a whole are in essence problem solvers. I have yet to find a person who is 0% creative (and alive at the same time). I have met people who’s creativity has blown me away in shame for my own inadequacy. The vast majority of people I have worked with, especially as a lecturer, have shown a clear desire and capability towards solving problems.
My anecdotal sample comes in the form of students (thousands over the last 10 years), from all parts of the world, representing many different age groups, ethnic backgrounds, languages, educational backgrounds, and the like.
Perception, intellection, expression
The bad news is that creativity is constrained.
David Owens, author of “Creative People Must Be Stopped”, a very creative title for a book I must say, lists three components of creativity: Perception, Intellection, and Expression.
Perception is about the ability of a person to capture raw data. Raw data provides the basis of new ideas.
Intellection is about the ability of a person to think.
And Expression is about the ability of a person to externalise (express) their thoughts.
Skills in all three components can be learned, as research shows, and can be developed through practice. Learning happens throughout a person’s life, but the first 20-30 years are extremely important as these are the formative years. The expression that “schools kill creativity” has a lot of truth in it. Often, creativity goes to school to die as that is where the first creativity-destroying constraints are introduced. I am an insider, I have seen the destruction!
Perception is enhanced when people know where and how to find high-quality raw data/information in general, and how to filter out the non-sense (something increasingly important and critical in the age of Twitter and Facebook).
Thinking, the process of analysis and design, becomes extremely powerful when a person uses appropriate cognitive models and analysis tools with which they can make sense of the world.
And the ability to speak out our thoughts clearly in order to be understood make it possible for our ideas to be adopted by others, and accepted.
But even when people leave school and university relatively unscathed (this happens from time to time), chances are that their natural creative forces will have to be reckoned by “normal” work life in organisations optimised for efficiency (similarly to schools and Universities). The constraints imposed to the creative mind by bureaucracy can be devastating.
The orthodoxy is that in order to improve efficiency, perception, intellection and some times even expression must be limited to the absolute necessary. Creativity is naturally wasteful and inefficient, so organisations optimised for efficiency are by nature unable to let their people be creative. People are taught to do things by the process, by the book.
Bruce Nussbaum, Professor at Parsons, The New School of Design and author of Creative Intelligence: Harnessing the Power to Create, Connect, and Inspire, notes (quoting a National Science Foundation report) that in the US, only 9% of public and private companies are doing any kind of serious innovation.
What a shame, and what a waste. Organisations, by design, seem to be switching their innovation engine off and only use a limited spectrum of their capabilities. Or, they may introduce an R&D team, tell them their job is to innovate, and report to shareholders of their new brilliant initiative while alienating the rest of their workforce.
Let’s switch from innovation theory to innovation practice.
At the heart of the invention process is prototyping. Prototyping is the construction of a working solution (a machine, software app, etc.), which people can use but which only has a limited set of features and capabilities.
The usefulness of prototyping in the invention process is that it provides a mechanism for validation of an idea for a solution to a problem. In organisations, resistance to innovation is not so much about squashing ideation (the process of generating ideas), but in the cost and effort necessary in actually pursuing an idea. When an employee mentions to their manager an idea for reducing, for example, the processing time for a loan application, the manager may say “that’s a great idea, but we can’t commit the resources to do it”. When the cost of testing an idea is high, ideas most often are left unexplored. The ideator will be left feeling unappreciated. And the organisation will be left without a potentially important improvement in operations and products.
Cloud computing is a way to radically reduce prototyping costs for particular types of problems.
A component of prototyping is tinkering. Tinkering is the activity of introducing small changes to a device (software, hardware). Prototyping is often the result of lot’s of tinkering. Kids, geeks, and makers are all proud tinkerers. Steve Wozniak is one of the most famous of them.
Tinkering allows people to test the workings of their ideas rapidly. Tinkering does not test the validity of the idea itself, this is something in which prototypes help the most. Tinkering is there to gradually build the prototype. It helps ensure that (1) the maker gradually learns the technology they are using to build the prototype, and (2) that the various functional components of the prototype work.
Agility is the capacity of a person or organisation to change direction readily with the arrival of new information. Small and inexpensive prototypes can return validation or falsification information. This can be used as the basis for an adjustment of the original hypothesis and solution.
There are two major problems with expensive R&D and prototyping in organisations (which really narrows down to one):
1) Committed resources. As money, people, and effort are being committed to a development project, the resistance to change increases.
2) As a manager’s reputation is more and more invested in the success of a project proceeding “according to plan”, the resistance to change increases.
Problem two can be seen as a spin-off to problem one. Therefore, because human nature is such that admitting error and loosing face is culturally not acceptable, the more one has to loose the less likely it is they will assess the evidence objectively and take corrective action.
The software agile development methodologies movement realises this and advocates rapid prototyping techniques that actually expect and encourage change. They make change easy because they make it cheap.
Cloud Computing is a technology that treats computing as a commodity. It focuses in making the deliverable of computing, the computation, easy. By easy, I mean that the basic premise of Cloud computing is that it hides the complexities of building and maintaining the infrastructure of computing so that developers can focus on developing the application.
In the last five years or so, computing clouds have revolutionised the way applications are built and consumed. The majority of applications we use on the web or on our smart phones, are powered by the Cloud.
Renting vs Buying
The difference between Cloud and traditional computing is similar to that between renting a hotel room just for the days of this conference, or buying a house in Sydney.
In Cloud Computing, computing services can be rented. In some cases, by the minute. In traditional computing, if one needed computing resources, they had to buy a server and host it in-house.
Not only that, but in traditional computing, knowledge for the installation, maintenance and operation of a big part of the software and hardware stack needed to be available in-house too. In traditional computing, an organisation must pay the full cost of the hardware and the salaries of the people to maintain it upfront, before they even get to run an application on it.
Cloud computing eliminates all this overhead, allowing people to focus on what is most important for them, that is building applications. As a result, the impact of Cloud Computing in terms of cost and productivity is revolutionary. A KPMG study in 2012 shows that just in Australia, the impact of Cloud Computing in the economy is an increase in “long-run GDP after 10 years of $3.32 billion per annum at 75% of Information and Communication Technologies spending towards Cloud services by 2022”.
“Modelling the Economic Impact of Cloud Computing, which was commissioned by the Australian Information Industry Association (AIIA) and launched by Senator the Hon. Stephen Conroy today, estimates that adoption of Cloud services across 75 percent of ICT spending would result in an increase in long-run GDP after 10 years of $3.32 billion per annum1. At 50 percent adoption levels, the GDP gain is $2.16 billion per annum.” According to a KPMG study, the banking sector is by far the largest adopter of Cloud Computing in Australia. Considering that in terms of local infrastructure we are behind the US and Europe, Australia is still at the early stages of adoption of Cloud computing.
Case: MyApp on the Cloud
As an example of tinkering, and prototyping, I will take a few minutes to discuss a real application that I built on my own over the last few months. I will just call it “MyApp”.
MyApp is an application which I built to solve a specific problem I had. It is not a revolutionary product, and it is not solving a unique problem. Other applications exist that do solve this problem, but I had the desire to solve it in a new way. The innovation is in the approach. Old problem, new solution.
I analysed MyApp and determined the basic components that it would have to combine. The logic (how it works internally) is unique, however many other bits and pieces can be rented.
First of all, the platform. I am not a systems administrator. I am a developer and want to focus on my own problem, directly related to the thing I want to build. So, I chose an infrastructure provider called Heroku. Essentially, Heroku is my systems administrator. I pay them $15 per month. I get the computing resources on which my application runs, and their expertise for keeping it running.
From the outset, I needed a way to to communicate with my users via email. I could build my own email system, with public-domain technologies like STMP, but this is not my problem. I should not have to reinvent the wheel, especially when others have done a great job in creating exactly what I need. So I rent email services (ingoing and outgoing) from Mailgun for $19 per month.
Then, I needed a way to offer online search capability. I started tinkering with my own search logic, but found that things were becoming more complicated than I thought they should be. After all, I am not the first person to need search for their application. And surely enough, somebody had done a great job in creating a search component I can integrate in my app. So I attached SearchBox to my app. Cost: $0 per month (they are in beta at the moment).
I also needed things like scheduling, performance analytics, access analytics, scaling monitoring, a database, a payment API, exception reporting, OCR processing, storage, heartbeat monitoring, stress testing, and a couple of other things.
An app I build just before “My App”, called “My Other App”, looked like this. A completely different application in fact, using largely the same components. No component or infrastructure reinvented, just a focus of the application logic itself.
Case: Cloud and Traditional comparison
In terms of time, the Cloud has saved me years of man-hours. I have no objective way of calculating this other than comparing my own productivity (of useful features) using the Cloud development model vs traditional.
In terms of money, it is clearer. To build MyApp in the traditional way, I would have to start by either repurposing one of my old computers as a server, and upgrading storage and memory, or renting one from a hosting company. I calculated that the electricity alone required to run my own basic server would cost around $5 per month.
Is innovation at the enterprise level any different to innovation at any other level?
No, the ingredients are the same. Thinking freedom, time to tinker and play, lot’s of prototyping, lot’s of testing. What is different is the level of bureaucratic control (governance) required for approving initiatives. Typically, organisations give a limited amount of discretion to departmental managers and to some key employees to deviate from approved processes on occasion. However, the process of innovation is too ad-hock and unruly to manage in a traditional organisational environment. Tools such as portfolio management and QFD (quality function deployment) do exist and can be effective, however they do not really address any of the issues of grass-roots innovation. Without grass-roots innovation, organisations will not be able to produce a critical mass of projects that can graduate from the tinkerer’s scrapbook to the portfolio manager’s spreadsheet.
And that is a shame, because an organisation that does not innovate, becomes solely dependant on products created for the past, or for the near future. If the organisation has money in the bank, it might purchase a small startup to look innovative, like Yahoo did.
Agility is the ability of an organisation to change itself in the light of new evidence that change is required. Competition, new technology, new Government legislation, etc, all from time to time generate conditions such that they require oranisations to change.
The OODA loop, conceived by US military strategist John Boyd, describes how decision making takes place as a series of Observe – Orient – Decide – Act cycles. Whoever can complete these cycles more effectively and faster, has a significant advantage over enemies or competitors.
Enterprises find that the Act step is the hardest to complete. How many times have you heard this phrase from a manager: “I know that’s what we have to do, but we simply don’t have the resources right now.”
Without Acting, the loop is broken, and the next set of observations, in which the organisation can actually influence the playing field, will not be made. Without acting, the organisation is unable to influence it’s own future.
I am not saying that organisations in general don’t act, this is obviously wrong. What I am saying is that organisations in general only act in a centrally organised manner, with organised R&D, budgets and processes, i.e. top-down. Moreover, most organisations usually only react to external events and show little initiative. This way, organisations can only explore a limited set of options and possible futures and give up the vast set of possible futures that can only be discovered if the threshold for acting is reduced significantly by allowing a lot more people to take part in OODA.
A major obstacle for making such acting freedom possible is cost. A major reason for organisations wanting centrally allocated R&D budgets and processes is to ensure that shareholder capital is not wasted. As I have said earlier, Cloud Computing represents an opportunity to significantly minimise such costs and an opportunity to significantly lower tinkering and low (bottom)-level acting costs.
This is what I have tried to do in the way I go about building my own apps, and what trully innovative organisations do day in and day out.
Another primary reason for centrally controlled R&D budgets and processes is an aversion to failure. Failure is of course tainted as something to be avoided in most cultures (the startup culture is the only one I know of where failure is a badge of honour). In the days of traditional computing, failure was an expensive proposition that would affect shareholder capital and the careers of those involved. Extra caution in such cases makes good sense.
But failure can be seen as just a step in rapid prototype development and an opportunity to learn from flaws in our designs and assumptions, and then correct them. The central motto in the Lean Startup movement is “Fail Fast and Fail Cheap”, which means that because failure is only part of the innovation/entrepreneurial process it should be neither avoided nor delayed.
Imagine an exited and focused employee doing an experiment for a proof of concept on a Friday afternoon, in which they utilise an Amazon virtual machine, a simple database, a queuing service, storage and workflow engine to test an idea. A simple proof-of-concept can be build in a couple of hours, and assuming that all the Amazon resources used are within their free tier, the total cost of these resources will be zero. If the experiment fails, the total cost total to the organisation will be the time of the employee. However the organisation now has both a more knowledgable (and happy) employee, and has eliminated a dead-end while possibly collecting actionable information that can be used in the next prototyping iteration.
To finish, I wish to talk about people.
It is common for organisations to refer to employees as their most important assets. Much emphasis goes to things such as compensation, sick leave, maternity leave and the like. All these are important, and also fairly easy to manage because they are quantitative and measurable.
However, people react positively to those things only in the short term, like when they are about to sign a new contract or accept a raise. In most cases, the traditional tools of human resource management are quick to fade into the background of everyday life in the office. The everyday reality, then, become THE reality with which the employee has to deal. And in many cases this reality is void of discovery, risk taking, goal setting and seeking, and learning.
In effect, a substantial part of human psychology that has to do with the need to discover strange new worlds and be useful beyond the mundane is left completely unaccounted for. The outcome of this is boredom. In psychology, boredom is defined as “an unpleasant, transient affective state in which the individual feels a pervasive lack of interest in and difficulty concentrating on the current activity.” (Fisher 1993, p. 396). Some psychologists explain that boredom in organisations can be more dangerous than overworking employees.
Curt W. Coffman, global practice leader at the Gallup Organization, says:
“We know that 55 percent of all U.S. employees are not engaged at work. They are basically in a holding pattern. They feel like their capabilities aren’t being tapped into and utilized and therefore, they really don’t have a psychological connection to the organization,” (http://www.washingtonpost.com/wp-dyn/content/article/2005/08/09/AR2005080901395.html)
Do you want employees like that?
Boredom results in luck of engagement. Luck of engagement leads to employees who are not motivated. It all becomes a mess for both the employee and for the organisation. I suppose, there is an argument to be made for companies allowing employees to use Facebook at work. It helps break the boredom and it might actually improve productivity!
Cloud computing is a perfect way for the creativity of countless employees to focus their creative forces and explore possibilities for innovation like never before. The barrier to entry has been lowered so much that not using cloud computing is most likely result of ill-information or ignorance. Although Cloud computing is not a universal tool for innovation (you need basic IT skills, and the problem you are solving needs to be compatible), it is still a tool that millions over millions of people could readily use or learn how to use with a small investment in time. Similarly for organisations, Cloud computing is not a panacea, but a tool that can help organisations to innovate from the bottom-up.
I wish everyone happy tinkering!