Convention Over Configuration in Technology

One of the classical debates in computer science is convention versus configuration. This debate manifests in programming-language design, product design, marketing, what hardware and software can and cannot do, and how they do it. 

In the early eras of computing, configuration reigned supreme. The first computers had no security, no encryption, no rules, and very little if any software infrastructure to sit on top of. The early computers were in many ways a clean slate, akin to John Locke's tabula rasa. Not only did the software epitomize configuration, so did the hardware. Before the Apple II, the only people who could acquire a personal computer were those who knew where to buy all of the components and how to assemble them. Assembly often times required soldering. Talk about high barriers to entry.

Apple has risen to become the world's most profitable company because it embraces convention and eschews configuration wherever possible. Apple pioneered the first personal computer hardware that people couldn't physically open or configure. And just as importantly, Apple makes software design decisions for its customers so that they don't have to. It turns out that people want computers that "just work." They don't want to deal with hundreds of settings and options. iOS represents the pinnacle of convention over configuration in the mobile OS market, where users cannot access or manage the OS’s file system or processes in any significant manner.

For most of Apple's history, it was the loner in the industry in its pursuit of convention over configuration. Most of the rest of the industry valued openness, flexibility, and configuration. It wasn't until the mobile computing revolution of the past few years and Apple's extraordinary corresponding rise that the industry began to actively push convention over configuration. Microsoft in particular has changed its views on the issue. The transition from Windows 7 to Windows 8 marks a very large swing away from configuration towards convention. Looking forward, it appears that most consumers will compute how Apple and Microsoft allow them to.

Although Android is quite flexible and open-source, it’s quite apparent that Google is actively working to  mitigate some of the major shortcomings of configurability. The Nexus line is intended to encourage Android OEMs to converge on Google’s vision for Android.

We can see a similar shift in software development itself. Abstraction in many ways favors convention over configuration. Why should developers worry about memory management, object models, file storage, and application distribution when Apple/Google/Microsoft can handle it for them? Why should developers worry about API management when Mashery can manage APIs for them? Why should developers worry about infrastructure when Amazon can build, operate, and scale it for them?

We can see the power of convention extend into web services as well. Although the Internet is hailed as extremely democratic, capitalistic, and thusly competitive, we've seen that most web-native businesses tend towards monopoly because of network effects and economies of scale. Amazon will be the Walmart of the 21st century, and no one stands a chance of stopping them. Google has owned search in every country except those where the government actively worked against them (Russia and China). eBay owns peer-to-peer physical goods auctions in virtually all markets that it competes in. These companies are the natural convention for the general public. "I go to Google to search. I go to Amazon to buy things. I go to FaceBook to share with friends." What percentage of the general population actually uses Path or Diaspora in place of Facebook?

Google Glass, the pioneer of one of the major new computing form factors, is extremely locked down. You will use it only in manners and conventions that Google deems acceptable (this is arguably a good thing to protect privacy). As we move into the hyper-mobile era of computing in which computers are everywhere, expect convention to continue to overtake configuration in most consumer and business markets.

We've Passed The Threshold: Computers Will be Everywhere

Computers are getting smaller every year. Back in the day, computers consumed entire rooms. Today we have Glass. Glass is a complete computer that's small and light enough to rest comfortably on the human face.

But Glass isn't the only new computing form factor hitting the market this year. There is an incredible number of new form factors coming out, including:

Leap Motion

MYO ArmBand

Neuroware

Oculus Rift

Ouya

Pebble Smart Watch

We haven't ever witnessed such a blossoming of computing form factors in such a short period of time. That begs the question, why now? What made it possible for all of these new form factors to commercialize in such a narrow window of time?

The simple answer is that computers were never good enough to power these form factors until recently. Today, an entire computer, packed into an integrated system-on-a-chip (SoC), consumes just a few square centimeters (approximately the size of a fingernail). SoCs are cheap, affordable, capable, and available in relatively small volumes (< 50,000), with availability to scale up to extremely large volumes (> 100M). Computers are small enough to quite literally fit everywhere, so entrepreneurs are trying to fit them anywhere they can.

But hardware is only one piece of the puzzle. In order to compute, these new devices need an operating system and connectivity to web services. Luckily enough, Android is available as a free, open source, extremely robust, extensible OS that's evangelized by the almighty Google and has a thriving developer community. It has been and will be the foundation for a large percentage of the new computing form factors.

And lastly, these new devices depend on high speed, ubiquitous Internet connectivity. Most of them will feature hosted app-stores, where all applications are hosted and delivered through the Internet. The Internet is the greatest distribution mechanism of all time. It's hard to compete with distribution that's global, instant, and on a marginal basis, effectively free.

The foundational elements are in place. The stars have aligned. Computers will be everywhere, in everything, all the time, so that we can compute whenever and however we want. We are witnessing an explosion of innovation in computing form factors. Incumbents will dominate some form factors such as smartphones, but startups will pioneer and lead many new computing form factors, especially those that aren't expected to be mass-market devices. Incumbents will focus on the largest and most profitable businesses, leaving smaller markets up for grabs. For the first time in a long time, we're experiencing a renaissance in hardware. It's refreshing and exciting, and is creating enormous new opportunities for hardware, software, and services startups.

Are Hospital EHR Vendors Primarily Software or Services Companies?

This post was originally featured on HIStalk.

Hospital EHR vendors are not primarily software companies with professional services divisions. They are primarily professional service companies with software divisions.

Although the core value of hospital EHR companies is the software they develop, the bulk of the value they provide is in training, data conversion, setup, logistics, and support. These services are built on top of the software that they support, but are collectively more valuable than the software itself.

I’ve seen the challenge of deploying an EHR in a hospital first hand. Most of the challenges are with the people, not the software. Hospital EHR vendors derive most of their value for their professional services.

Hospital EHR vendors employ more services staff than software staff. At least 60 percent of the employees at the major hospital EHR vendors are in deployment and support services, not software design, development, testing, or administrative functions. Employees, especially road warriors, are expensive. The more employees involved in a given division, the more expensive and valuable that division is.

But perhaps even more important than sheer costs, healthcare IT staff need to receive sophisticated training in hospital workflows and software systems. Before the HITECH Act, there weren’t enough people with healthcare IT skills to deploy the entire country in five or six years. Many argue that there still aren’t enough people. The EHR vendors had to develop large-scale internal training programs to teach all of these people how to set up, train, deploy, and support hospitals. This is one of the greatest sources of value that the big EHR vendors have generated: an educated healthcare IT workforce. The scale and scope at which they’ve done that has been remarkable.

Epic employs 6,400 people and Cerner 11,900. I would estimate that at least 60 percent of those – or 11,000 people in just these two companies – work in training, deployment, and support roles. These companies and many others have developed phenomenally large internal training programs for their employees, who are usually fresh college graduates.

To provide a sense of scale, the University of Texas at Austin — one of the largest universities in the country by enrollment (60,000+ students) and located in a major US technology center — boasts that it has graduated a grand total of 333 people in its 9-10-week-long healthcare IT program since its inception a few years ago. Vendors are educating the workforce, not the educational system.

Training and conversion costs usually prevent hospitals from switching EHRs. These costs are multi-dimensional, spanning financial cost, employee personal costs, and opportunity cost of working through other initiatives such as Meaningful Use 2 and ICD-10.

As an employee at a smaller hospital EHR vendor, I’ve experienced this phenomenon. We are trying to spearhead the replacement market for hospitals that are dissatisfied with their legacy EHRs. Most of them love our product and are even willing to pay, but aren’t willing to change systems because the non-financial costs of change are too great for the organization.

Because the costs of switching are high, the cost of choosing the wrong EHR the first time is even higher. Most large software projects that fail do so because of the people, processes, and cultures, not because the software isn’t capable. In that sense, the services surrounding the software implementation are even more important than the software itself. The majority of the value that the vendors provide is in services, not software.

Looking at hospital EHR vendors as service companies can help understand management decisions that may not have made sense when looking at them as software companies. Decisions are made based on training and deployment realities, not software limitations. This analytical framework can also help explain vendor practices and methodologies (especially hiring), release cycles, growth rates, stability, and many other operating metrics.

Proteus Will Use Social Pressure to Cut Healthcare Costs

There are plenty of theories about how we can go about cutting the healthcare cost curve. Most involve changing incentives, using more health IT for analytics, and using mid levels and other non-mss for an increasing amount of care. They may work, they may not. But they are all most certainly expensive and difficult to implement.

Proteus Digital Health just cut the cost curve with an beautifully simple invention: a pill that sends text message alerts when its ingested. This breakthrough can save hundreds of thousands of dollars for an untold number of patients. This is a groundbreaking invention.

Why will such a simple invention be so effective?

Because of the power of social pressure. It's been proven in study after study that social pressure is among the most powerful motivators for people to accomplish goals. I advocate for the methodology. I adopted it when I lost 80 pounds of fat and built 30 pounds of muscle over the course of 18 months in college. It works.

It's widely known that medication adherence is a big problem. significant portion of patients simply don't take their medications when they're supposed to. It has been nearly impossible for family members to monitor if and when children and the elderly take their medications. Proteus solves this problem by notifying family members if the patient has or hasn't taken their medication on schedule. The family can act accordingly and hold the patient more accountable.

I'm particularly fond of this technology because it's so simple and inexpensive, yet incredibly effective. It delivers an amazing ROI. Not only is it cheap, it requires virtually no cognitive overhead of any kind. Proteus is a brain-free technology. It's idiot-proof. Although it's difficult to forecast, this technology could in time save thousands of lives and tens of millions of dollars annually through avoided hospitalizations.

The Defining Characteristic of Eyeware Computers: Transparency

Google Glass won't be the only eyeware computer. There will be many others, including:

Contact lens computing (not real yet)

Meta-View (Like contact lenses, but in glasses form instead)

Oculus Rift (complete virtual reality headset marketed to gamers)

None of these devices directly compete with one another or Glass because very few of the applications for any of these devices would work on the others. Each device provides a totally different experience.

Although these devices differ in many ways, there is one key attribute that will define how the device is used: transparency. Why is transparency so crucial to adoption? Because humans need to maintain eye to eye contact for basic communication and interaction. It's incredibly de-humanizing to talk to something that looks and acts like a human, but that has robotic/android eyes. The eyes inject life into the human face. People will not stand to interact with those wearing a device that covers the eyes unless it's understood that the device is paramount to the task at hand. Transparency and the ability to maintain clear, direct eye to eye contact will dictate how eyeware computing devices are used.

Through the lens of transparency, let's walk through the transparency of each of the eye computing products that have already been discussed.

The contact lens solution is by far the most transparent. As the video demonstrates, contact lens computing could impact literally every aspect our lives. Contact lens computing presents the logical extreme of frictionless, transparent, and ultimately "real" computing. Unfortunately, this is probably still a ways out.

Next is Google Glass. Glass was designed from its inception to be, as Google phrased it, "there when you need it, and out of sight when you don't." Google is envisioning a future in which everyone can quickly access the information they want when they want. In order to accommodate that, Glass must be acceptable in public. Google has gone to great lengths to make Glass as invisible as possible.

Next is Meta-View. The creators of Meta-View share the same vision as the products of the contact lens video: "real" computing. Meta-View is simply bound by current technology limits. Unfortunately, that means no one will communicate with anyone wearing the device. Even still, it's amazing to see someone working on this idea.

Oculus Rift is at the far end of the transparency spectrum. It's beyond opaque. It places the user in an engrossing virtual reality in place of actual reality. Of course, this will never reach high usage mainstream adoption, but it will be incredible for the hardcore gaming market, training, and other enterprise applications. Perhaps the masses may use an Oculus Rift-like device as an occasional novelty. There could be opportunities for a movie theatre type model where people go to venues to use an Oculus Rift like technology for a short period of time.

Eyeware computing will manifest in many unique form factors. Each will have unique characteristics designed for unique applications and experiences. For at least 5 of the next 10 years, eyeware computing will experience an era of hyper-innovation, just as multitouch glass smartphones have over the past 5-6 years. This is going to be a hot industry for a long time to come.