Sometimes innovation comes about by transferring ideas between domains. Picasso is supposed to have said “Good artists copy, great artists steal”, but what he actually said was “When there's anything to steal, I steal” – which I think is much more useful. It goes along with not reinventing the wheel. Maybe the guy next door has a better wheel you can steal.
If you're advocating data standards or managing their implementation, you could steal the word “Story” from the health standards community. The Health Story project has been running for almost a decade now and has produced several important data standards which improve interoperability across the healthcare system. HIMSS, who look after the project, are working toward “Comprehensive electronic records that tell a patient's complete health story.” This goal has analogs in every area of endeavor.
Are you in a retail business? Then you're dealing with the story widely known as the customer journey. If you're in industry, your story is likely to be about the supply chain. In government, you'd be concerned with the rights and needs of the citizen – the story of everybody's life.
Thinking in terms of stories helps people appreciate the systemic nature of today's challenges. Stories help to make complexity manageable and to keep goals in sight. Above all, stories humanize activities that can all too easily become mechanical, hidebound, and inflexible processes. Health Story
The indemnification provisions of the Dodd-Frank Act have been removed. Rejoice! Okay, so this is an obscure piece of tweaking to a law governing the conduct of financial organizations and I guess it's not going to inspire many all-night celebrations. But it's a sign of a shift in thinking about data that shows how leaders are getting the importance of data sharing.
The provisions which have been annulled “effectively blocked the sharing of over-the-counter (OTC) derivatives data among regulators and complicated efforts to improve market transparency and mitigate systemic risk”. Where previously people thought walls were automatically a good thing, they're now seeing how artificial barriers increase risk for the whole community while obstructing oversight.
If even the government has got this message, there's no place for enterprises to hide. Sharing is the new default. In turn, that makes using industry data standards a no-brainer. DTCC
One area of the IoT that's getting traction is the world of consumer retail. The buzz phrase is Micro-electromechanical systems (MEMS) – which, as far as I can tell, are sensors. Retailers are looking at the sensor data captured in baseball and thinking how it can be applied to the shopping experience. Baseball has always been a great producer and consumer of data, so it's a natural leader in this area. They track items like a player's eye movements or a team's coordinated response to a fly ball. The idea is to use similar ideas in stores so as to better understand consumer behavior.
But in a familiar pattern, the early promise is already being obstructed by the old enemy, the lack of data standards. According to Josh Knauer, “1) Most sensor and MEMS data is proprietary; 2) The data lives in database silos and doesn’t come in one format; 3) Siloed data in proprietary formats can’t just be exported into a spreadsheet and compared easily”. And this means retailers are not making the progress they want.
Knauer's company, Rhiza, is in the business of helping device managers package and monetize their data. They're in a good position to advocate standards. It would be good to hear from the major retailers too, who must want to avoid proprietary lock-in.
Who are the data management leaders of the future? As I look toward my retirement from the role of President and CEO of ACORD, I'm optimistic about the future for data as a resource. ACORD has always been a strong voice for the value of data, the importance of sharing data, and the competences needed for good data management. I think the next generation of leaders will not need the same level of education about the benefits of data. But they may need more help in understanding the obstacles in the way of success.
It's a conundrum. Before, people in management were often hazy about data and not always able to appreciate the asset nature of data. But they were very aware of the difficulty of coordinating activities in a complex environment, and the challenges of managing change. Now things have switched around. People are comfortable with data because they have been interacting with it, producing it, and consuming it, all their lives. And maybe it's because they have a close affinity with the world of data – a realm of smooth interaction, instant results, and intuitive progress – that they are less aware of the issues around human processes. By which I mean, getting things done with, by, and for, people.
So I'm generalizing widely here. I'm drawing a caricature. The picture I'm drawing is a product more of industry comment and anecdotal evidence – maybe even prejudice – than science. But that's okay, because what I'm concerned with is the way the rhetoric is running. Received wisdom has it that the leaders coming up are technology savvy but organizationally limited. If there's any truth to this, then it's our duty to do something about it.
I don't mean we should spray cynicism all over everybody – tell them how difficult it is to get anything done “in the real world”. I'm all for simplifying processes, reducing bureaucracy, spreading empowerment. Organizations need to change in order to deal with change. All I'm saying is we need to look for leaders who don't just appreciate the value of data, but who are also people people. And this is a very positive message, because there are many more of these balanced individuals around then there used to be.
Metabolomics is yet another critical area of activity which is crying out for data standards. What is metabolomics? It's the “systematic study of the unique chemical fingerprints that specific cellular processes leave behind” – so it's very important for understanding cell biology and diseases including cancer.
The situation regarding data in metabolomics mirrors so many domains. Vital need for sharing data? Check. Huge growth in data volumes? Check. Problems with data diversity? Check. And get this analog to regulated commercial activities: “Reporting recommendations in metabolomics started to emerge about a decade ago and were mostly concerned with inventories of the information that had to be reported in the literature for consistency.” In other words, the only data analysis and design work done was in response to mandatory third party requests, not the primary mission of the data creators and users themselves.
It's becoming a challenge to identify any worthwhile sphere of human endeavor which does not require industry data standards in order to function correctly. I'm finding it hard to think of any significant (legal) business process that doesn't include some requirement for machine-readable data at some point. And wherever more than one person or organization needs to connect with that data, industry standards are logically implied. Standards can no longer be spun as a choice. They are a necessity. Wikipedia
GovLoop, a social network for people in federal, state, and local government, ran a neat piece on the benefits of the DATA Act. I'll summarize their points in an even shorter format: The benefits are better accountability, better management, and better compliance.
These all sound great to me. While people are right to focus on innovation, we must remember there's always a need to improve the mechanics of how organizations work, and how they serve their stakeholders. It used to be called management science – I guess it still is, but nobody seems to use that label much today. You've got to pay attention to accountability, management, and compliance – because if you don't, you don't get to play any more. Doing well means doing better. Always did, always will. GOVLOOP
Gotta love the 80/20 rule. It's a universal phenomenon that applies in every aspect of data management and exploitation. Here's an instance from a data scientist working in housing: “Anecdotally, a data scientist will spend 80% of his or her time cleaning a data set.” This is tragic. The valuable time and skills of motivated professionals are being squandered on dumb, repetitive tasks that could be removed at a stroke – with industry data standards.
Whenever we see people in this kind of data mess, we don't need to look far for either the cause, or the solution. Every community that suffers needless data complexity is generating its own problem, which it can cure with data standards. Let's stop wasting people lives with thoughtless data creation habits. This is a business hygiene issue. Down with dirty data! HACT
When is the right time to adopt industry data standards? Reading some robust remarks by Ross Koppel on electronic medical record (EMR) systems, I'm tempted to answer: Yesterday. Koppel says the health industry urgently needs data standards to serve healthcare professionals better. Existing systems were built with billing rather than clinical needs in mind. “Now the existing structures are so embedded that it’s very costly to reach the kind of standards that are needed to allow the interoperability that is the hallmark of a reasonable system,” he says.
Note the implication here – the current ecosystem is unreasonable. It doesn't make sense, and we shouldn't stand for it. No one can travel in time. But Koppel has a shortlist of immediate actions that could be taken to move the industry ahead in priority areas. He says his ideas could be implemented in six weeks. And he says: “We’ve been facing the same bloody problem now for 15 years. It’s enough. We just have to deal with it.”
He's right. It doesn't matter how long something suboptimal has been the case. If it's wrong, get rid of it. It doesn't matter how hard it's been in the past to create standards. Standards are still absolutely required. Past difficulties should act as signposts to prioritization exercises, such as the one Koppel has performed.
And it doesn't matter how much anyone has invested (monetarily or emotionally) in the existing set-up. Spending more isn't going to trigger a reversal in values – it's just going to compound the problem and further delay the inevitable.
The article citing Koppel is titled “Overcoming the Madness of Flawed EMRs”. Not using industry data standards is a form of madness. It's unreason. I see more and more evidence that people will no longer tolerate the unnecessary expense and limitations created by non-standard data environments. They're right to call out the madness – and to get mad. WTNnews
My latest book presents the challenges members face when adopting industry standards as well as the opportunities that come as a result. It features my discussions with many people over many years and follows the foundation I set in my first book "The Business Information Revolution".
Industry standards are never adopted in a vacuum. They become part and parcel of all the trials and tribulations managers face in their day to day work. ACORD Standards are always part of a larger software development project that brings along people, priorities and politics. Adopting industry standards isn't simple, but the benefits far outweigh the problems of building and maintaining proprietary alternatives.
I trust that you will appreciate my frankness, identify with some of the challenges and learn from what others have done to pave the way.
Previous Book
This is a PDF version of my book. You have my permission to view, save and print copies for your personal use. Use your browser "Back" button to return to the blog after you visit or print a chapter. If you want a clean copy, it's available at the Amazon bookstore.
