Gregory Maciag

Believe the Device

Smart devices are producing new data streams which will have profound consequences on the way we manage the environment, our health – and, of course, our businesses. But the powerful devices around us and in our pockets can also provide significant benefits simply through their status as reliable witnesses. It's not just the data that matters, but its source.

The Quantified Self site led me to a story in the Atlantic about a lady whose face froze – sending her to the Emergency Room, where they couldn't find anything wrong with her. A couple days later it happened again. She filmed the attack with her smartphone, and showed it to the doctors. They ran full tests and discovered she'd had a mini-stroke.  Serious stuff. Significantly, the professionals accorded the movie more credibility than the patient's original verbal account. Technology can act as an advocate as well as just an assistant.   The Atlantic

Data Standards for All Data

Making her predictions for the healthcare data area in 2015, Pam Baker makes an intriguing distinction. Her first prediction – about the growth of intelligent wearables, especially in the hospital setting – is acute but not contentious. I'm more concerned about the apparent divergence of the views given in Predictions 2 and 3.

Prediction 2 says that organizations will begin to address their problems with interoperability and data flow by using workarounds. And Prediction 3 is that healthcare data standards will improve – for big data.

I don't get why standards are supposed to be limited to big data only. Remember, big data only means “lots of data from all over the place”. Structured as well as unstructured. But it's still data.

The perspective of the article seems to be that big data is all about retrospective analysis. It's what we used to call Management Information (MI). So data standards are needed here in order to produce good analyses.

But operational data also flows better when standards are used. Why give the okay to workarounds at the cutting edge of the business while seeking standardization at the planning level? This is to neglect the huge power of data standards right across the organization and beyond – to partners and patients. It's also not in the spirit of real-time business, where resources and actions are managed in response to actual demand.   NUVIUN

NetVU Recognition

I was surprised yet very grateful for the recognition by the NetVU organization at the annual conference tonight. Thanks to Debbie Miner, Keith Savino, Kitty Ambers and the entire NetVU organization for supporting ACORD for all these years.

 

 

After Fannie and Freddie

Fascinating keynote remarks on the future beyond Fannie Mae and Freddie Mac by Edward J. DeMarco of the Milken Institute, speaking to the American Enterprise Institute. DeMarco says that both institutions are technically ready to exit government control. In fact, housing finance has had a substantial operational makeover, with industry data standards readying the sector for a bright future:

“For example, much progress has been made in developing data standards, both with regard to data definitions and developing industry-standard technology for reporting data. This work is important because it removes a key barrier to entry inherent in the Fannie/Freddie model, namely, the proprietary data standards and systems each used. Accurate, standardized data is also a prerequisite for improved disclosures to investors, which in turn enhances investors’ ability to price and manage mortgage credit risk.”

DeMarco sees standards as “building blocks for a stronger, more resilient and more competitive secondary market in a post Fannie and Freddie world. These changes have taken some time, and they have been costly, but the returns from these investments should be large and long-term.”  The investments have been well made. Now we need to start reaping the rewards.   Milken Institute

Standards Empower Customers

The Department of Transportation has approved the use of a data standard for exchanging data in the industry. The standard comes from IATA, the international air transportation body.

Why is the airline industry so late to the standards party? Here's a clue:

“The absence of a common, up-to-date data transmission standard has meant consumers must check multiple sources to identify competing fare options and information about the comparative cost of additional ancillary services, such as extra leg room, Wi-Fi and advance boarding, that many of today’s air travelers view as critical to a satisfying trip, DOT noted.”

In short, fear of competition. Many industries have tried to hold out against consumer comparison, worried that ease of substitution will erode their margins. Well, of course it will. But this is reality.

Trying to keep customers by denying them information or obstructing their attempts to compare offers just doesn't work. They'll move to intermediaries who can regularize the information for them. These aggregators then become the trusted brands and wield immense power.   It's better to be on board with an industry standard, helping to shape it and getting early experience with implementing it, than to hold out.    Travel Agent

Meta-Data Maturity

Adrienne Tannenbaum's article on the history and future of metadata is a masterclass in the subject. The whole piece is well worth reading, but I particularly like this:

“When looking for direction and advice for your metadata strategy, follow the leaders. In the world of metadata, quite a few industries survive and thrive based only upon the automated exchange of information. Information may actually be their product/service offering, but regardless, the similarity across successful industries is that the exchange is and has been a necessity. Without timely and accurate sharing of business information across organizations - including regulatory organizations - there is no functioning business.”

That's insurance in a nutshell. Increasingly, it's every other industry too.   Bloomberg

Fixing the Fab Five

Starting May 2015, the Department of Health and Human Services (HSS) will be the location for a two-year pilot of the DATA Act. The project will test how data standardization in a complex federal ecosystem works.

HSS has already identified the key areas where different agencies use different data definitions for the same concepts. They call these the Fab Five, and they are data to do with organizations, government programs, places, time periods, and “the concept of amount”.

The department's Amy Haseltine says: “It doesn’t make any of it wrong - it just makes it all different.” This is a diplomatic and positive attitude to evolved complexity which sets a good tone for the work ahead. But I can't help thinking that “different” is the new wrong. What's meaningful and usable at the local level is often dysfunctional at the collaborative level. And the collaborative level is where all human activity is headed.

Notice how it's pretty easy to outline the main areas of data difference. I guess the good people at HSS and their contacts have been struggling for years with the problems caused by the Fab Five. This is an example of the great effectiveness of the 80/20 rule. If they can transition to a Fixed Five, they'll make major gains and make space to reveal and tackle the next tier of unnecessary complexity.   FedScoop

Google Fit

Good quotes about health and fitness data from Derek Newell, CEO of digital health care platform Jiff, in Forbes magazine: “Google Health [shuttered in 2012] never took off because consumers actually don’t want to aggregate their data […] They haven’t wanted to. What they want is information. They want meaning, rewards and a feedback loop.”

Now Google is set to launch an open API data aggregation service called Google Fit. Newell continues: “There’s lots of little players, and no data standards. [Apple, Samsung, Google] are going to step in the middle of all that and facilitate the collection of that data and let developers develop on top of it.”

He's absolutely right. Consumers don't want “data”, and they don't want “data standards”. They want information, so they can take decisions; and they want apps, so they can take action.  Data is the lifeblood that supplies all applications. Data standards provide the rules for moving, sharing, and aggregating data. Customers don't need to know about any of this stuff.

But businesses do. If you're in the health business, or insurance, fitness data could be very valuable to you. You need to care about the data standards. You want them to at least map logically to the concepts you use in the business. Ideally, you want a close mapping relationship to the data standards you already use.

For this reason, businesses have an interest in the success of open data standards for health and fitness data. This needs to be distinguished from the competitive forces working in the smartphone and lifestyle markets. Decision makers will follow the progress of Google Fit with close attention.   Health Data

Data on the Hill

Hudson Hollister of the Data Transparency Coalition continues to keep a watchful eye on open data standards, and to point the nation's decision makers in the right direction.

The Senate has said it will release machine-readable summary and bill information from the 113th Congress and legislation from the upcoming 114th. But the House and the Senate are not using the same data standards. This isn't an academic point. Hollister points out that if both House and Senate adopted the same standard, “then legislative drafters could use software tools that would automatically show the impact of what they're doing on the underlying law – automatically redlining proposed legislation into the U.S. Code.”

That's joined-up thinking.  GCN

Need for Data Governance

Some thought provoking findings from recent Forrester research suggest that in prioritizing speed of data delivery over trustworthiness, organizations are creating new bottlenecks. Instead of improving decision making, they may be degrading it.

Here are the key stats: “Overall, 42% [of respondents] spend more than 40% of their time vetting and validating the data. For executives, 70% spend more than 40% of their time vetting and validating data.”

So it's good that people are checking the utility of data instead of acting on it without question. But it would be way better if users were given pre-validated data. Especially since the vetting and validation must be being carried out multiple times by different people on the same data.

It seems we've moved from Garbage In, Garbage Out to asking people to sort through garbage. We're wasting their time. Data governance is a critical element of organizational effectiveness.   Forrester

Data Standards Create New Markets

The Google-led General Transit Feed Specification (GTFS) has produced a slew of third party apps using open transit data. This standard is the poster child for open data – and it's encouraging people to think about how other kinds of open data can be exploited.

Now a coalition of parties is setting out to create a standard for building permit data. The challenges in creating the standard will be familiar to anyone who has worked on a data standard: “The tension in this process has always been how to make the data described in the standard generic enough that it accommodates different business processes but doesn’t gloss over any detail that’s important.”

Every authority structures building permits differently. This is because, historically, there was no need to share the data. One city wouldn't need to share with another city. It's only now that citizens, and the third party software development community, are connected that the data has wider currency.

This kind of data has apparently been called “civic exhaust” in the past – a new term for me but a vivid one. Organizations of all kinds are realizing that their by-product data may be useful in new, collaborative contexts. In this case, it's not that two cities will want to exchange building permit data; it's that other organizations may be able to derive value from the data – but not if they have to write new interfaces for every city's data feed.

Business data standards are vital for removing barriers to market formation. In fact, without data standards, many market opportunities can't even be visualized. You literally cannot imagine a viable industry for third-party civic data apps without common business data standards. That's how integral standards are becoming to our economy.    GCN

Share Umbrellas

There's lots of activity in data standards for IoT (Internet of Things). Jim O'Reilly gives an overview of the various initiatives and industry groupings under way.

But: “We would be far better off if there were an umbrella organization, such as IEEE, that addressed IoT standards, rather than partisan camps. This would remove the possibility of incompatible solutions on the store shelves. More importantly, it would bring more effort to bear on solving the security issues that are not currently being addressed well by any of the groups.”

Meanwhile, elsewhere another commentator notes about IoT: “A lack of consistent standards in data or connectivity, and what appears to be a reluctance to share data, is stifling innovation.”

Players in this industry need to think carefully about the relationship between competition and collaboration. Without a certain amount of collaboration, there won't be a viable market, and therefore competition will be moot. IoT demands data sharing. It's too big a concept to be owned by one player.

InformationWeek

Short-cut or Highway

Some developers resist standards because they think using standards is a short cut. Taking short cuts is unprofessional, and possibly dangerous. By creating their own formats, they feel they are taking proper responsibility for their software.

Using standards will certainly speed up development. However, this is not because standards offer short cuts. Rather, standards are highways.

Today's system development and integration isn't carried out on surface streets. It's not a neighborhood business. Everything we create or adapt has to work with other systems and other organizations. It really is a global world. And data standards are the highways of this world.

Interoperability Among Standards Bodies

The Healthcare Information and Management Systems Society (HIMSS) has suggested four interoperability principles in a recent submission to Congress. I wonder if they could have been a little tougher in the first principle, which starts like this: “Standards development organizations should offer educational resources and support in order to implement and develop data standards.”

Or perhaps what I'm looking for is a Principle #0. There's nothing wrong with education and support – it's vital. But before you get to that point, you have to ask why there are standards organizations in the plural, what each organization's remit is, and how they work together. My zero principle would be that if there have to be multiple standards organizations then there should at least be a single coordinating body to which all those organizations federate.

Without interoperability among standards bodies, you won't get interoperability among systems.

Exscribe

Standards Make Smarts

Data standards have many benefits. One benefit that is not often cited is how standards enable people to understand more than their immediate environment. Standards actually make you smarter.

Here's an example. About a decade ago, student mental health services came together to agree a standard for recording data. Thanks to this action, they now have data on over 100,000 students from 140 institutions. Analysis of the data shows, for example, that drug and alcohol problems are on the way down, but self-harming is on the way up. These findings can in turn influence how information and counseling services are designed and delivered.

Switch now to the business world. If you only know about the customers, partners, and staff in your own corner of the organization, how much do you really know about people? And if you can only look at your own transactions, achievements, and processes, how do you know how well you're doing? When you have islands of information, your metrics don't travel.

Switch up the focus and the same logic applies at the industry level. You can't achieve best practice unless you can meaningfully compare existing outcomes with other organizations. Also, you can't even know what collaborative opportunities you're missing unless you have systems of record which are data-compatible.   Choosing not to exploit data standards is to opt for limited knowledge, reduced agility, and missed opportunities.    Onward State

The Compatibility Continuum

Creating data standards is a community effort. Standards imposed from the top down rarely work. This is because however well crafted they are technically speaking, top-down standards always incorporate edge case decisions that alienate one or more sections of the community. In some cases, the standard may have opted to endorse a formulation used by the majority of players in the domain, and assumed this will be supported by everyone. It's not the solution that's the problem, but the assumption.

A phrase leaps out from a report about standardization in radiology. Dave Harvey, managing director at Medical Connections in the UK, is leading an effort to establish standards for metadata in the field. His group is aiming to enable interoperable systems which are “as compatible as local politics will allow”.

This is a very realistic approach. He says: “You’ll get trusts working on these projects, but when they try to do any sharing between them, they find the data models don’t work. There’s a better chance of integrating if you start with at least vaguely similar data models.”

He's right. Integration is a continuum. That's why it can be better to use a term like collaboration – “integration” sounds like it has to be all or nothing. Finding workable spots on a continuum requires negotiation. And negotiation requires flexibility. Remember, flexibility doesn't mean anarchy. It means tolerance – in the engineering sense, and in the social sense.

We work together because we're stronger together. We must build our data standards together as well.   EHI