Gregory Maciag

Standards Enable Insight

The incredible growth in the power of IT, together with the tumbling costs of technology, have brought profound changes to our lives and businesses. The most noticeable change is the shift to “always-on” expectations. We expect processes to happen faster. Our organizations work on a more real-time basis.

A less obvious but equally important result is the massive increase in the volume, variety, and velocity of data. The world is producing great quantities of data and the amounts are about to get exponentially larger with the dawn of the Internet of Things.

So we're nearing the point where marveling at the technical achievements of IT is giving way to realization that managing and exploiting data is a real challenge – and opportunity. This struck me with new force when reading about big data and genomics. Barely 20 years ago, sequencing one genome was time-consuming, complex, and expensive. Now it's routine and cheap. So much so that  “sequencing is merely data gathering for the biology”. But: “This deluge of data needs more standardised metadata – from which true insight will come.”

We've been here before and we'll be here again. As a domain moves from exploratory to production phases, prototypes are superseded by products, pioneering is replaced by processes, and disparate data is organized into coherent frameworks. For once, “paradigm shift” is an apt term.

Here's an authentic voice from the heart of the paradigm shift: “When I first heard about these standards developments, I was bored almost to tears. But now I realize that this is a very important aspect of getting the most out of genome data.” This researcher tells how, without full data standards for the domain, it's very hard to relate gene sequences to organs and body systems.

The movement for standards in genomic sciences is timely and well led. I wish everyone in the field success. I hope people in other walks of life will take note of their guiding principles – especially the advice that standards enable true insight.    BioMed Central

Smart Insurance

“There will be a saturation point at some time where everything is connected up – where you won’t be able to buy a doorlock that isn’t connected. After that, it’ll be all the services created on top of these things,” says Alex Hawkinson of SmartThings. His colleague, ex-Googler Kelly Liang adds: “We’re talking to all the key players who have interest in this space – whether they’re telcos, insurance providers, tech companies.”

The smart home is usually sold as a kind of control-freak dream that leaves a lot of people cold. How many of us are organized enough to want to control our ovens from the freeway, or have the garage door opener talk to the refrigerator?

The real excitement around the smart home is in the services made possible by interconnected devices. There's home security, for example. Care of the elderly is another potentially vital service that can be layered on to the smart home. And insurance is surely a massive opportunity.

Soon, switch on an appliance and part of its boot sequence will be to fetch a range of insurance quotes. Soon, insurers will have a pretty good idea of what stuff you own and what condition it's in. They'll even know how active you are, based on energy use patterns in different parts of the house. And these are just off the top of my head.  Forbes

Looking Outward - Industry Standards Trump Proprietary

WHAT.... I did a double take when reading a piece on the securities industry by Sapient. I'm nodding my head until the very end, when it goes weird.

The argument goes that regulatory forces will compel buy-side companies to put their data in order – break down the silos, connect up the sources. But the conclusion goes like this: “[Sapient's Duncan] Cooper believes industry standards will take some time to emerge, but firms should begin looking internally to help drive change. 'Most firms don’t even have internal data standards and this needs to be the first step. If you can get your own house in order, then external standardisation can come later,' he said.”

I disagree. If you don't have internal data standards then you're in the perfect position to adopt industry standards! You should grab hold of whatever industry standards are available, even if they are not yet complete. Also, you need to regard regulatory requirements as heavy, heavy hints toward core entity definitions. This is the fast way to get your house in order while minimizing your overall costs of transitioning to a data-centric enterprise that's fully connected to your customers, partners, and regulators. Creating internal proprietary standards and needing to then create industry standards is a waste of time. Go for industry standards first and ease the adoption of them internally. The Trade.

Standards are a Best Practice

There's an important article about IoT standards by Thomas Davenport and Sanjay Sarma in the Harvard Business Review. Did you know the term “Internet of Things” was coined in 1999? It's less surprising to learn the RFID standard – the authors' historical measure for IoT – took 15 years to develop and implement. And you've got to agree we can't take as long to develop new standards for the IoT.

Davenport and Sarma list a number of factors that worked well for developing the RFID standard, and make recommendations for additional factors needed to make IoT standardization a success. The ACORD community can be forgiven for feeling satisfied at the close match between the authors' prescription and ACORD's practices. Our standards achievements have benefited from collaboration across a diverse community, deep user involvement, highly influential players, focus on outcomes and being device agnostic.

We've also made great strides in two of the areas cited as “what needs to happen differently”. These are “faster process” and “bottom-up efforts”, where ACORD has made structural changes right across standards management, engagement, and education. I'd also have to say that I'd shift the authors' third recommendation from the “must do” list to the “already doing it” list, in the case of ACORD. This is the call for “more carrots than sticks”. I think ACORD members have developed a unique ecosystem in which we all encourage and help each other. We understand the benefits of collaboration – and we know it's ultimately the market that carries the big stick.    HBR

Unscrambling Eggs

There's steady and even growing business in the field of data reconstitution. This is the kind of processing required to create compatible data sets from disparate sources. And it's a waste of time.

Dr William Hersh explores this topic in an excellent piece about healthcare data. He uses great examples that are easy to understand while being taken directly from the real world. He makes it clear how implementing data standards right through the healthcare process has great benefits for everyone, not least patients.

Why take a stack of reports, PDFs, scanned images – and grind away at them to reconstruct the discrete data items they originally held? It's crazy. But it's happening everywhere.

It can look like a smart solution. You don't have to tackle any of the existing systems. You don't need to challenge anybody's processes. You can look like a light-touch, hands-off data manager.

However, from the viewpoint of the whole organization, this is a waste of time and resources. It's also a new source of risk. Reconstituting data involves interpretation, which can always go wrong. Also, if people use presumed-common fields in different ways, or with different value ranges, you'll never get back to correct data through a retrospective process that doesn't involve the originators.

Dr Hersh calls it unscrambling eggs. He's right.   Informatics

Drones for Claims

Catherine Smola at CSIO notes that insurers are interested in using drones for claims work. The scope for using unmanned air vehicles (UAVs) to check damage and so on is large. At the other end of the business, the scope for insuring UAVs must be good too.  Canadian Underwriter

IoT Wake-up Call

Speaking at CES 2015 in Las Vegas, (I was there) Samsung President and CEO BK Yoon said that the possibilities of the Internet of Things (IoT) are “infinite” – but also called for more collaboration and openness across industries to unlock that potential. In the meantime, all Samsung devices will be IoT-ready in five years time, with IoT-enabled TVs leading the pack in 2017.

How do we square the calls for collaboration with the announcement of a single-vendor implementation timetable? I guess it's pragmatic leadership. Samsung is busy inventing the types of devices that will populate the IoT and it needs to get them out there.

I hope this turns out to be a wake-up call. If the electronics industry gets this wrong, it'll make VHS versus Betamax look like a squabble in a stone age village. And if the business world doesn't wake up soon and start to appreciate how the IoT is going to force real-time business into every area of life, then there will be a shake-up that makes the online revolution look tame in comparison.  CIO

Data Standards in the AIr

Full disclosure: I haven't read SITA's new paper, Smart Thinking. (SITA is a key IT and communications body within the air transportation industry.)

But I'll take away two things from the peripheral coverage of the paper. First, SITA says: “Access to data is the foundation for seamless travel”. Second, a helpful press release from Sky Assist reproduces survey findings which show the top five priorities for airlines and airports. The top two priorities are “system incompatibility and data integration” and “insufficient data standards”.

Reacting purely on the basis of this surface knowledge, I've got to ask: Could those top two issues be related? Just asking.   SITA

Will Google Insure the Future

This topic comes up at every meeting I attend. You know who youu are!  Well.   Jack Uldrich pulls together some suggestive moves by Google and shows how they could add up to a major insurance play by Google. The evidence is there and his argument makes sense.  What do you think?  Google seems to spread its bets right across the board. I'm sure they're interested in insurance. They're interested in everything.  GoogleInsurance

Data for All

A call for the UK to follow the US with a DATA Act of its own signals that America is leading the world with its embrace of open data. The principle that the people's data should be available to the people is one that we should be proud to assert in the wider world.    Open Data Institute

Data Standards = Customer Requirement

This is a data standards parable, taken from a piece about the quantified self movement – that's people using personal performance and health data to improve their quality of life, get fitter, or manage a medical condition:

“The average person is not going to spend time creating data visualizations with the information gleaned from their FitBit. Even experts have trouble dealing with the data, especially when it's from hundreds of different fitness trackers with a unique definition of what constitutes a 'step' [… Quote:] 'It would definitely be easier if everyone used a Shine [which tracks health and sleep] and an iPhone […] I'm not saying it would be better – I think it's good to have competition and diversity […] But it's always easier if you don’t have to normalize data from different types of devices.'"

The key phrase here is “unique definition”. If people don't share a common, stable language which defines their data, life is hard. One quick alternative is to clear the market of competition, so everyone uses the same definition by default. But the de facto standard created by the sole supplier may not be the best solution for everybody. A provider can come to dominate a market in many ways, without necessarily using data standards optimized for the whole community.

So, simple choice creates a problem: incompatible data. Removing choice creates a different problem: inadequate data.

The way through this conundrum lies in rethinking the concept of choice. As customers, we want to make choices based on various criteria, many of which are implicit. I might want a fitness monitoring device that looks like a serious piece of sports kit, for example, rather than a medical appliance. Or I might want all my self-monitoring functionality available from my smartphone. And so on.

Providers need to recognize another implicit requirement: I want to be able to share my data across devices, analysis services, and other third parties. The implication of this requirement is that I want data standards.

Customers want data standards, even though (most) customers have never heard of them, and are not interested in sitting through a lecture about them. They want – and need – data standards just like they need all the other unmentioned standards built into modern products and services. Would you sell a fitness device that didn't take standard batteries? That couldn't recharge from the wall? That had to be paid for in goats?   Absent data standards, the quantified self movement will surely stall. Data Overload

Data is New Building Material

A striking thought: “Data represents the building material of digital cities.” Tim Adler's article says digital cities will lead to new markets, just as the creation of ports triggered a huge number of ancillary industries.

This insight can be applied in many other settings. Digital business is often explained just in terms of speed and connectivity, since these are the qualities that express its revolutionary nature in the most vivid ways. Customers don't want to wait and they want responsiveness, so digital is sold as the enabler of unprecedented service quality.

But the data at the heart of digital business is the real treasure. When customers and companies connect, it's data they trade. When companies respond, they do so on an informed basis – using data. When business partners collaborate, they share knowledge – derived from data. The new value added services and products created by digital businesses and their ecosystem partners are all founded on data.

CIOs, CTOs, CDOs – all should tape this message over the mirror: Its the data, dummy.

The data holds our enterprises together. The data gives our organizations purpose. The data powers our operations. The data benefits our customers. The data underlies everything we do. The data is what we are.   Digital Cities

The Steel Thread

My favorite metaphor for the role of data in organizations is flow. ACORD is all about moving data, because that's what business is all about. Many people call data the lifeblood of organizations and I agree with that.

Randy Thomas has a different metaphor: “Data is the 'steel thread' that runs through health IT applications and connects the frontlines of care delivery with our ability to measure, analyze, and improve.”

I like this too. Data flows, yes. But the flow of data connects the players in a business process. Data flow defeats silos. Data flow makes partnerships possible. Data flow puts customers in charge. It joins everyone together.

My caveat would be this. Data flow can be thought of as a thread. But it's only going to be a steel thread if it's functioning at 100 per cent. Too often, the thread breaks. Steel must be our aspiration. Standards put the steel in the thread.   HIMSS

 

CrowdSourcing and Data Conflation

Asking users to supply it can be a quick and easy way of gathering the large quantities of data you need for running analytics. If you're tracking usage of a website, you don't even have to ask – the data is collected passively. But for a class of data known as volunteered geographic information (VGI), collecting the data is harder: “due to positioning device inaccuracies, place name inconsistency and variable observer skills, VGI requires quality authentication.”

The smart way to tackle this issue is upfront. Require your donated data in a standard format: “Effective use of standards can simplify the process of conflation: unifying multiple separate sources of data into one integrated, all-encompassing result. Standards also can simplify quality assurance of multiple types and sources of geospatial data when included in the automation processes.”

I want to go one step beyond this good advice. Any organization that is offered data in some format that will require interpretation, transformation, or some other process to make it usable, must turn it down. If people give you something you can't use, they're not doing you a favor. They're creating liabilities.

If they care enough about you to volunteer their data, they'll be happy to provide it in the format you specify – unless they're already using an established standard, and you're asking for something different.

My belief is that a lot of the data disparity we see in the world is unintentional. Disparity arises because people design their own databases, craft their own spreadsheets. If you show them the standard you'd like them to conform to, most will go along with it.

This is one of the big differences between the era of the crowd and traditional information technology. People who volunteer data don't have the same goals or constraints as IT vendors. Nor do they have investments and decisions to defend, as organizational IT bosses do. They don't have to save market share or save face. And they don't see stuff like data formats as worth going to war over.  Baseline

The Data Revolution

The UN has a report out recommending a “data revolution” for sustainable development. There's a passage in the press release which sounds like Donald Rumsfeld on a bad day: “The report highlights two big global challenges for the current state of data: the challenge of invisibility, for instance gaps in what we know and when we know it: and the challenge of inequality, including the gaps between those who know and those who do not know what they need to know [to] make their own decisions.”

Joking aside, these challenges are indeed important – and they can be observed in many business settings. Generally, we do a better job tackling problems of invisibility than we do with problems of inequality. We might not call it data inequality in the business environment, but we still suffer from failing to get the right data to the right people at the right time.

Commenting on the report, Craig Fagan says that data transparency has a key role to play in reducing corruption. Again, the development community could take some lessons from business here. One rarely mentioned benefit of using data standards is that they minimise people's ability to hide stuff or account falsely.    Data Revolution Group

Transparency International

Standards are Critical

Every so often, an advocate or leader makes the case for data standards in a crisp and concise way that's simply impossible to resist. Here's Seth Foldy: “Health information exchange is a critical technology for public health […] Healthcare providers and health departments exchange tens of millions of reports and documents yearly. These protect communities from disease outbreaks, prevent crippling disabilities in newborns, protect against identity fraud, and assess the health of communities and the effectiveness of population health programs.”

The pattern in Foldy's case is replicated across industries. We find business partners exchanging huge amounts of data with each other. And it's this information sharing which enables the business goals of the participants. It's not busy-work. It's business.

Data continues to proliferate and the number of parties sharing data with each other continues to grow. But this isn't – or shouldn't be – a problem. It's what needs to happen. Business data standards  enable data collection, sharing, and analysis. They apply as easily to a trillion records as a handful. They really are the key to success.   HIE Watch

Fly Right

Paul Saunders has contributed to a new data standard for flight operations, created by the major aircraft manufacturers, airlines, and vendors. The standard ought to save everyone time, money, and complexity. But, he says, none of the vendors is going to use it, key participant Airbus isn't using it, and Boeing won't use it unless the airlines ask them to – and there doesn't seem to be a rush.

It's frustrating. You just know there are good people in the industry struggling to make business cases for something that's pretty much a no-brainer. You know they're having a hard time getting an audience. You know what they're up against.

Attitudes have got to change. In his intro, Saunders mentions an important driver for change – the sheer obviousness of standards to anyone entering the IT industry. Here's the insight: “Not to make light of the technical challenges involved in working APIs, but this kind of integration between apps and services like Twitter and YouTube are being taught in my son’s school… and he’s ten years old.”

If business leaders and managers can't see that data must flow, they should be thinking of getting out of the way.     Flatirons

Data Standards Do Good

As the collection, analysis, and especially sharing of data become more and more central to the way we live our lives, the social – and even moral – responsibilities of data professionals are growing. Good data management can no longer be ghettoized as a technical issue. It's at the heart of our society, and it's central to creating the kind of world we want to live in.

Here's a stark example of how failing to pay attention to data issues can cause real human tragedies.  The National Commission to Eliminate Child Abuse and Neglect Fatalities says the 1,620 child maltreatment deaths reported in 2012 “may be under reported by 1,000 or more children because of a lack of consistency in how child deaths are documented across the U.S.” The number of fatalities is closely linked to the resources applied to intervention, so getting this number wrong – and by such a large factor – is a major concern.

The Commission wants data standards put in place to address this problem. Standards should also help professionals analyze the success of intervention programs and crisis response processes, helping to identify and spread best practice.  Social Work

Start with the Core

Noted health commentator Dr Bob Wachter gives a lively and illuminating account of a decade's efforts to wire up healthcare, paying particular attention to the last five years. Data standards are at the heart of the story.  Wachter argues that standards have been too prescriptive and have interferred with health delivery and innovation. He wants to see a new focus on basic interoperability standards – which he likens to lightbulbs and electrical plugs.

Responding below the line, Ross Koppel says: “I think we could have had a far better set of EHRs [Electronic Health Records] if we had insisted on a set of data standards and usability testing by the time the funding was available. The decade-long delays in agreeing on standards was not due to technical reasons, but rather to proprietary efforts, attempts to capture market share and to avoiding allowing hospitals to switch vendors. Call it greed, wise marketing plans, whatever. It was destructive and we live with that failure.”

This debate resonates with me. There's a tension between creating standards that cover the whole domain and standards that energise the core business. Ideally, you want to do both in tandem. But experience suggests that you first need to build a solid, minimal core that brings immediate business benefits to the participants. This also allows working relationships and standards building processes to evolve, while stoking an appetite for tackling other data sets.    Wachters