The creation of business data standards is spreading out from commercial and government organizations to the big wide world. The motivations for new standards, and the methods being used to formulate them, are instructive. Looking at how these new standards movements are faring may well help business organizations to improve their own standards processes – saving time and money, and opening up new potential benefits.
A paradigm case for the new, in-the-wild standards world is that of food inspection. Food inspection could sound like something dull and technical – a box-checking exercise designed to add costs to restaurateurs. But badly sourced or prepared food has major health impacts. An e-coli outbreak in 1993 claimed more than 700 victims in four states. And this wasn't just a few upset stomachs: over 170 victims were left with permanent kidney and brain damage, and four children died. It was traced to under-cooked burger patties.
Data standards for this domain are a two-level job. The standards must define the key data elements from the ground up as well as ensure compliance with the definitions. In business, it's been usual for the first level of this task to be reasonably straightforward. Say you want to define a “product”. You can usually come to an agreement about what a product is, its attributes and relationships, even if there's a deal of discussion and compromise along the way. The data associated with food safety is much more diverse and there's less of a common mental model waiting to be dug out of the participants. And, which is very interesting, a large part of food safety monitoring comes from the experiences and impressions of customers.
Now, the challenges facing the creators of food safety standards look very much like those facing businesses in an age of big data and social media. Business isn't fully in control of the data model any more. Consumers, partners, and regulators contribute as much – if not more – to the model than the business entities who think they “own” the domain.
What can standards advocates and adopters in business learn from areas like food safety? First, remind people of the stakes. Bad food kills people – get the data right and you save lives. Characterize the risks of bad data in your own domain, whether those are lost sales, reduced market share, slow response, rework costs, reputational damage, blocked access to markets etc.
Second, expect to discuss data items that are completely new to the domain. Many of these will come from the Internet of Things, initially from wearable tech. What do we want the sensors in our environment to measure? What scales should these measurements use? What are the threshold points we're interested in?
And third – solicit and embrace data from other parties. A diner's assessment of how clean a room is matters to other diners, and therefore ultimately to the restaurant's owners. Your customers' feedback is not just useful for informing how you develop the business. It's also a primary source for other customers' behavior. Everything you do is now public. Government Technology