In a previous blog, we wrote that information findability is as worthy a topic for ISO 9001:2015 standards-making as say, risk-based thinking, preventative actions, or the process approach. Unfortunately, there is still a wide chasm between even the newly-published ISO 9001:2015 requirements and Information Management (IM) practices.
In this post, the intent is to look at the reasons why good information management can make a strong contribution to meeting the quality objectives of an organization. It also provides a high-level introduction to IM for those more from the Quality side.
Reinvention, Rework and Innovation
Reinvention can be a good thing if it is intended. Unintended reinvention on the other hand is one of the more common wastes in product development. The converse is re-use, which is the ability to synthesize previously generated good ideas into a new product design.
Redeka (2012) defines reinvention as "the need to redesign or rework something because previous solutions to a problem are not accessible to the problem solver or not generalized enough to be reusable".1 It is typically a waste of time, resources, and money. The quantity of reinvention in an organization usually indicates lower-quality product development environments. Any process or organizational structure that reduces rework and prevents reinvention is often presumed to also increase innovation.
But how can you reduce rework and reinvention? It's all very well to say that one of the responsibilities of a quality management system is to keep rework to a minimum. There are book-loads of case studies talking about the "learning practices" of companies. You could read a thousand examples where reinvention could have been avoided and still make a mistake the next day leading to rework in your own project. The "why" is very different from the "how". Practical value however seems to come from: (1) the timely use of reviews, checkpoints, checklists, and documented process to check the work in the first place; and (2) the use of systems to make documented information readily and easily available. The organization can't learn how to remove rework, but it can put in place the processes to minimise it. Those processes depend on IM.
When talking about IM as a way to minimise rework, we should not fall into a trap of considering information as just a passive repository. It is equally important to ensure that changes to knowledge are actively pushed out to users. After all, there is limited value in working on something if changes elsewhere invalidate your work. Even if you remembered to search for relevant facts before you started work, changes can happen while you are working. A good process of keeping people up-to-date is needed. A product design review might be limited to a small set of reviewers, but there could be a much larger set of people who are notified of the design decisions.
More and more information is unstructured
So, it's easy then. Just make a ton of information available to the organization and top-quality innovation will surely follow. If only.
To begin with, a significant percentage of organisational knowledge is tacit. It is the experience and know-how that employees use in their everyday activities. The more that can be captured as explicit knowledge, the better. Putting the captured knowledge into a common document repository is another big step forward. Two examples will help make the point. First, creating a standard operating procedure (SOP) as graphical content that can be used from a web browser is infinitely better than a diagram stored in a filing cabinet. Second, capturing the conversation that happens around a product design review as part of the document history in the repository is better than trawling back through a set of (private) email inboxes.
Even when it is captured, the typical organization holds knowledge in structured and unstructured form. The structured form is that found in databases with data contained in fields. The unstructured form is the rest: all the word processing documents, presentations, emails, PDFs, audio, video, and image files. The mix is typically around 10-20 per cent structured and 80-90 per cent unstructured.
There will be ongoing growth in the amount of structured data - transactional records from e-commerce systems and data from intelligent sensors, for example, will see to that. But even higher growth rate is expected for unstructured data. Yes, there will be big data streaming from the home heating device sensor but there will be even more in the customer service case files and in what they said about the Energy Company on social media. Analysts such as IDG say that unstructured data is now growing at the rate of 62 per cent per year.
Inside companies, that key piece of knowledge that could prevent reinvention may be stored in unstructured form such as field reports, maintenance notes, or surveys. The key is to tap it, and that means making structured data from these unstructured forms.
So, we'd better find a way to organize the knowledge. We've now entered the realm of taxonomies and ontologies.
Metadata, metadata, everywhere
Taxonomy is classification, and we use formal and informal taxonomies all the time. For example, your email client may have folders where emails are automatically filed according to a classification that you chose. In addition, you may add tags such as "to do", "important", "personal", and so on.
There are different types of taxonomy such as hierarchical and faceted taxonomies. We use hierarchical taxonomies for things like animals or plants arranged in one single hierarchy classification, and faceted classification when we use multiple attributes (facets) to sort objects into categories. We can also use indexes - for example there is no simple classification of employees at my company, but sorting their surnames by alphabetical letter helps me find someone quickly.
Metadata is a set of attributes that provides information about digital assets. A photo taken with a Smartphone usually contains default metadata embedded in the image file such as time, geolocation, date, and camera model. When I add that image to a photo sharing site I may add other attributes (or Tags) such as "summer vacation", "great sunsets", and so on.
"Metadata is a set of attributes that provides information about digital assets."
It is regarded as a truism that, the better the metadata, the better the chances of finding any piece of information. As soon as data is identified by metadata tags, it is structured data.
It is important that metadata is not some pre-fixed structure, and that companies can create custom metadata tags for documents in their repository.
Metadata also has a big part to play in an audit trail history. This is why it's usually OK to change a present or future metadata value (the next review date for a document), but not OK to change a historical value (person <X> approved document on <date>). The review history of a document may be as valuable as the current document content, as it will explain changes that may be applicable elsewhere. Metadata can provide knowledge about who in the organisation is responsible for knowledge. Knowing who last updated a document can be useful. This is knowledge which can be lost by storing documents in a file system or shared drive.
To bring this back to our theme - structured information aids information retrieval; better information avoids reinvention, and this in turn enhances quality management. The next question is: what are the best ways to turn unstructured content into structured metadata so that we can find information?
We will explore this in Part 2 of this series, where we will look at Enterprise Search and the value of Folders and Categories.
1 Radeka, K., 2012 "The Mastery of Innovation: A Field Guide to Lean Product Development", CRC Press, ISBN 9781439877029
This post was written by Paul Walsh.