In a recent CMSWiRe article by Boris Kraft, several examples of content management outcomes are offered. What’s extraordinary about the examples is that none of them have anything to do with marketing, accounts payable or insurance (go-to uses cases in blogs and articles for some reason). It’s also cool that the examples have to do with people accessing content via mobile devices, away from the office. Read the article; it’s good and the use cases are pretty interesting.
This post is about the importance of metadata in making the magic described in the article happen. I’ll use a project I am working on to provide some context.
The client I am working with builds electricity transmission infrastructure, and one of the projects on the go is to digitize their field reference guides. Their current process is to find the manuals/guides you need, print them out, put them in a binder, and hope you’ve got the right ones. If you don’t have the right information with you, it can be a really, really long drive to go and get it. Proceeding without it is not an option as the consequences could, literally, be fatal. Even if no incidents occur the organization could be penalized for all sorts of regulatory violations.
There are four broad categories of manuals/guides: 1 – Health & Safety; 2 – Environmental; 3- Maintain, Repair, & Replace; 4 – Engineering Standards. The documents are required reading by staff, contractors, and business partners at all of my client’s locations. The specific docs that one must read / use are determined by:
- Your role;
- Whether you’re an employee or not;
- Where you are / will be working;
- What specific task(s) you will be performing;
- What equipment you will be using;
- The weather and season;
- Environmental considerations.
Currently, the information that’s needed is stored in SAP, several SharePoint sites, a geomatics database, a health and safety tool, an engineering project management tool, and a couple of purpose built in-house applications. Getting the information requires searching the repositories you have access to, and asking others to search the repositories you don’t have access to. Oh, enterprise search is a foreign concept. There must be a better way.
Well, there is and that’s what I’m working on. We’re at the phase one part of the plan, which is a portal. I don’t like the portal approach because it’s cumbersome. It sits on SharePoint, it doesn’t provide access to all the required information, it’s not built for mobile field access, and it still requires too much effort of the user. But hey, that’s what was scoped and budgeted for in the first phase so that’s what they got.
Where this client, and any other organization with similar requirements, needs to be is at a place where the solution does the work for the user. If I’m about to jump into a helicopter to go and wash resistors (or whatever those white ceramic things are) on a 500kw line 150’ in the air, I should be able to pull the work order (using my mobile device) and have my phone or tablet updated with all the safety and procedure manuals I need. I should not have to access up to seven different repositories to print out documents that I hope are current and correct.
The truth is that the technology to automatically provide health & safety information to field workers exists. The problem is that organizations don’t understand the scope of the challenge, nor do they understand the potential long-term benefits. I’m not implying that there is some sort of magic button or potion that’ll make things happen; there isn’t. The key to making it happen is metadata. Without metadata nobody would find anything they need in a timely manner, leading to all sorts of negative consequences including increased costs, compliance violations, outages, and potential injury or death. With a properly defined and implemented metadata model the opposite is true. In addition to helping people find the information they’re looking for, metadata helps systems communicate with each other to pass information back and forth, execute workflows, and secure information.
Defining an organization’s metadata is not a trivial task, but it is certainly doable. It requires a team made up of IT, subject matter experts, end users, and information management specialists. Metadata should be defined with specific business outcomes in mind, but framed with good information governance practices. A good metadata model contains attributes that control the content (security, workflow, retention) as well as properties that describe the content.
In addition to defining and implementing a metadata model, organizations also need to figure out their content requirements. It’s all about understanding what information you have, where it’s located, how it’s described, and how you’re going to get it to where it’s needed. It’s also about understanding what information you’re missing and how you’re going to get it. For more of a detailed explanation take a look at the Principles of Holistic Information Governance (PHIGs).
The PHIGs aren’t about metadata; they’re about managing information for business outcomes. However, metadata is a big part of that. The key to metadata is to get detailed enough without getting bogged into analysis-paralysis. A government I worked with spent more than six years trying to develop an organization wide model. Big, big mistake. They should have focused on something more specific like a process, department, or content type.
With the trend of going mobile first or mobile only, metadata takes on an even greater importance. I’m convinced that the way forward is apps built on top of content-as-a-service platforms. In order for that path to lead to anything resembling success, the underlying content platforms need to have rock solid metadata to support them. If metadata and geolocation services can be used to tell you what to listen to, what to wear, what to eat, and what the weather’s going to be, there’s no reason they can’t be used to help you do your job and protect yourself. There’s no good reason why someone should have to search for manuals, guides, or procedures. There should be no more effort required than to submit a work order number (or other identifier) and show up at the work site. In fact, do it right and the order of the two actions doesn’t matter.