I find once a few key people get the Enterprise 2.0 bug and read all the content on the web about changing demographics, retaining new and younger talent they lose clarity on where a more immediate impact can be had. I have been guilty of this myself and want to share some thoughts on the subject.
First of all, I am a huge proponent of bringing Web 2.0 behind the firewall. I think it has a trememdous amount of potential to democratize an organization and give people a way to have a voice and contribute to a unified effort. Blogs, wikis, tags, folksonomies etc. are all great ideas and definitely have their place. But attempting to implement them without first understanding the language of your organization is a mistake.
Firstly a discussion of structured vs unstructured data is in order. I like this excerpt:
“We all know that structured data is boring and useless; while unstructured data is sexy and chock full of value. Well, only up to a point, Lord Copper. Genuinely unstructured data can be a real nuisance – imagine extracting the return address from an unstructured letter, without letterhead and any of the formatting usually applied to letters. A letter may be thought of as unstructured data, but most business letters are, in fact, highly-structured.”
I consider structured data to be information that can be easily described, has a defined container to store it and is easily searchable. The problem is that the vast majority of business interactions take place in an unstructured environment: email, IM, RSS etc.
I saw an excellent example somewhere where someone used the example of New York City. In a structured scenario New York could be described with a latitude and longitude, a type of City, possibly a parent State and Country attribute etc. In an unstructured scenario New York could be referring to the city or state, the Yankees, Rangers, Jets, New York Stock Exchange etc. Gleaning the context and intent from unstructured data is much more difficult.
I’m going to analogize in the context of MOSS 2007 but the same principles apply regardless of the technology platform. I’ll use the example of an HR department who finds out the IT has implemented SharePoint (I’ll discuss why this is a mistake in a later post) and believes that this is the silver bullet for getting people to collaborate and share documents with each other. The HR lead sits with IT and runs through a vague set of requirements which ends up with a few document libraries in SharePoint that grow over time to become complex enough that internal customers find it too tedious and over time revert to the previous ways of doing things.
Unfortunately the question that is usually asked is, “What content would you like to share?”. The questions that should be asked are:
- Who are your target audience(s)?
- What business processes support the interactions with those audiences?
- What is the language of HR? In other words – what is a policy, an employment standard, leave of absence , vacation etc.
This then leads to going through the process of identifying metadata for each type of content. It also will allow for the idendification of standard templates and workflows. Having this data well-thought through is the foundation of an effective enterprise search strategy.
So, instead of a customer meadering aimlessly through a semi-structured repository they are able to interact and search based on the language of the business. Without going through this process you will exacerbate the problems you currently have with your enterprise data. In fact, if the adoption rate of the portal is near expectations it will accelerate the mayhem.
There is a growing suite of vendors that have offerings on the market for integrating unstructured data into your business intelligence and analytics landscape. I submit again, that if you can’t manage structured data and your business intelligence capabilities are limited you will be striking a Faustian bargain by attempting to engage a partner or implement a solution that focuses on unstructured data.