I attended a very inspirational session at SPC14 run by Scott Jamison (Making SharePoint Collaboration Rock by Increasing Discoverability), and thought I would build on what Scott covered in the session.
One thing we have seen over the past few years as SharePoint consultants when looking into collaboration is how many people seem to use information in different ways.
Probably the biggest ticket item that has been in need of a flak jacket is that of folders. We have been taught (and teaching) for many years that folders are bad and instead metadata is the way of the future.
This doctrine has held true for most people where metadata was (and rightfully) seen as something important to increasing the discoverability of content. As with anything it’s all about applying a common sense approach. Metadata for example is only as valuable as the data specified.
What this translates to in practise is that if you require too much compulsory metadata from your users, then you are not going to get quality metadata tagging, hence negating much of the need for metadata in the first place.
As an end game however, metadata is still a goal worth striving for. Metadata is key to improved search results within SharePoint, simple filtering (e.g. metadata navigation), and of course the introduction of consistent classification of information within an organisation.
Over the last few years, metadata has continued to be pushed almost exclusively as the way to store information. The only problem with this is that Microsoft have now gone and told users that there is this new technology called OneDrive. This almost magical technology that makes it easy to save your files locally and have these synchronise either centrally with Microsoft or directly into your SharePoint environment (OneDrive for Business).
Brilliant – end users now have a simple way to manage SharePoint documentation and can even choose to remain unaware that it’s SharePoint driving things behind the scenes. From a user adoption and migration perspective this is a reasonable approach – those who are used to saving everything into a folder on their home drive can now almost behave in exactly the same way, yet their content is now backed up and is searchable.
Whilst this is a positive step for end users, what seems to be missing in many situations is how this newly discovered ability to keep working the way that everyone really wanted to, will impact your information architecture, and more importantly your metadata assigned to documents.
In fact, you have just lost your metadata by default as OneDrive won’t require this from you (your document won’t be checked in if it’s missing anything but you won’t get a prompt to complete this in your windows explorer client).
The answer is indeed a hybrid approach. Seem familiar?
Full credit to Scott for this analogy (it was so good it did not make sense to try and make up another one):
If you are looking for a can of soup in a supermarket you probably want a label on the can to know what flavour it is right? Else you are going to spend a lot of time opening up cans in the hope that it’s the flavour you are looking for.
If we consider a label on a soup can to be metadata, then the analogy makes sense – we can read the labels (metadata) and see what sort of information (or soup flavour) is contained inside.
The only problem with this approach is that it relies on you walking down the soup isle in your supermarket to read the labels. What if soup cans were scattered throughout the supermarket alongside other food? You can certainly see what the flavour is, but you still have to know where the cans are.
This is where isles (folders) come in.
With a isle (folder) you can narrow down to what you are looking for and then rely on metadata to converge on exactly what you are looking for.
There is a symbiotic relationship here.
Moving away from Scott’s analogy above, the message hopefully makes perfect sense and you may be wondering:
a) Why I’m not doing this already
b) How can I easily do this?
The answer is actually simpler than you think. From SharePoint 2010 there is an obscure feature in your document library settings called Column default value settings.
This menu takes you to a screen where you can assign the default metadata that will be applied against a particular folder combination. In practise, users who upload documents into folders using the likes of OneDrive for Business can have their documents automatically tagged with metadata based on the folders where the document was stored.
I will flesh out an example of how to do this below. There are some manual steps required to set this up and continue to maintain the relationship when new folders are added. From a practical point of view, you probably want to apply this method to the first 1-2 levels of folders only and leave your end users with some training/guidance on applying metadata manually to documents below these levels.
Firstly, let’s look at a basic folder on my local machine. This is essentially what we are trying to replicate – but seamlessly get the content somewhere safe i.e. into our SharePoint environment which we can control, and of course get some labels on those soup cans!
If we set this same structure up in SharePoint with a couple of metadata fields then we end up with what many of us would consider a good starting point for document management.
We can tidy this up by activating Metadata Navigation and tweaking the default view to not show any folders. What we end up with is something similar to the below.
Next, let’s crack open our OneDrive sync client and connect to this new SharePoint library for offline usage.
What we now have is a solution that has gone full circle at a basic level:
Local folders -> SharePoint library -> local (cached) folders
We are still missing the key ingredient of automatically bridging the gap between those users who want to easy contribute information (save something into their new OneDrive synched folder), and those that want to peruse and consume this content through some high level metadata filtering.
Let’s jump back into our document library settings in SharePoint, and more specifically open the Column default value settings.
Here you can browse through your folder structure (hence why the suggestion is to not go too deep here), and decide what metadata will automatically be tagged to each column when something is added to the specific folder.
Let’s perform a quick filter in SharePoint to compare results “before” we contribute some fresh content.
Jumping back to our local OneDrive, we can now upload a new document.
After refreshing our filtered view in SharePoint, we can now see the new document where we expected it – why? – because our metadata has been automatically applied based on the folder location we saved to locally.
So what we now have is an information architecture approach that supports the ease of contribution that end users expect through the use of tools such as OneDrive for Business.
For content contributors, the OneDrive for Business experience is the most likely interface where documents will be added, whereby for content consumers – the document library with metadata navigation (and collapsed folders) offers a more dynamic experience when filtering by metadata.
So as with most things in business, it’s all about compromise and aligning a balance of experiences between how users want to work and how we need them to work in order to provide value to their peers and the rest of the organisation.