On August 6, 2015 Box announced the release of their Metadata Template Editor (MTE). I’ve been looking forward to the release for a couple of reasons: 1 – having to rely on Box to set up metadata templates is inefficient, to say the least, and; 2 – the announcement is, in my opinion, another step on Box’s journey to being an Enterprise Content Management (ECM) platform. I also think that MTE will help Box stand out from others in the space that don’t offer strong metadata capabilities.
I played around with the metadata template editor au naturel. That means I didn’t have any integrations, interfaces, or anything else of that nature. I tried it on my Box Enterprise instance, completely standalone. Also bear in mind that this is Box’s first cut at the feature; I fully expect that it will mature and improve over time.
Before I get into my thoughts and observations about the metadata template editor, I’m going to take a few words to explain metadata a bit.
Metadata is data that tells us about stuff. It can also tell us what to do with stuff. When I talk to clients about metadata I break metadata into two broad categories: 1) Descriptive – tells about the object, and; 2) Administrative – tells what can / must be done with the object. Sometimes a metadata property can be both descriptive and administrative. E.g.: “Vendor Invoice” can tell us what the object is, as well as provide guidance on how we’re supposed to manage it.
In addition to describing an object, metadata also helps to determine what security settings to apply, what workflows to use, and what retention rules to apply. It also, and this is really important, helps to find the objects we’re looking for.
Without further ado here’s my initial impressions of Box’s Metadata Template Editor.
Very, very easy to do. Major caveat is that you better model out your metadata before creating any templates. You’ll also want to not repeat what Box already provides in the file metadata. You’ll have to have an account type (biz or enterprise, I think) that can enable metadata templates, and admin access.
It’s easy if you’re doing only a few files at a time and they’re all in the same folder. You’ve got to go to the previewer and add the templates from there. It will be a pain if you have to do more than a handful of files at any one time.
What I’d like to see in subsequent releases:
- Inheritance – files should inherit properties from the folder they’re in, as well as properties unique to the individual files.
- Intelligence – being able to drive security and retention properties based on selected metadata values. Being able to trigger / route workflow actions would be cool, too.
- Expand the list of data types – Currency would be an awesome addition.
- Dependent choice lists – If I select Canada as country, province selection should be limited to Canadian provinces.
- Apply metadata upon ingestion – best practice for metadata is that you apply it at the point of capturing the content.
- Master template – have a master template that I can then modify as needed, to apply to content.
- Automation – adding metadata to content is purely manual at this point. I’d love to see some sort of automation enabled to make it easier for the users, and to enforce some consistency.
- Security – it would be good if security could be set to allow users to see the content and metadata, metadata only, or nothing. This is important in some scenarios.
- Retention – when deleting / destroying content via retention rules, it would be useful to be able to select whether or not to keep some or all of the metadata associated with the content. There are a number of scenarios where this is useful, e.g.: analytics and compliance.
- Mobile – it would be really handy if the mobile apps supported managing and searching on metadata.
Search is okay. The addition of metadata templates to the standard Box search box is welcome. However, it’d be cool if I didn’t have to do a standard search first (i.e.: give me a dedicated metadata search UI) and if I could apply multiple templates (I’ll very briefly explain why this is important later). As it is, searching on metadata is really about applying filters and “searching within” after you’ve run a general search.
I think that the MTE release opens up opportunities on the services front. Box Consulting and Box partners ought to be engaged in modeling and implementing metadata for clients. The truth is that too many organizations have neither the skills nor the time to do a decent job when it comes to defining metadata models, and then implementing them.
Box’s approach to metadata can seem a little weird to those of us who are used to ECM platforms from OpenText, Oracle, and other vendors. However, this is not a bad thing. In fact, after playing with Box’s version for a bit I can see how I’d go about modeling and implementing it for a client. The key is to chunk up the metadata into multiple templates for various purposes. Start with tombstone type metadata that applies to all content types, then get more specific.
Box’s MTE is a v1.x release and it shows. However, it’s obvious that the real value will come from the metadata API and integration. Metadata is such a vital part of managing information and extracting maximum value from it and I’m pleased that Box is taking it seriously. That said, I’m looking forward to seeing what improvements they make to MTE over the coming months. For now the biggest improvements I’d like to see are in assisting the user to apply metadata and having metadata search available on the mobile (iOS) app.
For a little more help / guidance on using Box’s metadata template editor, check out this post from cloudfind.