Uncorking the Value of Metadata in 2019 And Beyond
Last week, Graham Allan wrote an article over on LinkedIn with some great points which echo those we have made previously on DAM News about digital asset supply chains:
“A modern DAM is used for creative workflows, as an API-integrated source of truth for other information and content systems, and as a distribution point for content and product information. Assets will generate new data about themselves – and about the people and processes they touch – as they progress through their lifecycle. Imagine your operational dashboard showing real-time analytics of your content supply chain, modulation of search results based on business data from your ERP or even from live consumer behavior, or the ability to elevate querying to BI-levels of sophistication (“show me all the assets I have rights to sell for SVOD in the UK, in genres which have been at least 30% profitable there over the last three years”). ” [Read More]
While the article is good (and I made sure to ‘like’ it) I disagree with the core premise and have a perspective which is diametrically opposite to Graham’s. I added a comment to the article where I asked others who had also liked it to explain their reasoning, but as I write, no one has done this (although Graham’s boss has complimented his colleague, which is obviously highly commendable). I am assuming these people all understood what Graham meant but if they can’t rationalise why they thought it was a good piece, why did they like it? In the absence of any other analysis of the topics discussed, I will provide my own.
So, how can I like this article yet simultaneously hold an apparently contradictory opinion to the author’s? I will explain myself via the use of a metaphor and an illustration of why I think it is an appropriate one in the context of the article.
In the opening paragraph, Graham opens with this:
“After a long and dominant run, we at KlarisIP pronounce the end of metadata as we have known it in the world of DAM. Instead, the new mindset we must adopt is (drum roll….) – data!” [Read More]
Typically after drum rolls, bottles of carbonated beverages (of one description or another) are often produced and opened to toast the momentous occasion. As readers might be aware, the superior technique to extract the liquid from the bottle is counter-intuitive and involves removing the bottle from the cork, rather than the other way round. On a number of occasions, I have witnessed people who have forgotten this essential fact visibly struggle, needlessly wasting the contents by spraying it around after a period of frenetic activity, or being forced to procure corkscrews, knives and other unnecessary sharp metal implements to gain access.
In asserting that metadata should play a subservient role to data, I believe Graham is making a comparable error. The instinct to get at the good stuff is one we all share, but his technique and reasoning is the wrong way round and may be wasteful and inefficient as a result.
Astute readers will probably have already guessed where this counter-argument is heading, but for those who would prefer I explain my reasoning in more detail, I invite you to examine the two terms that are key to this discussions: ‘metadata’ and ‘data’.
Metadata frequently gets defined as ‘data about data’ as we have discussed many times on DAM News, a more precise definition is that metadata is contextual data about assets. Why do you need contextual data? To find whatever you are interested would be one very good reason, but more or less any scenario where you have to isolate an entity like a digital asset (or group thereof) requires some context otherwise it becomes virtually impossible to differentiate one asset from another. As such, the more context you have about a digital asset, the more likely it is to be valuable (i.e. useful and worth something).
There are a variety of definitions of the term ‘data’ and I had to read multiple sources to get some examples. Here are a few:
- ‘information in digital form that can be transmitted or processed’
- ‘factual information (such as measurements or statistics) used as a basis for reasoning, discussion, or calculation ‘
- ‘any sequence of one or more symbols given meaning by specific act(s) of interpretation’
In other words, the definition of ‘data’ changes according to the context (that word again).
The essential problem with shifting from a metadata-centric view of digital assets to a data-oriented one is that rather than adding context, it removes it, with a corresponding diminution of the value of an organisation’s digital assets. Where previously there was a concept which emphasises structure, order and having scalable frameworks, you move to one where none of these things exist by default (hence the multiplicity of definitions of the term ‘data’). The most likely outcome for those who try to ditch metadata is that they will end up re-inventing it, but with some alternative name which means the same thing. Re-inventing and re-branding wheels is stock-in-trade for the tech sector, but I wonder what value it really adds for clients and the advancement of human knowledge in the wider sense.
Returning to Graham’s article, there is a phrase he has used, which I agree with absolutely:
“they [DAM] are central hubs within a broader, technology-enabled ecosystem.” [Read More]
The key term here is ‘central hubs’. In other words, the DAM is at the centre; users and other applications regard it as definitive or ‘the single source of truth’. The rest of the data in an enterprise is contextualised by relating it back to the asset, i.e. it’s not just data, it’s metadata.
The reason why lots of enterprise stakeholders who lack Digital Asset Management expertise tend to use generic terms like ‘data’ or ‘digital’ is because they are a convenient shorthand (or perhaps a shortcut) to avoid needing to be more specific about what is actually meant. In reality, almost everyone has some pre-existing metadata model to contextualise ‘the data’, the problem is they either can’t or won’t find ways to relate it to everyone else’s, so the shorthand ‘data’ is settled on as a kind fudged compromise term. Even if you talk to IT hardcore people , who are often the ones who prefer ‘data’ (in the generic sense) they still have their own metadata model, but it is based on where the data will reside, so ‘database’, ‘storage’ etc. No one ever thinks about data generically without also considering the context it will get used in, i.e. the metadata. This is why metadata has been integral to DAM since it first came into existence as discipline nearly 30 years ago.
A few months ago, I discussed Magan Arthur’s article where he talked about the need for ‘business objects’ and I contended that these were ‘digital assets’ so a new concept wasn’t needed as it exists already. Even though I disagree with the terminology, I fully concur with his analysis. Enterprises need more methods to classify, slice and dice their digital assets, in whatever form they might present themselves – not less of them. While a big picture perspective is required, at some point you’re going to have to get your hands dirty with the details and have a robust method for dealing with them which you don’t have to re-invent endlessly and constantly explain over and over again.
Rather than acquiescing to using the term ‘data’ in preference to ‘metadata’, since the DAM is now a central hub for many enterprises, a preferable course of action would be to take the initiative and start persuading everyone else in the room to understand that what they too are dealing with are digital assets – and metadata in particular, not just data.
A frequent issue DAM consultants have to deal with (and possibly other fields too) is where stakeholders will attempt to resolve a problem by deliberately limiting their understanding of it, rather than really coming up with a genuinely simpler solution. I believe this is what is taking place here and I would encourage Graham to metaphorically take the bottle out of the cork and contextualise data in order to transform it into higher value digital asset metadata.
For anyone who is in London on Tuesday 27th November and would like to discuss this article (or others) with myself and a group of other DAM enthusiasts, The London DAM Meetup Group is holding one its all too rare socials at 6pm in the cellar bar of The Cittie of Yorke pub on High Holborn near Chancery Lane. You can sign up for this free of charge event here: https://www.meetup.com/LondonDAM/events/255632599/Share this Article: