Cloud Brokerage – A Solid Foundation For Digital Asset Management Resourcing Or Just Blue Sky Thinking?

I was interested to read Stacey Higginbottom’s article yesterday regarding her thoughts on the Forrester report ‘Cloud Broker – A New Business Model Paradigm’.  Like Higginbottom, I am intrigued by the proposition that cloud providers might become one-stop shops for all computer resources, including short term human resourcing.  However, I was equally struck by both the intricacy of the business-model proposed and the potential de-personalisation that might occur if implemented for real, in particular for Digital Asset Management projects.

There are numerous issues to consider.  Immediately apparent is the inherent complexity of the full cloud broker model.  With an architecture that resembles a ‘Frankenstein’s monster’ of traditional business paradigms, I wonder about the risks of such a multifaceted approach if put into practice.  It is certainly my opinion that the more moving parts, the greater the potential for failure, whatever the mechanism involved.

The flaw with many of these theories, especially when applied to human resourcing, is the assumption that you can reduce any activity into a commodity service model.  This may be appealing for technologists who spend all day trying to rationalise complicated problems but the actual results, in my experience, can be mixed.

Thinking about how this might be used for Digital Asset Management (as a business process rather than simply a software product), there are some process oriented cloud services that can (and do) work quite well but there are others that are less suitable.  Highly appropriate are simplified ‘platform’ services like storage and bandwidth.  Moving further up the complexity curve are plausible services like transcoding and asset conversion where service consumers will undoubtedly run into some kind of problems but ones that can be managed.  Beyond that are ‘soft’ services such as metadata and cataloguing.  At this point, you run into the inevitable flaw with this approach: human beings are not commodities and trying to run services that assumes that they are creates bigger problems than the original scalability issue.

Also, Higginbottom identifies that in order for the cloud broker model to work, the tech industry will need to make “cloudbursting” a reality first.  However, the obstacles to realising effective cloudbursting facilities goes far beyond interoperability, standards or other technical considerations.

One barrier that presents a far greater hurdle to seamless, globally dispersed resourcing is legislation, or lack thereof.  Without reliable and unanimous legislative measures, it is difficult to perceive how businesses will be able to procure services in this ‘any time, any place’ fashion.  Other factors are quality assurance and a mismatch of purchaser expectations and the priorities of those individuals on the other side of the cloud providing the service.

This disconnect can create negative consequences.  An parallel example is the sub-prime mortgage market, which you can arguably call a prototypical cloud finance infrastructure.  The chief reason for the disastrous consequences of this model was the way that loans were chopped up, securitised and the relationship between debtor and creditor completely divorced.  These cloud models for brokering resources seem prone to the same effects.  Indeed, on the finance subject, many P2P finance websites that aim to match borrowers and lenders now try to emphasise how there are ‘real people’ borrowing and lending precisely because of fears originating as a result of the sub-prime crisis.  Establishing a working relationship with hundreds or even thousands of cloud workers isn’t practical and the alternative seems to be a ‘jobs board’, but these don’t solve the scalability issues for clients of metadata services (or any other more complex human resource problem).

There are some services, for example, keywording and metadata where a cloud model has been employed to try and take advantage of concepts like crowdsourcing to solve short term HR constraints.  Do they work?  The photo library, Magnum previously used this model to tag a lot of images but could attract interest in the project from the labour pool with their prominent brand.  For a corporate or public sector organisation with a less glamorous asset cataloguing requirements, the up-take is likely to be much lower.  As others have observed, the quality of service can be poor and may need to be carried out again by professional picture editors.

In summary, the research conducted by Forrester and others is both challenging and valuable.  It illustrates to those of us who live and breathe the tech industry some exciting new possibilities for business.  However, until some of the fundamental obstacles outlined above are solved, it remains just that – research, pure and simple.

Share this Article:

Leave a Reply

Your email address will not be published. Required fields are marked *