Problems with Total Cost of Ownership

By Jonathan Morgan

For someone to sit down and argue that one technology is less expensive than another is a foolish way to work and here is why…

Let’s start at the middle. In general, I’ve been glad to see that over the past decade, the argument has moved from the headline cost of buying the nuts and bolts of the archive solution to “total cost of ownership” – however, we all know there are some major problems with that:

1. There is little if any consensus about what makes up TCO
2. Most TCO comparisons are woefully weighted towards one solution or another and/or extremely limited in scope
3. You pretty much have to break laws to even publicly publish/compare price lists (how crazy is that!)
4. Most importantly, TCO is only PART of the story – why then is it treated like the nirvana to reach in calculations?

How can we calculate TCO? How long should an analysis period be? Can there be a single formula when no two organisations are alike?

TCO is a Predictive Model – That’s Problem #1

By looking at TCO we open up a whole can of worms; Nostradamus might have been pretty good at seeing into the future, but, whilst in hindsight we can see who the prophets were; you kind of need to be a prophet yourself in order to correctly chose who the prophets of today are! So I don’t like to guess too much about the future, other than to extend trends seen today.

Perhaps the hardest part to put down in a TCO model is the analysis period. 5 years? 10 years? 100 years? They make a massive difference to the end calculation.

100 years? It would seem ridiculous to me try to base a comparison on 100 years. But it’s an easy mistake that’s made. Easy that is, because for sure, in the case of something like a film archive, we want to plan on keeping the films we make for at least that period, but totally ridiculous, because of course archive technologies are going to evolve beyond what we can imagine today in that period.

I’d go for 10 years. But that’s a generalisation. Every situation is different.

Why is the Storage world Stuck on TCO Anyway?

I don’t like TCO in isolation.

There’s also something I call TOO: Total Opportunity of Ownership.

There’s an opportunity being missed every time a piece of data isn’t re-used, there’s an opportunity to reuse data in a variety of ways, an opportunity to take all those creative assets and to make them more available for internal and/or external reuse. Digital assets can be veritable gold mines.

So move aside TCO. And welcome in TVP (I’m in acronym mode today!)

The Total Value Proposition = TOO - TCO.

The value of an archive shouldn’t be measured in its cost – it should be measured in its value, which equals the opportunity it presents less its cost.

Menno talks about that in his 5 minute video about the 25 years of archived broadcasts at EO in the Netherlands here:

Back to TCO, what should it include?

To repeat: no two organisations are the same. But nevertheless, a few pointers to some of the things that I include when calculating TCO:

• The Analysis period
• Initial storage requirements
• Annual growth requirements
• Base hardware and software costs
• Capacity % utilised of storage media
• Annual predicted % change in cost of media
• Replication cost (comms link vs manual actions)
• Employees required / costs / % annual change
• Energy usage / energy cost rises / better future energy efficiency % changes
• Servicing costs / employee training
• Failover facilities

Comprehensive enough? Probably not! Some might want to simplify the TCO equation, but to do so is to risk making a TCO argument that doesn’t calculate TCO at all.

NOTE: in any period of analysis above 10 years – the overwhelming cost turns out to be not hardware, software or energy, but (if there is a bit of a difference in the amount required between two solutions): manpower.

What Should go into a TOO Argument?

The benefits of one archive over another may well outweigh the TCO argument by a large factor.

What if one solution allows better, more frequent reuse of data over another? What is the yearly, monthly or even daily benefit of that?

Does the solution allow you to reuse your pool of created intellectual assets for new purposes, either internally or externally? Can a value be put on that?

I’ve captured all of the above calculations (and more!) into a spreadsheet – you can download it here.

In any analysis period of more than a few years:

The overwhelmingly significant numbers that affect the TVP are the manpower required and opportunities gained/lost. After that the picture comes down to a straight head-to-head.

Archiving Film

Not all archiving is alike. Take film and broadcast production. A film today can be stored on tape but that task is growing in size and complexity.

Traditional complexity of storing film on tape is well known. Once a film has been shot it will be stored as film on an original colour negative (OCN). Generally, the OCN is considered too delicate to be used more than absolutely necessary, so it needs to be stored, at correct temperatures, and in multiple locations. Film also needs to be stored in conformed OCN, Color Match, and Color Separation type formats. Film must be handled carefully, stored at correct temperatures, re-scanned, fetched, stored, restored… etc, etc: all requiring manpower.

And the amount of data to be stored is growing. Take Pirates of the Caribbean III as an example, which is said to have used up 90TB of disk space. Pirates of the Caribbean II (same basic technologies) is said to have taken ‘only’ 60TB. Add to that changes in technology and it is clear the figure will continue to grow:

One hour of uncompressed ultra high dev video (UHDV) consumes 12 terabytes. In-use 3D videoing techniques will consume even more since they take multiple shots from the same position.

I could go on (and I do!) but I’m sure you know the story of HD, etc.

It is straightforward therefore to conclude that film and broadcast will continue to push at the limits of what can be achieved by an archive.

So, what’s the Right Solution?

• Total Value Proposition = TOO – TCO
• The analysis period is a key input to calculating this
• Manpower and Opportunities Gained/Lost could outweigh all else
• Avoid solution/vendor tie in that don’t allow you to easily move on when that times comes
• Do not pay a premium for hardware/software – you may need a lot more storage in the years to come

Of course headline cost matters.

But we’d argue strongly that as data is captured digitally, edited digitally, so of course, store it digitally and the opportunity of reuse will outweigh all other factors.

External References

What’s in a terabyte:
3D broadcasting is here:
Menno video: