The decline of traditional enterprise storage and the unstoppable rise of software defined
In 1991 Geoffrey Moore published ‘Crossing the Chasm: Marketing and Selling High-Tech Products to Mainstream Customers’. This seminal marketing textbook explored the ‘diffusion curve’ – the rate at which new technology products and services are adopted across a market. The book is still in print today, is still a best seller, and as Tom Byers, Faculty Director of Stanford Technology Ventures Program, Professor of engineering and leading light in the US tech venture capital scene put it a few years back ‘is still the bible for entrepreneurial marketing’. The reason for this is simple: Moore’s description of the way new technologies are adopted illustrates a clear truth: as ordinary mortals might put it ‘he nailed it’.
Even if you haven’t read the book (essential reading for marketing professionals like me is not necessarily essential reading for subscribers to storage magazine) you will be familiar with his terms: the ‘innovators’, ‘early adopters’, ‘early majority’, ‘late majority’ and ‘laggards’.
Innovators are the IT people that take up products with exciting potential whilst they are still in their infancy. In many cases this is unproven, ‘beta’ product best described as a working test with a live customer base; it’s a huge risk compared to ‘off the shelf’ proven approaches but offers something new and unique. Next come the early adopters, the visionary first movers who see the potential in new technology as it becomes workable, and rapidly adopt it for their own advantage. When a technology succeeds with the early adopters, it can then ‘cross the chasm’ into early and late ‘majority’ territory: mainstream acceptance and sales nirvana for the vendor. Finally, we have the laggards, the uber-cautious who take on ’ technology only after everyone else has. But its not called a chasm for no reason: for every vendor who successfully bridges the depths there are legions who fall into the depths to become footnotes of tech history – even when the technology is demonstrably superior (Betamax vs VHS is the classic example: Betamax offered superior picture quality and a smaller storage form factor to VHS, but a small price difference was enough to kill it – early adoption did not give rise to mainstream acceptance).
Moore’s description of technology adoption works perfectly in a ‘me too’ markets where there is little concrete differentiation. Nowhere is this more true than amongst the old giants of the enterprise storage market; take a look at proprietary storage from any of the 20th century’s household names – be that EMC, Dell, HP, IBM – whatever – its a struggle to find something that really makes a concrete, provable difference. They all have management software which is broadly proprietary, broadly comparable hardware, broadly comparable prices, and broadly comparable performance – its the choice between a BMW, a Jag, a Merc or a Lexus: everyone has their preference, but much of that comes down to familiarity and preference: : no model costs half as much as the other at retail, travels twice as fast, or cost half as much to run.
Yet, every so often, Moore’s take on the diffusion curve is wrong. Circumstances combine to create a situation where a technology doesn’t just cross the chasm, it builds an eight lane motor way on a suspension bridge and charges across at speeds which exceed normal limits. For this to occur, two things need to happen:
- The technology needs to deliver a genuine step change. Real benefits and obvious advantages as opposed to a marginal brand differences.
- The problem or advantage the technology delivers has to be big enough to offset the reservations customers on the early adoption side of the chasm feel. The “need’ of the buyer has to be genuine and pressing.
Both of these circumstances are now present in the enterprise storage market, and accordingly the market is going to change quickly and permanently: traditional enterprise storage vendors are facing inevitable decline in their current business models and a ‘change or die’ moment because the advantages of software defined storage are so great as to represent an end-of-an-era step change: when the Iron Age arrives your highly optimized manufacturing processes, slick manufacturing and motivated sales channel count for nothing if you are working in bronze.
So what’s the step change advantage in software defined? Firstly, the cost advantage of software defined storage is huge. Software defined storage separates the physical storage plane from the data storage logic (or control plane). This approach eliminates the need for proprietary hardware, and freed from the need to buy proprietary appliances running proprietary software IT teams can work on commodity x86 hardware and discs in cheap racks, generating as much as a 50% cost saving. So that’s the first of our necessary market conditions for rapid adoption dealt with: real advantage in avoiding vendor lockdown coupled with huge cost savings.
So what about the second condition, the pressing problem or advantage? As a storage professional, I doubt you need to be informed on what the key problems are, as you will be dealing with them day in day out: more and more data to store, bigger unstructured data, and indefinite storage duration. Here’s your top seven storage pain points as measured by 451 Research.
The biggest of these is problems is data growth which is a major problem for more than half of all enterprises, large and small. The next problem is managing the cost – which harks back to my previous point – and the third is capacity forecasting. Put these three together and you are creating a serious budget and management headache. Storage is growing a fierce rate, its difficult to predict how much you need, and the costs are getting out of control.
In these circumstances, given the clear advantages of the technology and the serious nature of the problems faced by IT, is it any wonder why analysts from Gartner to IDC are united in forecasting the unstoppable rise of software defined storage?
In a word, no.
[By Jason Phippen]