OK, I’m expressing a little frustration. Every time I see the term “collaboration,” I shudder. Why? The term is actually worse than “cloud” or “big data” or other terms that end up obscuring the products and vendors in a market. Ask five people what they mean when they say “collaboration,” and you will end up with five completely different definitions. Am I taking this personally? Well, yes. I am trying to collaborate every day. I am creating content that needs to get out to peers and the public. I am trying to get the word out to colleagues in UK, France, Germany, Belgium, Singapore, Tokyo, down the hall, next door, at the desk next to me – and failing. Or at the very least, spending about five times as much time as I think I should be.
As you speculate on who’s buying whom, who’s breaking up, who’s on the rise, and who’s in trouble, consider the Enterprise Technology stack. Hats off to Geoffrey Moore and the EMC M&A team who first taught me to think about the tech landscape in this way.
Technology markets are often controlled by “gorillas” or market leaders in particular parts of the stack. Consider the stranglehold Microsoft had for a long time in computer operating systems (dramatically reduced by Linux), or the continued dominance of Oracle in business databases (gradually being challenged by open source and the Cloud). Showing who has what in a stack orientation allows you to quickly scan gaps in a portfolio (that my need to be filled), alliances that may need to be formed, niche players that are being marginalized, etc.
This is a version 1 and would appreciate input. I had a tough time deciding whether to split Cloud vs. packaged/appliance into its own set of layers or not. I kept them as options within particular layers (e.g. database), but it may be more helpful to split them. In any case, here’s version 1. Let the speculation continue!
I’ve recently come across the statistic that “90% of the world’s data has been produced in just the last two years.” However, I can’t find the real source for this statement, so I’ll try to quickly break it down below:
In 1997, Professor Michael Lesk, the Chair at the Department of Library and Information Science at Rutgers University, made the statement that all the world’s information amounted to about 12 Exabytes. This is from looking at “traditional” information in the form of cinema, images, broadcasting, sound, and telephony.
If we look at the last two years, the leading sources would be IDC with their Digital Universe study (sponsored by EMC), and the University of California, San Diego’s Global Information Industry Center. IDC is more of an apples to apples comparison, and they indicate that in 2012 and 2013, 2.8 and 4.4 zettabytes were created respectively.
If we directly compare 12 Exabytes with 7,200 Exabytes, 99.8 percent of all information was created in the last two years.
This bears additional investigation, but at least we’re now talking numbers. ;)
The headline of the latest IDC storage report says a lot:
Weak High-End Demand Results in Worldwide External Disk Storage Systems Revenue Falling at Rates Not Seen Since 2009, According to IDC
A tweet from Dave Reinsel, a Group VP at tech analyst firm IDC was even more evocative:
Three years pass and twice as much capacity is shipped at the same revenue. This is not just Kryder’s Law of disk density at work, this is an inflection point in the type of storage being purchased.
Like in my attempts to size other markets, it turned out to be hard to gauge just how big the top video sites are globally. None of them publish the same metrics, and the analyst firms often only cite North American numbers. Taking matters into my own hands, I combined Sandvine’s interesting peak bandwidth consumption numbers (the basis for many articles about the growth of Netflix) with Cisco’s annual Visual Network Index, which forecasts the overall bandwidth consumption in each geographical region. Some surprising results:
- Youtube is #1, consuming 7,875 petabytes of bandwidth a month – this is no surprise given a 10 to 1 lead in active monthly users over any other site.
- Netflix is #2 at 6,103 petabytes a month – this is surprisingly close, and shows Netflix’s impressive growth, as well as the effect of full length HD shows and movies.
- Bittorrent is #3 at 3,862 petabytes a month – Bittorrent use has been dropping over time, and while some of this still impressive number is video (some say as high as 85%), it includes all types of content being transferred.
- iTunes is a distant #4 at 817 petabytes a month – this is all downloads, and includes other media. This is an interesting data point reflecting the limits of the rent/buy model versus a subscription streaming service.
- Dailymotion is #5 at 489 petabytes a month. This has some heavy extrapolation since Sandvine only lists the top ten vendors in each geography, but I think it’s fair given Dailymotion’s still large monthly active user base. I will attempt to validate this number further.
- Amazon Prime Instant Video and Hulu should be in the top ten, but one needs to remember that they are not yet available broadly around the world.
Last week, we presented a counterpoint to the recent negative press on IBM. I’ll take it a step further. I believe that IBM is actually the vendor that is closest to providing “Enterprise Cloud.” But what is sold, how it is sold, and the vision of the end result still needs a lot of work.
I’ve both competed against IBM at EMC and worked with IBM as a partner. When competing, our fear was getting maneuvered by IBM out of the deal when they went right to the CxO. When partnering with IBM, I was frankly surprised and disappointed at the fumbling. As a shareholder, I would have been livid.
- Because of internal turf issues, IBM reps got stuck on whether to propose product, services, or cloud instead of focusing on what the customer wanted
- Because differentiated core technologies like GPFS were positioned as products instead of part of a strategy and delivery model, they lost in feature-to-feature comparisons
- There were repeatedly missed opportunities for leadership, where the customer was looking for a new vision of computing, blending attributes of traditional enterprise technology with the cloud – instead, the customer got generic positioning (or worse, marketing-speak)
I have three recommendations to address these issues.
There’s a recent survey by IDC in which the vast majority of enterprise respondents name IBM as the vendor able to most effectively provide Infrastructure as a Service (IaaS). Surprisingly (sort of), the megascale public cloud providers, Google, Microsoft and Amazon Web Services come in 5th, 6th and 7th respectively. As a former AWS employee and cloud analyst who firmly believes that a public cloud with essentially unlimited scale, relentless consistency and automated metered service is the “real” cloud, I generally agree with the sentiment that the old line IT companies like IBM and HP have fallen way behind. But from a business strategy and perception viewpoint, it might be a different story.