Taylor's Take on Sun Storage

Mapping the DNA of our customers...

The Human Genome Project (HGP) has been credited as one of the great scientific feats in history when it announced it mapped the human genome in April 2003.

We don't have any molecular biologists on staff - but we think we mapped 70.1% of what makes up our Customer's Experience with Sun Storage Products (Give us a couple years to map the rest, it took 50 years to map the genome after DNA was first discovered after all...) 

This is important because if you are in the business of trying to make customers happy, then you need to know what makes them happy (or not happy for that matter.)    

The person in charge of our "Customer Experience Genome Project" is our Research Manager, Hernando Gonzalez, a Ph.D. in International Development and research statistician at heart.  You can read some of his published work in Quirk's Marketing Research Review.  In fact, if I were to guess what Hernando is thinking right now - it would probably look something like this:
How to Map Customer Experience

At the risk of spilling our secret sauce recipe for customer research over the Internet, this is what we did...

First, we sent a survey to more than 40,000 Sun StorageTek customers in 24 countries (and translated that survey into 14 different languages).  Then, we asked each customer to pick just one Sun StorageTek Product to review and answer questions about (product features, services, sales surrounding it, delivery, etc.) Finally, rigorous analytics like Factor Analysis were applied to the responses. 

Here is what we found (and I think you will quickly see that this can be applied to multiple industries)
 

• 19% of our customer's satisfaction or experience is determined by the type of relationship Sun has with a customer account.  How well sales anticipates needs, whether customers view Sun as a strategic business partner, etc.
• 10.9% of the experience fell into the product deployment and operations category - how well the product integrates, how easy deployment was, how easy upgrades are.
• 10.5% of our customer's experience is impacted by product RAS (Reliability, Availability, Scalability).
  
• 9.9% of the product experience is determined by how our sales folks communicate with a customer before and after a sale, as well as how well we do in keeping customers apprised of new technologies.
• 5.8% is impacted by product fulfillment - product delivery and install.
• 5.3% is impacted by ongoing technical support or customer support services. 
  
• 5.1% (and this came as a surprise) is impacted by the invoicing process - order accuracy, clarity, timeliness, etc.  (I guess when you are dealing with budgets and cash flow, the invoicing process becomes increasingly important.)
• 3.6% is impacted by how we communicate and deliver our future products - our roadmaps.

And finally, our analysis tells us that there is 29.9% of "something else" that dictates whether a customer has a good or bad experience with us.  ~30% unknown is higher than I like it, but mapping about 70% of the DNA of customer experience isn't a bad start...        

Measuring (and Improving) Customer Experience   

The next logical question is, "If this is what impacts a customer's experience with Sun Storage, then how is Sun Storage doing?"   To get this answer, we simply have our customer's rate us on a 0-10 point scale - 10 being great and 0 being not so great.  We can then see how Sun is rating with customers on each of the major categories above (by product line AND geographic region).  

But here is where my sharing of the secret sauce stops - since this blog is public, and our customer scores are confidential.  (But, if you are a Sun employee you can log into SWAN to see how our customers rated us - if you are not with Sun, you will just see a "server not found" page).

Now for the hard part... 

If the above work seems like a lot - it was.  But we are only half way there.  A large step in the right direction is knowing what impacts customer experience, satisfaction and loyalty on a macro scale (nothing replaces a good 1:1 with a customer btw).  The next step is to develop and implement plans to maintain the categories we are doing well in; and improve the categories we are not doing well in.

So, as we move into our new fiscal year, you don't have to guess at which areas we'll be focusing on...  

Posted at 05:04PM Jul 09, 2007 by Taylor Allis in Storage Intelligence  |  Comments[0]

Lacrosse and the future of storage...

I like storage, strategy and sports - so I really liked Scott Tracy's "Telegraph" Blog.  I commented on it, but will elaborate further here.  He talks about Storage running on general-purpose Solaris, and shows a nice OpenSolaris Storage Platform diagram. I'll serve up my own sports analogy (and re-live the glory days while I am at it...)

I played Football and Lacrosse in high school and college  (yes, I was a UPS Logger) - but I'll stick with the Lacrosse theme. 

In high school my team competed in Florida's Cocoa Expo Lacrosse Tournament. We were quickly mocked by all the teams there - they represented the best of east coast lacrosse and we were from cow-town Colorado. We were at first intimidated by their "moves" - quick sticks, behind the back passes, etc. But our coach told us to stick to the basics - two hands on the stick, straight forward passes, etc.

What happened?

We won the tournament that year - first place, #1 (and went on to win State by the way). So what's the storage tie in?

There is an appliance or point-product for everything in storage today - virtualization, data movers, back up, CDP, encryption, etc.  A lot of fancy stuff that can solve individual problems, but adds to the overall complexity of IT storage. I like the OpenSolaris Storage Platform approach because it sticks to the basics - data volume management, data services, file systems - open and residing at the heart of any system or solution - its OS.

That's a winning strategy...

Posted at 02:12PM Jun 19, 2007 by Taylor Allis in Storage Intelligence  |  Comments[0]

Does a complete portfolio matter???

When you look at Sun's Storage portfolio, it's unique in two aspects: its breadth and quality.

In fact, after the StorageTek acquisition, Sun is now one of only two companies in the world that truly offer an end-to-end portfolio of storage solutions - see the table below:

But this begs the question..."In a world where customers can buy from multiple vendors, why does a complete portfolio matter?"  I heard the best answer from our Storage Technologist, Randy Chalfant.

The answer is best illustrated in how the world's data is distributed. According to analyst firm Horison Information Strategies:

• 5% of the world's data resides on Enterprise disk (Used for OLTP, Database, and Data Warehouse applications)
• 10% is on Midrange Disk (Used for File Serving, Web, Imaging, Video, CAD Fixed Content, and Reference Data)
• 20% is on Entry or Capacity Disk
• 60% is on Tape (Used for archive, back-up, and disaster recovery).

The pyramid diagram below shows the distribution: 

Customers really start to see the significance of this distribution once dollars per gigabyte ($/GB) are applied to each storage tier. Horison calculates the $/GB of each media type in 2007 as:

• $43/GB for Enterprise Disk
• $25/GB for Midrange Disk
• $9/GB for Entry Disk
• $2/GB for Tape

So, a vendor that offers a complete storage portfolio can support all applications at a cost that meets budgetary needs.

So what happens when a customer engages with a vendor who only sells disk as their primary hardware platform?  (I can think of three major ones: one starts with an "E," another with an "H," and a third with an "N" - not to mention a million disk-only start-ups).

The fact is, without a complete portfolio, storage costs on average 69% more (yes, using the numbers above - 69% more) across the Enterprise. The below graphic demonstrates Sun's portfolio advantage compared to a disk-only vendor:

So, next time you hear a vendor say they offer complete solutions - be cautious.  And, next time you hear someone say something like "Tape is Dead" - now you know the real reason why they said it...they don't sell any.

Posted at 10:55PM Jun 16, 2007 by Taylor Allis in Storage Intelligence  |  Comments[0]

Redshift Storage...

In astronomy, "redshift" refers to when the Sun's light is reflected off of an object and is shifted towards the red end of the electromagnetic spectrum.  Our CTO Greg Papadopoulos described what Redshift means to Sun Microsystems and its customers at the '07 Sun Analyst Summit.    

In a nutshell, Redshift refers to a segment of IT users applications whose computing needs far outstrip Moore's Law (i.e. the doubling of transistors every 18 months).  While processing power is adequate for some of today's applications, a segment of customer applications need more than what Moore's delivers.  

The Redshift segment includes last-mile bandwidth provider applications (BW), high-performance computing (HPC) applications and on-demand Internet application providers (*Prise).  Sun is meeting the Redshift need through projects like multiprocessing, crossbow and multithreaded networking cards.  The illustration below shows Moore's Law and where our Core customers (Green Line) sit vs. our emerging Redshift Customers (Red Line).

 

So what about Redshift Storage?
To see the relation between Core and Redshift customers, applications and their Storage needs, we drafted a model "borrowing" Greg's methodology - but using Storage numbers:

Moore's Law for Storage:  This has been mapped in terms of Hard Drive Capacity, but let's use the same method as the diagram above.  All those 1Ms and 100Ks are actually how many MIPS (million instructions per second) a customer could buy for $1,000 USD at a given point in time. 

To determine "Moore's Law for Storage" let's look at how many Gigabytes (GB) customers could buy for $1,000 over time.  (And since Tape and RAID storage differ in price and density, let's look at both lines over time): 

Core Storage Demand:  Now we need the Core customer demand line, the Green line.  To best capture demand, let's look at a metric that sums up what our customers deal with every day - "How much storage can a singe IT administrator manage?"  Fortunately, Horison Information Strategies tracks such a metric.  So below is our new chart - with Moore's law vs. Core demand:  (We added the Tape and RAID "blue" lines together for simplicity)

The Information Management Gap:  The most interesting difference is that the Core "green" line is below Moore's line in Computing; but the Core line is above Moore's line in Storage.  Simply put, Storage technology has not kept up with Core storage demand.  This constitutes the Information Management Gap - customers need to manage more storage, but storage technology has not kept pace with the explosive amount of data the world is generating (ever heard that one before?).  Also, when you look at our emerging Redshift customer applications - needs are even greater:   

So what's the Solution?  A great first step for anyone everyone struggling with Information Management would be to simply determine how equipped a "data center" is in managing data.

This current market trend has also brought terms like "virtualization", "partitioning" and "volume management" to the top of IT manager's minds: Storage Virtualization helps grow systems without disrupting operations and Virtual Tape speeds backup and recovery without changing backup infrastructures.  Technologies like Thin Provisioning help utilize more in a storage system.  In addition, file systems are gaining more and more data management features that help IT admins with Information Management.

But the intersection between Information Management and core storage needs like application support, data protection, business continuance and archive requirements is where customers need the most help - We'll cover what Sun is doing in these areas in later blogs...

(Inside Sun, see our Redshift Storage Wiki)




 

Posted at 05:54PM Jun 08, 2007 by Taylor Allis in Storage Intelligence  |  Comments[2]

"man it feels good to win"

Sun midrange disk grew 38.6% in revenue Y/Y - and this quote came from an internal e-mail from a Sun Storage Product Director (not surprisingly...)

The number came from IDC's Worldwide Quarterly Disk Storage Systems Tracker results for Q1 CY07 (released two days ago).

I have to give credit to the midrange disk storage team - over the past year they have focused on driving customer value through clear value propositions, vertically focused solutions and training sales and customers on how to benefit from our disk systems.   (It also doesn't hurt to have a product that is the world record holder for the SPC-2 price/performance benchmark - hmmm, why aren't EMC products on this list?)

Some highlights from IDC's report:

• We showed our 5th straight quarter of Y/Y revenue growth in overall disk systems, growing 7.4%
• Our midrange disk (classified by Sun as systems priced between $15K - $300K) grew 38.6% in revenue - this is our Sun StorageTek 6000 family, and specifically, the 6540 did very well
• Sun retained the #1 Worldwide position in UNIX disk units shipped - for the for the 14th consecutive quarter (And seeing how Horison Information Strategies reports that today, 47% of the world's data is generated by a UNIX OS - this isn't a bad position to hold...)
  

Posted at 10:58AM Jun 08, 2007 by Taylor Allis in Storage Intelligence  |  Comments[0]

Tape's Dead? Heard that one before...

This article was e-mailed around Sun Storage just the other day: And tape walked quietly from the spotlight (applause)

I worked for a founder of StorageTek who said he first heard the "Tape is Dead" argument in 1987.  I understand the argument, especially with disk prices dropping, recent VTL growth and the entrance of data de-duplication.

But I don't agree with it - it's far too short sided, and usually the person saying it has not done their homework. The best approach is not to pit Tape vs. Disk - they should be used together to accomplish storage goals.  Makes perfect sense to me.  To see why certain vendors like pitching "Tape is Dead" articles - see Randy Chalfant's Blog.       

But to those who say tape will be exiting the data center, we have this to say: Tape's value proposition remains maturity, longevity, volumetric density, declining costs, removeability and scalability (and this is from someone who supports tape AND disk systems at Sun). 

But for those still pitting Tape vs. Disk, I'll put forth these comparisons calculated by our disk analyst Bruce Norikane and tape analyst Chris Ilg (calculations made from public website info and do not include software costs):

Which solution would you recommend for a petabyte archive?

 
To further illustrate this point:

• According to Freeman Report's Whitepaper, "The Growing Importance of Archive"...37.1% of all archived data reside on Sun StorageTek solutions
• This equals 11.924 exabytes (to be exact)
• Storing just 1 exabyte of data on EMC's Symmetrix DMX-3 disk array consumes around .97 billion kilowatt hours (KWH) per year (according to their website)
  
• So, if 11.924 exabytes of data were stored on EMC disk:
  • It would consume 11.5 billion KWH per year (enough to power the city of San Jose)
  • This would cost over $1B ($1,488,000 to be exact)
  • And you would have to burn 3,768,885 tons of coal per year to generate enough power

An insider's view: For my last point, I'd like to share an insider's view as to how a vendor can use market data to tell a story they want to tell. (credit goes to Chris Ilg for this great insight)  This is particularly relevant to the "Tape is Dead" presentations lingering out there.  Take a look at the below graphic from a customer presentation.  Both charts are based on IDC data - but the first chart is one you would see from a disk-only vendor, telling you that disk is converging on tape costs.  The second chart is closer to reality - the rate of price erosion for disk and tape is about the same.  The only difference in charts is the scale, the X axis - the first chart gives an absolute view, the second a relative view.  

 
So, as long as we need to store more data at lower system and power costs - Tape (along with cheap disk) has a life. 

Lastly, I love Randy Chalfant's take on Sun's approach to the disk vs. tape debate posted in his blog linked above: 

 "Part of our approach to this, is to lead with a business value
assessment, where we can look at the operational needs and practices of
you business, and then map the technology into a supporting foundation
for the demands of your business."

You can't do that if you don't have tape and disk (not to mention software, services and servers) in your portfolio...
 

^{(Note: If you are a Sun employee, you can find more in the Market Fact Book section on our Storage Intelligence Wiki)}  

------- Update --------

Another disk-only salesman brings up the old argument, see:

• Marc Farley's blog & video:  Is tape dead? What do you use it for? 
  
• StorageStop picks it up at: Tape is dead, long live tape
  

Posted at 03:57PM May 30, 2007 by Taylor Allis in Storage Intelligence  |  Comments[0]