Tony's blog

Frank Talk: Data Domain CEO Discussing 2009

I had a discussion with Frank Slootman, CEO of Data Domain. We talked about Data Domain’s outlook for 2009, what is working for them and the trend of more and more customers moving away from tape. 

Tony:  Frank, tell us about Data Domain’s business outlook in 2009.

Frank: We gave guidance to the capital markets – a lot of people refuse to give guidance to the capital markets. For us – it was important to give an indication of what our operating assumptions both for Q1 as well as for all of 2009. Take a look at that guidance. 

Obviously we are getting into the laws of big numbers. Its heavier lifting growing from zero to one million than it is from 100 million to 200 and 200 to 500. On the other hand we are guiding to significant growth even in 2009 – where the overall storage market is going to compress probably by 5-10%. 

We are planning on growth and we are hiring. We have the advantage of running on a secular trend – as opposed to a macro trend – that’s a generational shift from legacy tape infrastructure to one that is disk and network based. We think we are in the first or second inning of this ball game – very low penetration. We will have head winds like everybody else but we think we are in a good category and still considered by customers even in this economy. 

Tony: That is an important point – you believe there will be growth with Data Domain while the rest of the storage market is planning for the opposite. And keep in mind that 2007 and 2008 weren’t great economies either. Nothing compared to 2009 but the last few years haven’t been anything to write home about and you still have seen amazing growth. 

Frank: Yes and the reason is we sell pretty hard on ROI and payback. In other words, at the end of the year the customer should be better off with this new infrastructure on a cash basis – you should have more money left at the end of the year for having gone with Data Domain as opposed to just staying the status quo. Net new incremental spend is only justifiable if on a cash basis I am better off. 

Typically where people are making their investment back they are not spending as much money on tape media, they are not spending as much money on record management services – you start adding that up it starts paying you back pretty quick. When people have investments in legacy tape the maintenance is really high and it gets higher every year. You very quickly find a cross over point that people are saying it makes sense to go with Data Domain. 

Tony: IT professionals are telling me – no more investment in tape. 

Frank: I call that drawing a line in the sand. You get a mandate from above that says – it isn’t a little bit here or there – they are going to make this change.

Tony: There are also other benefits - like not putting tapes on trucks. 

Frank: Right – no hands. 

Tony: Yes and there is a security and administrative aspect to this.  But one other thing that is harder to quantify is the recovery issue.  This might be harder for them to sell their management but for the IT guy it is a huge win – recovery is fast and reliable.

Frank: And they can have deeper retention. They can have far more data online than they could have otherwise. Individual file restores and you can do full system restores much faster than they could before.

Tony: Data Domain has been saying this for years but now the customers are saying it.

Frank: That is right. Recovery speed and reliability isn’t what CFOs embrace because they are really looking at the measurable economics. They want to see how Data Domain pays them back within a certain amount of time. The other stuff is all gravy.

Tony: I am surprised at how many people I am seeing moving away from tape. I was skeptical it would ever happen. 

Frank: Look at the numbers. IBM was down on tape by 31%, Sun was down 15% and Quantum was down 25%. The decline is accelerating. For a while they were going sideways but now they are dropping off a cliff. 

Tony: I am also seeing Data Domain becoming the de-facto standard.

Frank: We are not going to get too big for our britches. We always want to feel like we are the underdog with these giants around us. We are not going to get too cocky. 

Tony: Right, healthy paranoia – according to Andy Grove.

Frank: Yes and you need it. I find a lot of products in the market to be inferior and not doing customers any good. People often buy because they have relationships and they trust their vendor to deliver a good product but often this is not the case. It is often that it’s just downright bad. 

Tony: What’s next for Data Domain?

Frank: We are going to keep on trucking. We will continue to build bigger, faster and more broadly capable systems. We will be a very broad class of storage that is highly optimized. That’s our game.  Our dedupe is a bone crushing technology that is enabled in a fully realized solution. 

Tony: My view is that the way VMware has virtualized servers - Data Domain has virtualized data. 

Frank: That is exactly right. 

Panel on Storage Optimization

I just participated in a panel on storage optimization for Pacific Crest - an interesting firm that focuses on technology research for their investors. I was one of the panel members alone with Stephen Foskett, an IT consultant that works with end users and Mark H., an IT manager for plastics manufacturing company with 3,000 employees and 40 locations. 

Both Stephen and I talk to a lot of IT professionals and it was good having Mark H. on the line for his direct input on his IT operations.

Here were some key takeaways:

1. Both Stephen and I are seeing many of the same things - IT budgets haven't been dramatically cut but spending has slowed down and some projects have been put on hold for now.  However, high priority projects are moving forward since business can't come to a halt. Mark verified this within his IT organization. In the last 30 days I've personally spoken to about two dozen IT pros - and that has been the universal position among all of them.

2. Primary storage and backup are seen as mandatory. In many cases IT professionals are taking a look at what they have on the floor and seeing how they can better utilize existing IT assets. 

3. End users are more open to moving away from the status quo and leveraging innovation if the economic value is compelling and quantifiable. For example, Mark H. actually is eliminating tape and instead using Data Domain for backup. This is becoming more of a common theme among the Data Domain customers that I speak with. There is an opportunity to make big changes in the data center based on economics - especially in 2009. This isn't just true in backup but across the board.

4. Stephen and I both agreed that what we are seeing is continued storage usage - which includes optimization of what we already have and potentially buying new capacity.  Mark pointed out that storage was core to their business and was not being cut or delayed. 

5. Another area that is core to IT operations is backup and that investments are being made in backup hardware - with data deduplication being key to data center optimization. 

The Leader's Dilemma

There is an interesting dynamic in the business world - when you are not the leader in a category or segment you go out of your way to make some leadership claim. And when you are the leader making strong statements about this position is often perceived as boasting. 
 
Data Domain is the leader in dedupe disk-to-disk (D2D) backup - they just don't say it often or loud enough. Beyond the marketing impact of this reality there is great value to the customer. Data Domain has the largest number of installs with years of field experience. In the IT infrastructure world quantity is quality - because the market will throw anything and everything at you.  The more your product is in production environments the higher quality it will be - and there will be fewer unknown variables because you've been exposed to more and more of them. Additionally, Data Domain has built an organization and ecosystem that understands the solutions and issues for D2D backup - it is their core competency. 

Steve Duplessie, senior analyst for the Enterprise Strategy Group blogged about IT leadership recently,  Steve states "The Blue Ocean concept is absolutely true, yet 90% of all companies will spend far too much time and effort attempting to compete in a market with an established (perceived or real) leader..."
 
Data Domain went after a Blue Ocean opportunity and has become the leader - and others are following in order to compete.  According to Steve and his years of observing IT markets - it is very unlikely that anyone will dislodge the leader.  This isn't his opinion but an analysis based on historical trends.
 
EMC is the one big player going after this market in earnest with three different products (which will confuse the market and themselves) and they will make progress because they are a giant.  But I submit that they will continue to face the issues that Duplessie points out in his blog - the leader creates the Blue Ocean (Data Domain) and the number two guy started the race too late and will never catch up. 
 
The number two guy tends to throw mud and go on the attack. An EMC blogger did this recently comparing Data Domain and Avamar. The blog is overly biased trying to bash Data Domain and shine a bright light on Avamar. But look at the comments posted and you will see end users attacking the content of the blog because it is - well - ridiculous. 
 
I just had a conversation with a big EMC customer that went with Data Domain after giving careful consideration to everything that EMC had to offer. Even though EMC is a strategic partner they didn't want to go with Avamar or CDL:

"We didn't believe in the VTL paradigm. We were really looking for stuff that was native NAS, either CIFS or NFS protocol, so to keep it "simple stupid" kind of mentality.  And there's a lot of vendors that have since integrated even with Data Domain with proprietary APIs to get a better performance out of the Data Domain. But the tradeoff there is kind of performance versus openness.  So we looked at Avamar with EMC. They had just acquired them about the time we were looking at Data Domain. We looked at a few others, but they really weren't big name players.  We searched out there and we went to people that we typically do business with. The real issue of the Avamar solution was that it was a complete replacement rather than an augmentation of the existing environment.  

Data Domain just stepped in and instead of tape, now you're writing to a disk-based target.  And it was fairly easy to set up, because it's CIFS and NFS. Avamar was 'rip and replace' both the software and the hardware. They're a combined platform. So [Avamar] was a much more unilateral replacement if we wanted to go there.  We thought Data Domain was less invasive in the environment at the time."

This is common sense thinking that is the majority view. That doesn't mean that everyone will go
down this path - there are always exceptions.  
 
From an end users perspective Data Domain is the leading solution and the product works extremely well, there are thousands and thousands of happy customers, it is easy to install and the value is quantifiable. What other IT products can you say that about?  
 
Another big false negative that competitors will say about Data Domain is that it is a one trick pony.  So was EMC for years with the Symmetrix and that strategy made them billions of dollars.  Eventually they added more products to their portfolio but it took years. NetApp is still fundamentally a single product company and that has done extremely well for them. And even though EMC claims they are the number one NAS vendor (nod, nod, wink, wink) we all know that the real leader is NetApp. I predict they will make the same claim against Data Domain going forward. And Data Domain, just like NetApp, has an extremely healthy business with impressive growth rates and high margin products - that is clear to measure. 
 
I have a very different view than the naysayers - Data Domain has the leading product with no clear and strong competitors in a market that has endless opportunity. Pretty much every company has budget for backup devices and it is core to their operations. This is a multiple billion dollar market and Data Domain is the leader. Additionally, you will see more and more customers using the same Data Domain appliances for archival storage. This is a powerful value proposition - using the same system for backup and archive and the combined dedupe ratios are compelling and hard to beat. 
 
The leader's dilemma is that every competitor will come gunning for you - but that is okay - because they invariably miss. 

Pondering The Future of Tape

I once thought the notion of tape as the primary backup target being replaced by disk in the data center was not likely for a long time to come. However, data deduplication has enabled this to occur in many data centers over the last few years. There are literally thousands of companies that use disk-based systems as their primary backup solution and tape is secondary. An interesting dynamic occurs when you implement D2D backup - you begin to reduce the importance of tape as a part of the backup process. In many cases tape is just being used but for weekly or only monthly fulls. Additionally - tape is used for off-site storage for long term vaulting. But even this process is being replaced by leveraging remote replication of backups from the D2D backup device to the off-site vaults (e.g. Iron Mountain). 

I once thought the notion of tape being reduced and eventually eliminated in the data center was going to take a very long time to occur. But I see it happening at an accelerated rate - in companies of all sizes. There are really big companies that are reducing, big companies that are planning on elimination and many of the SMBs either have or are well on their way to eliminating tape completely. The market indicators support this with IBM tape 2008 sales down 31% and Quantum tape sales down 25% compared to 2007. 

I have actually been conservative about the decline of tape, but there is a trend occurring and it is happening at a rapid pace. Tape will exist for the foreseeable future but it will become more and more marginalized - being ultimately relegated to a small part of the market. Certain members of the IT community hold onto things for a very long time - and that will be true of tape - but there will be numerous companies that will move away from tape completely - and there will be new companies that will never use it. Someday soon it will be like when your kids ask you what records and cassettes were.  The next generation of business may never know what it was like to backup to tape.