Joe Tucci - Chairman, President, CEO
... I want to point out that we did have numerous incentives in place for our sales force to bring in orders earlier. Yet as I said, approximately $550 million of orders came in during the last week of June. By the way, we were confident of our ability to achieve this end of quarter goal, as we had ample bid coverage and firm field commitments to make our plan.
To answer a question that I know is on a lot of your minds, namely, when we announced our intention to acquire RSA on Thursday afternoon, June 29, did we expect to achieve our revenue and GAAP EPS guidance for the quarter? The answer is yes. Again, the quality of customer names along with strong field commitments gave us confidence.
So what happened? The primary answer lies in the wrong mix and factory inventories, specifically with Symmetrix DMX. Now let me state that I absolutely believe this issue was a self-induced execution failure on our part. There is no excuse.
As we were progressing through the quarter, bookings on both DMX 2 and DMX 3 were tracking well relative to their respective plans. During the final week of the quarter, DMX 3 orders literally exploded, while DMX 2 stalled. A significant piece of this phenomenon happened in the last two days of the quarter. These bookings came from a deal pipeline that was larger in size than what we booked and included both potential DMX 2 and DMX 3 deals.
What surprised us was that a very high percentage of deals that ultimately came in were for DMX 3, causing a severe shortage. While we did have contingency capacity to ship more than 100% of our DMX 3 plan, we did not have sufficient systems built to meet all of the demand. So in essence, we had too many DMX 2s in inventory and too few DMX 3s.
Remember, in most cases, these DMX 3 orders had other EMC hardware and software products on the same PO or purchase order; products like Centera, products like CLARiiON, products like Connectrix switches, which in many cases -- indeed, most cases -- we could not ship because of customer contractual requirements to only ship complete orders. Due to lateness of the quarter, we did not have the ability to go back to customers and secure their permission or consent to ship these partial orders.
So what did we learn from this experience? Well, we learned a lot. Number one is we need to make new product transitions faster, even major product transitions. We need to trust our ability to quickly prove to customers that we have done very extensive compatibility and regression testing, and trust our sales organizations to get the transition job done faster.
Secondly, we know we played it too tight with our supply chain. We need to make sure we have ample new product parts and components to build out extra systems.
Third, we kept DMX 2 in production too long. Obviously, making a single product line, or a single DMX product line in this case, is more efficient and far easier.
The good news is we are now through this transition. All new SIM builds will be DMX 3. The best news is that customers love the performance, the scalability, the functionality, and the availability of the DMX 3. DMX 3 clearly is the benchmark by which all other storage systems are judged...
Laura Conigliaro - Goldman Sachs
Thank you. I understand that you are in a mode where you are really trying to build more backlog; and therefore you appear to be holding back more on revenue expense, go-forward actual results. But you really did have a substantial additional backlog. Maybe you can go into a little more detail about why? What is it in the environment or anything else that is causing you to hold back on what you're characterizing as go-forward targets?
As part of that, maybe you can give some more detail on the kind of backlog build targets that you have for the end of this year, and how we will be able to metric that as the year goes on? Will you be giving us more information on this?
What is causing customers to order so late, I can't put my finger on it 100%. Obviously, I have said that I do believe that the environment is more competitive. I do believe there are more bid processes. I do believe there are more kinds of checks in the system. You have chief security officers getting involved now in decisions in storage, which we had not seen before, for instance. Purchasing departments are continuing to flex their muscles. And again, more formal processes. So the demand is still good. It is just taking longer to get through a pipeline.
Obviously, we have been complicating the issues with going through this transition cycle. As you got from our remarks, we did not execute the DMX 2 to 3 transition even close to adequate; and that is our fault. I think the CX 3 transition is going great. Just timing. So it's a little bit of all of that.
I do see, as I said, sufficient budgets. I mean, your own survey, Laura, as well as others I have read, storage is still right up there. As a matter-of-fact, storage and security still dominates the list; and other things we do, like VMware and content management, are hot. So I think our portfolio of products is good. Our sales force is phenomenal.
I would use a baseball analogy here. I just find three quarters in a row now we are doing a tremendous number of diving catches in the outfield here. Twice, even though we caught the ball, you haven't got enough time to throw it to get the runner out. So I just think it's more prudent to say, okay, we just have to go with bigger backlogs and a little more conservatism.
To the same point, I have indicated that Bill and I are going to go after more of the synergies around the acquisitions we have done; not in terms of the go-to-market or development, but in terms of everything else we do; and really work the cost the side hard; and just run the business a little more conservatively. And hopefully -- Laura, to your favorite word -- be able to be in a better position to gain our footing and to create some upside.
So it is prudent, and I think over time we will try to give you -- Bill and I have to talk about it -- but we will try to give you better metrics on what is happening with this backlog. But we're talking of backlogs here that are more than double, closing in on triple what we would have run, say, a year or two years ago. So they are growing substantially.
We need it because you never know what customers are going to order. We are chasing thousands of orders at the end of a quarter, and hundreds of those orders come in. If the mix is wrong -- we don't want to be sloppy and just carry huge inventories. One of the reasons EMC did great transitions in the past was we carried bigger inventories, and we had turns of four; and we certainly don't want to go back there. We need to make sure that we run efficiently.
So we just think this is more prudent. As we regain our footing I absolutely believe the market will like the answer...
On EMC's fiscal 2Q conference call, CEO John Tucci accounts for the company's failure to execute on DMX2 and DMX3 orders: