This blog was published in 2016 and may contain outdated information. 

IDC published its 2016 Q1 WW Enterprise Storage market share report (IDC: Worldwide Enterprise Storage Systems Market Update, Q1, DOC #US41520216, JUN 16, 2016), and Pure Storage was pleased to find two pieces of good news within: not only were we the fastest-growing storage vendor named in the report, but we cracked the “top 10” storage vendors (for revenue) globally – not just for all-flash storage, but for ALL external enterprise storage.

AFA and Storage Markets Become Synonymous

As AFAs become increasingly mainstream, the “AFA market” and the “storage market” are becoming synonymous, and in our view it is most interesting now to keep tabs on how vendors are doing in the entire external storage market.  Pure, for our part, in just four years of selling achieved a 2.2% market share and clocked-in at 96.7% YoY growth with a $119M quarter.  To put that in perspective, EMC and NetApp both saw declines of >10%, and HPE was the only of the large vendors who grew at all, albeit <5%. In a world where some of our legacy storage competitors have been touting their AFA gains, the “inconvenient truth” is that those AFA gains aren’t offsetting their dramatic loss in overall storage market revenues, and in fact much of their AFA progress is a result of share-shifting deals from disk to flash (more on this below).

What’s really satisfying here is that our still-stellar growth is now coming off of a substantial base, and we’re just getting started. As interesting as this data is on a percentage basis, this growth is even more interesting when you look at it on an absolute dollar basis, where Pure Storage was, far and away, the largest absolute dollar gainer in the external enterprise storage market:

Dietz Blog Image 6.20.16

In catalyzing the shift to all-flash storage, Pure has emerged as the disruptor to which the incumbents on IDC’s list are most likely to draw comparisons. Not surprisingly, the talk from legacy vendors is about closing the innovation gap, which goes something like this:

“By loading up our Retrofit Array [most any disk-centric vendor and array can be inserted here] with big SSDs, we have caught up to Pure. Our all-flash Retrofit Array offers the same kind of performance as Pure, but it also protects existing customer investments in Legacy Vendor storage infrastructure. Sure Pure has better data reduction and they have Evergreen, but we can discount around those. Pure’s one real trick was being first with flash and data reduction, but that no longer matters now that flash is cheaper than disk. In addition, Pure is not profitable, is burning cash, and is facing an EMC lawsuit. Most of all, the all-flash portion of our business is now growing faster than Pure, so we’re going to beat them.”

Pure Storage cracking the global top ten seems an opportune time to refute these arguments, and illuminate what it is going to take to win in this market going forward.

Who’s really growing?

Part of the reason the legacy vendors are eager to tout all-flash array (AFA) share growth is because, according to IDC, their overall businesses are shrinking. (HPE is gaining, but at an anemic 5% per year they are growing 20X more slowly than Pure.) This gets to one of the tactics that legacy vendors have used to create the illusion of success in the AFA space: Aggressive bundling, maintenance forgiveness, and other programs that essentially transition disk customers to flash during refreshes.  Pure, of course, enjoys none of these advantages, and all of our growth comes from selling a high-value product to customers who make the choice to purchase it.  So if a legacy vendor is succeeding in shipping some AFAs to their installed base, but that installed base itself is massively shrinking, can you call that growing?

Who has sustainable business practices?

Legacy vendors do indeed seem to be discounting aggressively to get around their lack of effective data reduction. In proofs of concept (POCs), Pure’s FlashArray will typically deliver anywhere from 2-5X better data reduction (and that’s not counting the additional savings from deduping data protecting snapshots). As a result, Pure often requires far less storage media for the same workload. Moreover, Pure tends to make more efficient use of Intel multicore processors, meaning we also need less CPU as well. And of course our FlashArray chassis packages all of this in a dramatically more efficient form factor, using less data center space and power.  As a result, for configurations north of 100TB, we estimate that our street price is often lower than some competitor’s hardware cost of goods sold (COGS).

Discounting below COGS is painful for a vendor, which is why achieving cost parity with Pure​ typically also depends on maintenance forgiveness (reducing the support fees the customer pays on their existing storage). Maintenance forgiveness has the added benefit of shifting revenue from a business investors value less (legacy support) to one they value more (AFA product revenue). The challenge is that cutting maintenance fees on the installed base is a gift that cannot keep on giving: When Pure occasionally loses a deal to such financial engineering (customers often thank us for getting them offers too good to refuse), we stay close to the customer because on their next storage refresh (say, 6-12 months later) those dollars are no longer available to tip the scales.

The legacy vendors are also addicted to the unfriendly business practice of making the customer rebuy the same storage every few years, and raising out-year maintenance fees in order to force such repurchases. It is as if your mortgage broker showed up five years after you moved into your home, and threatened to raise your monthly payments twofold unless you agreed to buy a new house down the street. It’s not just the money—switching houses disrupts your life, something will get broken and something lost. With EvergreenTM Storage, Pure customers get a subscription to new software and hardware innovations that allow them to run the same storage for a decade or more. Evergreen Storage is refreshed non-disruptively over time with hardware and software upgrades, and do so for the same flat and fair maintenance fees. As a result, we estimate Pure Storage can save customers an average of $2,000/TB per year versus the legacy solutions, and that’s just over the initial six-year period.   On that front, we have some exciting expansions to the Evergreen Storage program that you can read more about here.

Who’s Risky?

Pure’s business is on sound financial footing. Pure has >$600M in the bank as of Q1 end, had zero debt, was cash flow positive in Q4 of last year and expects to be sustained cash flow positive in the 2nd half of 2017. We have all the money we need to achieve profitability (with a several $100M cushion I might add), and the only reason we are not already there has been choosing to invest for growth. As an aside, it is ironic that EMC is the vendor that we see most consistently challenging Pure’s finances, given Dell/EMC will soon take on some $50B in debt (that’s substantial leverage by any measure).

Who’s Innovating?  

Our position remains that loading up mainframe-era storage technology with SSDs will not enable the leap to the modern data center. Pure remains highly differentiated within the all-flash array category:

  • Simplicity – Delivering a SaaS-like level of ease to storage for the first time, with supremely easy automation interfaces for DevOps;
  • Resiliency – Across 1,000s of arrays in production, Pure is very close to 99.9999% uptime (each array averaging less than 32 sec. of downtime per year), and we achieve that with zero maintenance windows (another first for Pure we think);
  • Fast, global data reduction – It’s not just about economics—data reduction makes snaps, clones and replication dramatically more efficient;
  • Management via the cloud – Our Pure1 cloud automation, proactive support and predictive analytics deliver an infrastructure as a service like experience in customer data centers; and
  • Strong security – Pure Storage encrypts all at rest data with 256-bit AES encryption, and we do so without relying on any SSD native encryption (which some of our security savvy customers have found lacking);
  • Future proofed and Evergreen – Modern protocols like NVMe are going to supplant the disk legacy SATA & SAS, because they enable dramatically (think 100X) higher levels of parallelism between the storage controller (CPU) and storage media (flash). Pure Storage is designed to take advantage of this new technology, a leap we expect to be hard for the disk-centric designs.
  • Use of consumer grade flash – Innovations in flash memory first target consumer devices. It is also the most cost effective on the market, not being 40% overprovisioned like many of the enterprise SSDs. Finally, we have found consumer flash to more reliable (less than .1% annual failure rates in our experience), likely because its firmware is tested for use in tens of millions of consumer devices rather than tens of thousands of data center devices.

Evidence for this differentiation can be found in

  • Pure’s industry leading customer satisfaction – Our competitors scores much closer to the B2B average of a 16 Net Promoter Score than our 79 (as audited by Satmetrix);
  • Our best repeat purchase behavior – Our top tier of customers have spent another $12 in the next 18 months for every $1 they spent initially;
  • The industry’s fastest revenue growth; and
  • That Pure is most often the choice of next-gen data centers – 7 of IDC’s top 20 SaaS vendors use Pure Storage, as do many of the consumer cloud. Reference customers include LinkedIn, Intuit, Workday and ServiceNow. Roughly 25% of Pure’s business to date has come from Software as a Service (SaaS), IaaS and consumer cloud.

Who’s Litigating?

Regarding concerns over EMC’s patent suit, in my opinion their motivation is much more about trying to slow down Pure than belief in any wrongdoing. In our view, we do not infringe any valid EMC patents. In our litigation, we were successful on 4 of 5 patents outright, and for the 5th patent, should we not be successful in post-trial motions or appeal, we’ve already released a superior implementation that avoids the patent (while sometimes called a workaround, we would have done this regardless as it improves our product). We look forward to putting this issue behind the company in the months ahead, and move our competition with EMC back in the market place where it belongs.

Looking forward

As we have remarked on this blog, storage will change more over the next 5 years than it has over the last twenty. We find it entertaining that each of our largest competitors are essentially leading their all-flash strategies with disk retrofit designs, relying on technology that is often more than 20 years old, designed before flash, cloud, virtualization, DevOps and Linux. At Pure, we’re going to keep doing what we’ve been doing—working hard to innovate in ways that better serve our customers and partners, we look forward to graduating from the “top 10” to the “top 5” global storage vendors before long, and we hope that you give truly modern all-flash storage a chance to transform your datacenter, and your business.

Forward Looking Statements

This blog post contains forward-looking statements regarding our products, business and operations, including our expectations regarding technology differentiation, customer adoption and business model advantages, our ability to maintain growth and take market share, and our financial outlook and statements regarding our products, business, operations and results. Forward-looking statements are subject to known and unknown risks and uncertainties and are based on potentially inaccurate assumptions that could cause actual results to differ materially from those expected or implied by the forward-looking statements. Actual results may differ materially from the results predicted, and reported results should not be considered as an indication of future performance. The potential risks and uncertainties that could cause actual results to differ from the results predicted include, among others, those risks and uncertainties included under the captions “Risk Factors” and elsewhere in our filings and reports with the U.S. Securities and Exchange Commission, including, but not limited to, our Quarterly Report on Form 10-Q for the quarter ended April 30, 2016, which is available on our investor relations website at and on the SEC website at All information provided in this blog post is as of the date hereof, and we undertake no duty to update this information unless required by law.