Core to Pure Success…

This Barron’s article, which I may have linked to before, pretty much sums everything up about Pure. It is nearly a year old now, but the gist is still the core to Pure.

FlashArray is a fine product, likely to maintain (and gain) marketshare. However, if they just maintain marketshare I do not think the stock will move much. They need to continue to gain marketshare.

However, FlashBlade is the wild card. Analyst projections are in the article. They are WAGs, except for the first year as that is based upon PSTG’s announced sales goals in year 1, using the low end of estimates from PSTG, which is $80 million.

As you can see by the announced expected growth rates just how important and material FlashBlade is to PSTG. PSTG has announced in at least one interview that there is more market potential for FlashBlade than FlashArray.

FlashArray holding its own will maintain value, but FlashBlade is what will move this stock.

Management refused to tell us how sales are going with FlashBlade. That is problematic. Valuation is still quite low for PSTG. I think there is a reason for this, and that is FlashBlade concern.

I can find nothing, anywhere, as to how well FlashBlade is selling. Certainly FlashBlade is created for uses today, but more so for uses tomorrow. It is a new product.

The article states that at the then valuation (quite a bit lower than now) that FlashBlade is a cost free option for PSTG. I think the same still holds.

I also think that FlashBlade is not selling as quickly as PSTG would like at present. If it were there would be more noise about it. That can change. Clearly FlashArray is kicking butt given last quarters growth rate. Clearly stealing marketshare and on fire.

I will continue to look at FlashBlade, but good to better understand what is going on in the business.



trying to catch up on Pure…

what proportion of the total FlashBlade is taking now and at what rate is it growing? and compared to the rest of their products?

why do you say that Pure thinks that is not not selling ‘as quickly’? did they put out some targets they want to hit?



There is such a dearth of information on how FlashBlade sales are doing that my post, that I put up, what 30 minutes to an hour ago made the front page of Google as a result.

Dang, that is fast matriculation through the system. No wonder why facial recognition cameras are capable of probably finding anyone, anywhere, as long as they are in the zone of cameras just like they show happening in movies these days.

Scary living in a world with tighter and higher enclosed boundaries under the tyranny of information everywhere on demand. The potential for abuse is far higher than it ever has been. And yet even without this look what North Korea has been able to do.

I guess this means there is no information on the web as to FlashBlade sales, so will stop looking for now. Perhaps PSTG will announce something.



just a step back…

my idea of flash SSD was that there were a reliability issue coming from read/write cycles (although I came across a study from Google that that was not the issue but uncorrectable BER was) and lifetime in general.

Has these issues been solved or corrected with some sort of coding or software?

on another topic, do you think Pure will be acquired eventually?


I will see if I can get a sense on flashblade success (and internal focus on flashblade wins/revenue) from their sales teams. In the meantime, here is some general overview info I found that others may find useful.

June 2017 blog highlighting key announcements at that time:…

This blog also contains a general overview on Object Storage and where Pure Flashblade fits in:…

Positive article highlighting how Flashblade works for a Hybrid Cloud environment (utilizing AWS or Azure as target for backup storage)…

“We could envisage snapshots of VMs from a data centre server going off to Amazon and then being instantiated to run in AWS, providing cloud-based backup, restore, migration and disaster recovery features. SNAPDIFF is an openly available ASPI for third parties to use and Pure has a range of partners using it to send their data to its arrays, such as; Actifio, Catalogic, Cohesity, Commvault, Rubrik and Veeam. We can imagine that could help data move onto its FlashBlade kit.”

More of a negative article on lack of traction for Object storage vendors/solutions.
Basically seems to be saying that now that the data is winding up in the cloud, the Object Storage vendors wind up competing against other multi-cloud data mgmt tools.…

“wiftStack CEO Don Jaworski says object storage has: “grown beyond on-premises object storage to genuine multi-cloud data management.” It then has to compete with other multi-cloud data management. products whether they ate object storage ones or not.”

Another article’s take on CEO comments (or lack thereof) in conf call from Nov:

Oh, so FlashBlade growth is slowing. We don’t know if this is significant. There was an $80m revenue expectation for the full year. Answering a question, Giancarlo said: “We are very pleased still with the FlashBlade growth and it continues to be much faster than FlashArray in its first full year relative to the $80m, and this will probably be the last time I talk about that, we’re in the ballpark.”

Sounds like it could be less than $80m.

President David Hatfield talked some more about FlashBlade markets: "There’s two distinct markets for FlashBlade as we’re getting more stick time with it. First is the next-gen AI, machine learning and real-time analytic segment. That is a great fit and completely differentiated from anything else in the market in that world.

“I think the second area is the replacement of legacy IT and this is going after the NAS in the object stores that are out there. And this is actually taking the fight to NetApp where we really haven’t focused in their file-based world and it is a first generation product.”

Couple takeaways: they view NetApp as a key competitor for Flashblade, and they mention it is “first generation”. While this could be viewed by some as a cop-out, the reality I have seen is that first generation (whether Nutanix, Simplivity, HPE Converged Systems, FlexPod, Vblock, HPE hyper-converged, etc etc ) products tend not to be gangbusters with sales. Most clients don’t want to be the first to try out an unproven solution in their production environment. Nutanix, a few years ago, was this weird/new thing that had a lot of buzz and a lot of tire-kicking, but it took time before large enterprise clients would admit that the product line was mature/stable and legit for Enterprise use.

So I think it is more important to see if Pure commits to 2nd and 3rd gens of the Flashblade product and how sales ramps up as more and more customer references and success stories pile up, which tends to fuel new sales.

Here is a year-old article referencing a Flashblade win over Netapp:…

If you want to start at least hearing about wins, look to LinkedIn or Twitter for Pure sales teammates and see what they are touting around Flashblade. Pure has overall account managers and then specific Flashblade Acct Execs it appears:…

Found this article on the head of Pure’s public sector biz:…

Article on Flashblade winning AI award back in Sept:…

Feb 2018 article touting some flashblade use cases:…

Hadoop is the big data play. Cloudera, HDP (Hortonworks) and MapR (private company) are the big 3 players for Hadoop. Here is article on flashblade for on-prem Hadoop via MapR:…

Flashblade win case study via quick linkedin search:…

mention of first Japanese company to use Flashblade:…

from last Summer, audio of lead Flashblade architect talking about the product at 2.0 phase.…

This means it technically isn’t first gen anymore, but still a new line for sure. Sounds like 3.0 should be out soon.



Fantastic stuff Dreamer!

Yes, the quote from the last conference call you cited…there is really only one interpretation to that and that is the FlashBlade is running below the bottom of their sales projections that they gave for the first year.

As you wisely state, first generation product, unless you have some real pain that other products cannot reasonably alleviate, why be a guinea pig.

Too early to tell if the product will come into its own or not.

I think, OF COURSE it is growing faster than FlashArray grew. When FlashArray came out Pure had no customer base. Now they have a large customer base they can sell into, plus sales partners/references such as NVDA, CSCO, ANET.

So talk of sales ramping faster than FlashArray is facetious.

This is what I like to do. Dig qualitatively into what is moving a company and its valuation. This is what is doing it. A failure of this product, I believe, will be catastrophic to the future of this company. So there is real risk here.

Before everyone gets up and panics or defends otherwise, it is still early days for this product. Later this year if sales have not picked up (analysts, I believe, expect FlashBlade to do $200 million in sales in year 2) then we know there are real problems.

First, you cannot hide $200 million in sales. There will be news one way or the other.
Second, how can PSTG justify not filling investors in on a material product like this if it does not come close to meeting goals, or not tell us if it exceeds goals, when you are talking 10-20% or revenues for the year, and more so in future years.

We shall see. But we do know the core issue to follow with PSTG.

I will review the other links you provided as well, as well as your insight some more.




I also think that FlashBlade is not selling as quickly as PSTG would like at present. If it were there would be more noise about it. That can change. Clearly FlashArray is kicking butt given last quarters growth rate. Clearly stealing marketshare and on fire.


I agree that their not specifically stating sales always makes one suspicious but here is what they said at their earnings call:

Steve Milunovich

Thank you. You had a year ago talked about FlashBlade being about $80 million of revenue, did you achieve that goal?

Charlie Giancarlo

As we mentioned Steve, last time around we weren’t going to be tracking things on a quarter-by-quarter basis, but what I can say is that we had a great quarter with FlashBlade last quarter. The total strength across the portfolio was 48%, almost 50% and we just very pleased with the growth in FlashBlade and we are expecting very good year in FlashBlade next year as well. So it’s going to be one of the fastest growing new products, I think ever in the market and we are pleased with that as the performance.

That is certainly not sounding like they are disappointed.

Also, if that article you linked is to be believed, the Blade is more often enabled by the previous Array purchasers so as goes ARRAY…one might surmise follow-on to Blade.

Further in the earnings call, more corroboration here:

Erik Suppiger

Okay. Then can you talk a little bit about your go-to-market with FlashBlade you had talked about having to refigure that out on the last quarter’s call, it sounds like things have gone well, do you feel like you have got your go-to-market pretty well nailed down for FlashBlade at this point?

David Hatfield

Yes. We feel great about FlashBlade. I think as you introduced the new product, you have a broad number of use cases and you are trying to get into the ones that are going to be highly repeatable and then drive those through your channels. And so we have seen a great uptick in rapid restore, a great uptick in AI and machine learning. And so we feel really good about the motion there. We have got a specialist team that are focused on DevOps and data scientists in these next-generation workloads and that’s working well. And look the 50% top line growth that we have achieved is a combination of the portfolio selling motion that we have pulled together and the focus on the net new opportunity for FlashBlade.

One thing that caught my eye in their earnings call was this statement:

Product gross margins were 65.7%, down 0.6 points sequentially, primarily from mixed dynamics as FlashBlade increased its overall percentage of revenue.

Obviously not a big drop in margins but I was hoping Blade would improve margin rather than go the other way…but again, a sign that Blade is becoming more material to over all revenue.

Just wish they had guided more than 35% YoY revenue for 2019.



I can still remember Chambers giving great Cisco projections for a year ahead right before the 2008 crash.

The decisions about buying a lot of this network stuff, ANET and PSTG in particular do not require a long lead time. It’s not like building a battery Gigafactory or a steel mill.

So I a going to lean to the non cynical side and say they just don’t know anything specific more than a few months ahead. They are not taking predictions beyond what the period where they have good insight. Being responsible
If our conjectures about AI and big data being revolutionary, and that new types of networks and storage will be needed are correct, a quarter or so is not very important in the larger picture.

thanks Tinker and Duma for the discussion


Obviously not a big drop in margins but I was hoping Blade would improve margin rather than go the other way…but again, a sign that Blade is becoming more material to over all revenue.

I don’t know about the industry in general but I have observed before that newer products can start off at lower margins even though you may sell at a higher ASP then the older models due to higher bill of material costs. Over time the costs get driven down as volume raps up and the products get a bit more mature.



Excellent post on PSTG and FlasbBlade sales.

I like naming things after myself :wink: so I will again call up the Tinker Ratio, that I recently defined on the NPI board.

It once had a much longer acronym that I forgot, so I unselfishly just titled it the Tinker Ratio.

It is Actual real world business performance (forward looking, but one must also look back)/ expectations.

This is not a full proof ratio as often a stock that is underperforming does so because we are improperly evaluating the numerator, and as we discussed many times, stocks that go up tend to continue going up.

This said, at present, PSTG has one of the highest Tinker Ratio scores of the stocks I follow. It is of course somewhat riskier as so much depends on the success of one product. But it is hard to imagine that PSTG has screwed up the product (that if I recall correctly, has never been done before because others who tried to do it were unable to pull the technology off…at least I recall reading that, and as many may know I read a lot and bits like this come back to me) and that they will mess up the marketing.

I guess I have confidence in management here. I have confidence in the product here. And I have confidence in the market they are shooting for here.

I think therefore the Tinker Ratio calculation is legitimate.

Further bolstering the case is that it is not like PSTG has underperformed at all. Over. The lat 6 months PSTG has been one of the best performing stocks, and has made the radar at IBD with ever improving fundamentals. It has not made the elite yet, but the tier just below the elite.

When do you want to buy, when the Tinker Ratio narrows, or before hand, having to wait a bit, and hold the company while it narrows the Tinker Ratio?

A bit riskier, and it does call for some self confidence (as so much in the world does…confidence, not cockiness), but if FlashBlade is rolling out well, then I believe the Tinker Ratio number I am getting is legitimate.

Fear, uncertainty and doubt of course. Always our friend and fiend.