Thread regarding SAS Institute layoffs

This isn’t a software company. It is a Corporate entitlement program doing business as a software company. Hire your friends and family while you still can.

@3hce+1umZyqji

In my opinion, "catching up" is not the intent. The company is a reflection of the primary owner. It will wither away as he does.

think that's why so many of us left. we already made up our mind the company couldn't adapt, at least not fast enough.

however, that may no longer be the case (low % but never say never ... perhaps there is hope)

@3xzh+1umZyqji That sounded an awful lot like a question you’ve already made up your mind on. Maybe I’m wrong but engaging on that question in this forum seems pointless.

question now is what if anything the company is doing to try to catch up, over a decade too late, or if there is anything remaining for which catching up would even be worthwhile

"pretty much all of my customers were wanting to move to SaaS"

Maybe SaaS had it been fleshed out to be more viable, would have slowed the fleeing away from SAS? Especially for the V9 customers having zero interest in Viya.

Hind sight is always clearer than fo-e sight...

@2foa+1umZyqji yes you could argue that comparing SAS with Snowflake and Databricks is not a fair comparison, but I'd argue it is. Here's why...

The reality is that they are what SAS is now competing with, that's why it's a fair comparison despite SAS being at a disadvantage. Also, other legacy software vendors have managed to successfully transition to SaaS....Oracle, SAP, Adobe, Microsoft, even IBM. SAS has failed where many others have succeeded.

In terms of demand for SaaS, when I was still at SAS (4 years ago now) doing pre-sales, pretty much all of my customers were wanting to move to SaaS. The trend was to reduce their IT department and focus on their core business. There may be some sectors where that's not the case, but I suspect they're becoming fewer and fewer.

Other on premise vendors such as Msft did "hybrid" steps such as letting customer IT run SharePoint or Exchange servers in the cloud instead of or supplementing the on premise servers, keeping the end user experience the same. What did SAS attempt? "Hosted" but no "as a service"?

@2ipd+1umZyqji We are mostly on the same page.

It isn’t entirely a fair comparison with Snowflake and DataBricks.
They were founded in 2012 and 2013 respectively and as such their journey to SaaS was a no brainer.

Out of the 3+ billion in existing revenue at SAS what percentage of those customers do you think would be all in on moving to full SaaS?

There in lies our real struggle. It is one that has been faced by many other old time well respected companies in the past.

• fade into the sunset and die slowly(while still make a shitload of money).
• fully transition and potentially alienate existing revenue streams while not sure of success with new
• walk the tightrope for some time and hope you get the balance right

We are walking tightrope but not in a way that gets us to a clean modern place if we pull it off. Stuck in limbo forever.

@2tay+1umZyqji
I didn't mean to suggest that cloud computing is simple or that SAS's transition to Kubernetes was not necessary. But yes, you're right to say what I was really talking about is SaaS rather than cloud computing per se.

But I guess my point is that customer's expectation of a cloud-based commercial off-the-shelf analytics platform, is that they are delivered as SaaS. And SAS claims that Viya can be delivered as SaaS (frequent reference to it on www.sas.com). But it's stretching the truth.

Snowflake is SaaS. Databricks is SaaS. They abstract Kubernetes away and you manage the containerization within their respective interfaces. Not so with Viya.

The fact that SAS has failed to do something similar in the 10 years since Viya was first conceived, is yet another example of how the company no longer has the technical capability to remain relevant.

Last several posts reinforce the point that not enough R&D and other SAS technical professionals were tasked with cultivating deep technical expertise in the evolution of the cloud in general and specifically security and data management therein. But this goes right along with the political shenanigans, micromanaging and stifling of innovation at SAS that began in the early 2000s. Meanwhile, over that decade, the SAS9 stack grew progressively more complex, convoluted, and less amenable to anything resembling cloud native or SaaS.

Going back to 2005 or so, with rare exceptions, only a few executives went to conferences or otherwise kept up with emerging cloud, data and security trends. They would bring tid bits back with a mandate to somehow “integrate“ with SAS, so some data access engines and Hadoop-related stuff got created … yet this fell far short of the colossal paradigm shift that cloud computing represents.

Compare this to the mid 80s when MVA was conceived and R&D embarked on porting SAS to many disparate host architecture and operating systems. Most host teams had deep technical expertise ICs in all aspects of the architecture and operating system they were porting to. A few developers were even nationally and internationally known for their work.

The lack of a clear cut and progressive architectural vision along with stunted technical expertise related to the cloud is a significant reason why SAS was unable to significantly innovate for this new computing paradigm. By the time CAS/Viya was conceived, modern cloud infrastructure had a full head of steam and technologies like Docker and Kubernetes spread like wildfire.

SAS failed to integrate these technologies into the early design of CAS/Viya. Things have been remediated quite a bit since then, there still seems to be a mismatch between SAS and cloud native operations. Can anyone further elucidate this?

@2yno+1umZyqji While I hear what you are saying I think it is a bit misguided. There is nothing simple about cloud computing.

Kubernetes isn’t the problem. It is a core piece of creating cloud native software. It was a necessary transition.

The bigger issue is taking the next steps to be SaaS. I suspect that is what you meant and yes at that point the SaaS vendor takes all of the pain in maintaining/architecting.
But even then there tends to be friction points when it comes to customer data.
Our problem is not being able to commit to SaaS because of those friction points.
There are solutions but all complicated and they do put some smaller pieces of the burden back on the customer in terms of securing their data.

it's ironic but unsurprising that SAS simply couldn't adapt to the era of SaaS. the more that cloud native startups and open source made things ever easier to run anything as "a service", the more SAS made deployment ever more difficult and costly. complex software that YOU have to deploy "in the cloud" or hire "professional services" to do for you. a ludicrously self-limited market to go after.

The comments around Kubernetes is worth reflecting on. The fact that Viya customers need to know how to administer Kubernetes goes to show just how broken SAS's cloud strategy is.

One of the main attractions of cloud computing, is to take away all those layers of complexity so customers don't need to be experts in the underlying infrastructure such as the containerization tech.

SAS's roll-your-own approach to Cloud, actually adds additional burden on the customer, rather than takes it away. Who's going to want that?

I agree - this sounds like AI or second or third hand info, cobbled together. Honestly there's really not much need to learn CASL; one can still work with Viya without it for probably 99% of things, maybe more.

@OP+1umZyqji

I dunno OP. AI-generated images and the text also smells AI-generated to me... and there's no link to a SAS press release. Probably fake news.

“SAS Sunsetting Mainframe Solutions…”

Fake news! Russian much?

Sounds like an expensive assortment of many things glued and cobbled together under the covers. A fragile mess and a nightmare to administer?

how in the he-- do you sell and support that? what kind of cloud experts does it take if it's not set up to be easy? my IT is pretty much terrible at everything. how would they be able to admin this? so many questions with probably no good answers

Wow, SAS has made some d-mb decisions over the last few years, but this has got to be one of the d-mbest. I'm staggered that anyone within SAS would think that's a good idea.

It's almost like JG wants the company to shut down. Let's give SAS's most loyal customers the best possible business case for completely moving off SAS because that's exactly what is going to happen.

"the hefty price tag of migration"

Adding to that, come November, the economy could tank. And who wants to be spending big when that happens?