Thread regarding SAS Institute layoffs

Speak of the devil…

Max severity RCE flaw discovered in widely used Apache Parquet

A maximum severity remote code execution (RCE) vulnerability has been discovered impacting all versions of Apache Parquet up to and including 1.15.0.

https://www.bleepingcomputer.com/news/security/max-severity-rce-flaw-discovered-in-widely-used-apache-parquet/

Gotta get patchin’.

“The releases aren’t requested by customers at anywhere near the pace with which they are created. It’s not cost effective.”

Exactly. I still wonder just why SAS, with customers who lagged 1-2 years behind the original release cadence, thought monthly releases would be a huge attraction. I suppose much of it was predicated on attracting new customers who were accustomed to that cadence, but that didn’t work out.

Open Source is free like a puppy. One of the hidden costs is keeping up with the rapid release cadence (something to be admired, eh? a bunch of dedicated folks getting it done quicker/faster/better) than a well paid commercial fleet of developers.

OP has a case of wants their cake, and to eat it too. NSTAFL and all that. Modern software is an exercise in compromise- maybe hard to adapt to when you live inside the country club walled garden.

Hitching the cart to Viya for over a decade was stupid. It was a huge gamble and the cost of losing that gamble is the biggest reason for the decline of SAS as a company.

It is a damn shame.

The releases aren’t requested by customers at anywhere near the pace with which they are created. It’s not cost effective.

The OP describes a specific case of a general problem.



SAS early on established a dominant market position, particularly once the FDA made SAS the tool of choice for dr-g development. The money flowed in steadily, for forty years.

The deal was: SAS won’t give you equity, or pay top dollar. But the benefits are great, and you’re only required to work 35 hours per week. To many people, this is a great deal, and SAS had no trouble hiring.

Many of the 35-hour people actually worked 40. And there were always a few, say 10-20%, who worked 50 hours or more. To produce high quality work, that was sometimes the only way; and one heroic effort could carry a small project.

But there was no way to reward these people. You could promote them to Principal — but they made the same salary as the 35-hour Principals.

Still, the motivated few were self-motivated, so they did the work. And while the business was stable and dominant, this model worked well.

Then Open Source came, and some of that Open Source was R and Python — free competitors to SAS/STAT. Microsoft, already on everyone’s desktop, began competing from the low end, adding simple statistics and good graphics, enough for most business reports. The transition to the Cloud was underway, so our market was moving, and we had to follow.

Suddenly, from a dominant market position, we had acquired three different sources of fierce competition. At the same time, all the tools changed, and all the libraries changed. It was not stupid to integrate with Open Source; it’s the cost-effective approach. But Open Source was only one of our challenges.

And we tried to meet all these simultaneous challenges with a workforce who had been promised a 35-hour workweek.


That is the general problem: our model worked, but only while our business was stable and dominant. When competition arose, and technology began rapid change, we needed greater effort. But there was no way to reward greater effort, so we didn’t get it.


Long ago, SAS was a startup. 95% of startups fail. SAS lasted fifty years. Objectively, this business model was a brilliant success, though all good things come to an end.

‘The “considerable denial” is attached more to them not recognizing that at most 20% of developers historically were willing to consistently put in both the intensity during work hours and overtime necessary…’

Certainly, many managers did not recognize that. They were not competent to differentiate between their more productive people and their less productive.

However: ever since Viya failed, that differentiation has not been the corporate goal.

“If you want your best and brightest to leave, what are you left with?“

A more marketable company, with an improved ratio of sales per employee.

That is the goal. This company isn't trying to grow; it's trying to shrink, in order to get itself sold. I guess that decision was made, once it was realized that Viya would not reverse the revenue decline.

“The fact that [productivity] wasn’t valued more really shows how ineffectively managed SAS has been for the past 15 years or more.”

Absolutely. We all saw it happening, and we couldn’t do a damn thing about it.

The “considerable denial” is attached more to them not recognizing that at most 20% of developers historically we’re willing to consistently put in both the intensity during work hours and overtime necessary advance SAS technology, maintain current releases in the field AND curate any form of third-party or open source integration. This is the bread and butter of what keeps the company relevant, and the fact that it wasn’t valued more really shows how ineffectively managed SAS has been for the past 15 years or more.

If you want your best and brightest to leave, what are you left with? This problem has been accelerating at SAS for going on 20 years and has now reached critical mass. The founders have lived long and are so wealthy that they basically don’t have to care. It’s all that certain their heirs are already set regardless of how much SAS ultimately yields in a sale or IPO.

“Upper management and HR seems to be in considerable denial about [experienced employees leaving].”

They wanted us to leave. By the time of the first buyout, plans for stealth layoffs and attrition were also in progress. As a result, headcount dropped from ~14,500 to ~12,000. This was quite deliberate.

SAS values experience; certainly JHG always has. But at this point in the company’s history, reducing headcount is valued more 😔.

Open source integration and working with third-party libraries has always been a challenge, and especially at SAS, given such a long history of developing its own proprietary systems-level software. Most of the developers doing this integration are also responsible for other complex components within SAS infrastructure. Hence, it’s more than a full-time job for one person, even someone at the Principal level with a couple decades of experience.

Over a span of many years, it’s difficult to motivate competent engineers to consistently work the long hours necessary to do all of this well, without providing above market salaries or some form of equity. That’s why many of the top Devs who were skilled in this integration either took the first or second VRBO or left for significantly better compensation during the Covid stimulus bo-m. Upper management and HR seems to be in considerable denial about this.

Is this that Open UI5 stuff from SAP?