Thread regarding SAS Institute layoffs

@4wcg+1oW1I0bm. FTW

A few of us were working to reduce CAS’ RAM requirements by making it stream directly from high performance storage formats and data providers. This is sufficient for a variety of use cases because the entire data set often does not to be RAM resident, especially when the column projection is minimal.

I wouldn't put Elon Musk on any type of pedestal. Firing 80% of the staff at Twitter is nothing to get excited about. Remember, he made an arrogant error in his $44B bid for Twitter, and the stock market and his net worth dropped soon after it was accepted.

He tried many strategies to get out of that contract, and when it was clear that he would be held to it, he pretended to embrace it rather than being held to the deal by the courts.

Cutting 80% of the workforce was an effort to save face and reduce his losses. He overpaid and knew he was stuck. Now he has to do whatever he can to try to make it a "winner", or else he has egg on his face. Narcissists can't handle egg on their face.

As an alternative, maybe Elon can make rock or mineral collecting his passion? Or does the passion fuel the rock collection?

Is VDMML that three-letter-name guy's vanity product?

My understanding is that V9 is bloated, and needs modernizing, because that’s the natural tendency of software when not carefully managed. I’ve monitored the LinkedIn job listings too, and they make clear that Viya is incomplete, lacking features that the market wants.

The two platforms undermine each other. One should be invested in, and improved, while the other should be gracefully terminated. And if SAS must eliminate other positions to support that investment, that’s a necessary business decision.

I could make a technical argument for investing in either platform. But going backwards to V9 looks like a marketing disaster. How would SAS present that? “Oops, sorry, we made a huge investment in Viya, but we weren’t competent to build anything useful in eight years, so now we’re throwing it away?”

SAS seems positioned between two stepping stones. I can’t any path to success going backwards. So I think SAS must either improve Viya, and go forwards, or go nowhere.

...I actually think if SAS had taken data storage more seriously, had architected Viya to be not so "in-memory" centric, and hadn't been so wedded to legacy SAS programming baggage, they could have built something more akin to Snowflake or Databricks (founded in 2012 and 2013 respectively) and could now have a promising future. Sadly, that opportunity has long gone.

My objections to Viya were perhaps not so much the underlying architecture but the quality of the products that were built on top.

VA never caught up to more feature-rich competitors like Tableau and Power BI and was extremely resource hungry requiring massive hardware for relatively small data volumes, and even then the performance was often woeful.

VDMML was far too d-mbed down and lacked the flexibility most data scientists needs, especially in data prep / feature engineering.

They were also full of defects...at least the versions that I used were (2020 and earlier).

But to my earlier point about the lack of jobs asking for Viya...I'm sorry, but there is no hope for an architecture, platform (or whatever you want to call Viya), that has been in the market for 8 years and is simply not being used.

It doesn't matter how clever the underlying architecture might be, if the product on top are poor quality and the target market isn't interested.

One has to define what "Viya" is. For many of us it's not some ethereal mishmash of technology and products, but instead core technologies (CAS, Compute Services, GO Microservices, etc.), that really cool stuff can be built upon. It is especially useful for scalable/elastic compute situations, although some significant work likely remains to be done to make this fully functional in the various public clouds. I have personally seen it beat open source in large enterprise/industrial strength, analytics situation's that have to scale massively.

What are your specific objections to Viya? Is it products like VDMML or VA, compatibility issues with Version 9 stack and more traditional SAS workflows? Please be more specific. Viya Covers a lot of technology ground. I think it's SAS' best and only hope if the company is going to retain any long term semblance of its current technology and historical computing language presence.

The one flaw I see in the business plan, is the assumption that investing in Viya is a good idea. I think that'll be throwing good money after bad. It doesn't matter how technically good it might be (it isn't, I've used it), It's been in the market since 2016 and SAS has pushed it as hard as they can...yet hardly anybody uses it.

SAS can talk it up as much as they like, point to one or two success stories, but the real story can be found when you look for how many jobs are asking for it. I just searched for Viya jobs on LinkedIn....a grand total of 3 jobs across all of the United States. I've run that search a few times over the years, and every time since pre-covid, the numbers have dropped. It looks like Viya usage peaked about 2019 and has been declining ever since.

I can only guess people tried it out based on what the SAS sales guys told them, struggled with it for a while, then gave up on it...probably reverting back to SAS9 or switching to open source. Another guess is that most Viya product is shelf-ware now and likely to be cancelled before long.

Sadly, Viya won't ever save SAS....I don't think anything can at this point.

@3zgj+1oW1I0bm

Agreed on it being a sound business plan, but for thousands of departing SAS employees worldwide, the necessary RIF would certainly seem mercenary.

Acquiring SAS is fraught with too many headaches IMO. I can’t think of a single major enterprise or cloud player who could benefit significantly from integrating SAS technology into their own stack. Not sure about the value of an IPO, unless it’s accompanied with the implementation of a similar business plan, with a retrenchment that would cut the number of current employees in half.

@2jwg+1oW1I0bm

Investing in Viya, and the products that are working well, by ki-ling off what’s not, is not mercenary. It’s a sensible business plan.

However, if I were 80 years old, I might not want to embark on such radical change. I’d certainly be reluctant to lay off thousands of employees.

Anyone purchasing SAS, even if they share this vision, will have alternative investments for their money.

Look at what venture capitalists are doing right now. They could buy SAS, but they’re not. Instead, they’re investing in risky AI startups, and risk-free bonds paying 5+%.

You’ve proposed a darn good business plan. But, yes, it’s unlikely.

@1rub+1oW1I0bm

The academic and grassroots support that you speak of has been trailing off for the past three decades. Modern SAS has succeeded by being an enterprise software company with deep analytics capabilities. It still has many of those strengths and could succeed building niche products on top of its own infrastructure with strong open source interfaces like Viya provides.

The key to SAS’ economic survival/thriving is to eliminate useless/unnecessary positions, including a significant quantity from middle management. Yes, this is mercenary, but it’s the only thing that will save the company from continuing to slowly bleed out. This would require cutting several thousand positions worldwide, and then hiring a limited number of world-class engineering and business talent, who are ready to seriously kick @$$ on an aggressive an extremely focused plan that focuses only on existing verticals (e.g., fraud and risk) and other advanced capabilities like visual analytics/statistics built atop Viya.

Once that is solid, new vertical products can be considered. Viya also needs stronger data management infrastructure. This is very hard to build correctly, so a targeted acquisition might be the best alternative.

I just don’t think JG has the will to do it, nor do I think the financial engineers who might purchase SAS will have the vision to do it.

SAS is not going to recover. They can implement strategies from whatever phony guru they want, but it won't help. SAS has had it's day in the sun, and the sun is setting.

Limited competition, and a grass-roots support movement amongst academics is what built SAS. Now there are other options, and more proponents embracing open source solutions. That moment in time is gone, and you can't claw it back.

It's too late for all that. The company is keeping all of the highly paid managers, directors and VPs in place and they have been shuffled around and put in charge of all the brilliant plans to modernize all the things. What got SAS here can't get us there now or ever.

It will just be more of the same. The people way down at the bottom who do the work, the ones who have not been able to advance because the people the people above them never left, will continue to suffer and carry the multiple layers of people put in place above them. It's the most asinine thing I've ever seen.

…and the echo chamber continues. Twitter != SAS. It’s going to take a shakeup of some magnitude for SAS to turn around. JG will ultimately determine if this occurs under his continued reign/ownership or some successive executive/owner initiates it.

Most of the success SAS continues to enjoy, which includes its legacy revenue stream, is brought to you by SAS employees who embody Musk’s principles in spirit and practice —. Including JG for several early decades. Without their competence and sacrifices, the salaries of less productive/effective SAS employees could not be funded.

SAS != Twitter. Think about the number of times Musk’s principles rescued Tesla and SpaceX from failure. That’s the point being made in this thread.

Frankly, the world will probably be a better place if Twitter/X dies. One thing is for certain, the innovation engine that has driven success at SAS will never be saved by suits.

Elon Musk has confirmed that he has fired over 80 per cent of the entire workforce of Twitter so far. The billionaire took over Twitter in October last year and carried out multiple rounds of layoffs after the $44 billion-takeover. Musk, in an interview with BBC, on Twitter Spaces announced that only 1,500 of the around 8000 Twitter employees are still employed at the company.
Source: https://www.businesstoday.in/technology/news/story/elon-musk-confirms-he-has-fired-over-80-of-twitter-employees-so-far-377045-2023-04-12

Seriously, does anyone really want to see a guy like Elon leading/taking over SAS.
God helps all remaining SAS employees.

Buying Twitter was a mistake. The $44B Musk paid for it is gone gone.

But Musk’s other companies demonstrate that his principles generate profits.

Instead of buying Twitter, he should have used half that $44B to buy SAS. Both he and SAS owners would have been better off.

"Sorry, how is Twitter doing?"

LOL... Twitter is no more. It is now called X...
I don't want to see SAS X-out (close for good...lol).
The last thing SAS needs is a leader like Elon.

"Sorry, how is Twitter doing?"

Swimmingly! Last I heard, it had only dropped in value by 80% since the brainiac guy purchased it. Our CEO should definitely take advice from that guy and respect "only the laws of physics" (but not his workers).

The dynamics surrounding social media, politics and the Twitter business model and AD-based revenue model are very different from SAS.

SAS needs to re-energize its current customer base and reimagine already successful industry solutions like Fraud & Risk atop Viya (with whatever new services/capabilities are necessary) as a beginning point for new product revenue. It is all but certain that SAS can only succeed based on the strength of its platform. AFAICT, it can be safely observed that attempts to build products apart from this platform legacy have to date not been successful — certainly not at any cost effective long-term scale.

Sorry, how is Twitter doing?