Thread regarding SAS Institute layoffs

I’ve worked at a handful of large companies, as a developer.

In all of them except for one, developers were strongly encouraged to interact directly with customers. That included going on trips alongside the sales folks, working with other companies’ developers on joint projects, participating in standards bodies and open source projects, etc. etc.

The one exception? SAS. Management will do anything at SAS to make themselves seem important and to hide their utter lack of technical skills or basic industry knowledge. Therefore, there is no room for actual technical exchanges that mean anything, and no room for folks with actual technical acumen to get with customers to understand requirements first-hand.

It’s shocking to experience, and no developer worth anything would stay at a company that does not trust technical staff and prioritizes management politics and hobnobbing over results.

Product Management for SAS used to be easier. It was a statistical and data management tool that was being developed by statisticians and data managers (more or less.) SAS outgrew this model. But, SAS found some success in related markets.

SAS used to even have surveys of users for enhancements to guide product development.

IMO, product management is best when it includes by actual users. Researchers will never really understand the users and market like a real user. Seldom is it just about the technical requirements. A user works in a whole eco system that includes their own systems, work flows, management, and purchasing. Even if the product is perfect for a user, the user may not be empowered to purchase it.

Was Viya really that much of a new market for SAS, or was it more of an "upgrade?" IDK.

Those were important questions to solve a decade ago, but now with LLMs, it's easy to rewrite or port old analysis code to new languages. Moving old code to a new platform that might not fully run it isn't even contemplated as a need. We are lucky to run our legacy software for some existing analyses.

LLMs already write, translate, explain, and run multiple languages. The mega vendors with LLM and code and analytics platforms/ecosystems will rapidly add more multi-lingual support. Msft should be in a great position with Co-pilot, Github, Visual-whatever, PowerBI ... maybe keeping that partnership expanding will help. Good luck to all.

“ and those folks would be flummoxed by having to deploy and support Viya in the cloud.”

I’m not sure you understand what “in the cloud” and ultimately SaaS means.

"Too much ignoring, perhaps taking for granted is a better way to word it. The big German's passion for Viya was surpassed only by his disinterest in V9."

When the BG announced that Viya would NOT incorporate any of the SAS9 capabilities (yes he did say that) - I knew that was the beginning of the end. He just thought his VIYA baby was SO great that all the SAS9 uses would flock to it. A total lack of understanding of how SAS9 was being used.

I have a question for those here who know our existing customer base... I understand that the large banks/insurance companies/pharma have adopted the cloud technology and are comfortable with the way Viya is deployed. But I always wondered if that was enough of the revenue from SAS9 to be successful. When I was in sales years ago - it seemed that most of our customer base was at the departmental level - and those folks would be flummoxed by having to deploy and support Viya in the cloud.

Is my perception correct that without that departmental component of SAS sales, we would be hurting? Those are the folks - it seems to me - who are deciding to go to R or some other software rather than try to figure out how to go to Viya - which doens't have a lot of their required capabilities anyway.

When I was still working - I can't tell you how many times JG said "I want ONE SAS - put SAS9 and Viya together"... and everyone would say "yes sir" - and then NOT do it. To me THAT is why SAS is struggling...

Hope so as well. With LLMs being able to do analytics and write code, and big data companies laying off SWEs en masse, it doesn't seem like the big opportunities will continue to be in analytics tools areas. Game is over. However, vertical problems/solutions will likely benefit from improved AI ... hopefully the company can improve in those areas.

V9 compatibility would help SAS sell Viya to existing customers.

The greater opportunity was to acquire new customers. At the turn of the millennium, SAS had the premier product in the world for analyzing data. Since then, the amount of data in the world has grown 1,000-fold — most of it in the Cloud.

In such a huge market, SAS could have grown by taking only a small piece of it. Yet we didn’t. By now, most existing and potential new customers who wanted to migrate to the Cloud have already migrated. The Cloud was a generational growth opportunity, missed.

Now AI is the next one. Like the Cloud, AI is a huge market; perhaps SAS can take a small piece of it. SAS has a new leadership team; it’s fair to give them a chance.

I expect SAS will maintain most of its existing V9 customers, and continue reducing headcount by ~2% every year, until the company is sold. Based on history, this continued slow decline is all I expect.

But I hope they do better. Good luck to all.

"Can v9 DATA step code run under Viya with out tweaks? Same question but for all v9 PROC step code"

Any other answer other than yes,yes creates obstacles to migrating to Viya because:

1. They are waiting to see what happens post sale/ipo.
2. They do not have the manpower or time to sift code for potential changes.
3. They do not see enough value in Viya to justify the migration cost to Viya.
4. The overall V9 experience is preferable over Viya.

Can v9 DATA step code run under Viya with out tweaks? Same question but for all v9 PROC step code.

I know I’ve said this before, but some product(s) did establish V9 compatibility. It made sense for our users. My product, SAS Optimization, has a clear migration path for PROC OPTMODEL, the primary modeling and solution procedure. Same for other optimization procedures. We highlighted this via doc examples and in demos at SGF. We didn’t ask for approval from upper management beyond our own development director.

"Did the BG ignore marching orders from JG that Viya be compatible with V9?"

If true, that is an stunning example of destruction having two significant components:

1. The BG self destructed his career at SAS.
2. The BG destructed the future of SAS.

Again, if so, the only way this could have happened would be for the BG to have 100 percent of JG's ears and was feeding JG a constant stream of deception.

Lastly, and again if the quoted matter above is indeed true, secrets never stay secret forever, which means that JG eventually found out he was being deceived and cut BG loose. By then the damage to the SAS brand was beyond repair. The Titanic had already hit the iceberg at full throttle. A decent comparison because the SAS story and the story of the Titanic seem to have many similarities. It takes years to build greatness but only a few very poor successive decisions to make it sink quickly to the bottom. Sometimes happy voyages have tragic endings.

I don’t know; but when the history of the company is written, the BG will be blamed for SAS’s failed transition to the Cloud.

That was the growth opportunity, and it was missed. By this time, most customers who wanted to move there are already gone.

Did the BG ignore marching orders from JG that Viya be compatible with V9?

SAS did ignore that base.


SAS users, like other users, wanted to move to the Cloud. They didn’t want to rewrite their old SAS jobs, just wanted them in the Cloud.

SAS failed to meet that need. The BG wanted to replace SAS with something better. Maybe he did, but that’s not what the base wanted.

The decline of SAS began well before the BG was in charge. But because of this large, prominent failure, he gets more than his share of the blame. 



It couldn’t happen to a nicer guy 😉.

"own a company that has 3 billion in SUGI Sam revenue and tell me you are going to ignore that base with a straight face."

Too much ignoring, perhaps taking for granted is a better way to word it. The big German's passion for Viya was surpassed only by his disinterest in V9.

Catering to the need of SUGI Sam. Is a gross oversimplification.
You can and should innovate while also protecting existing revenue. 3 billion reasons to…

It’s easy to think otherwise from anonymous site but own a company that has 3 billion in SUGI Sam revenue and tell me you are going to ignore that base with a straight face.

Underlying those organizational choices, settling a debate on "whose job" doesn't really matter because even if both product managers and engineers listen to customers, it only exacerbates the innovator's dilemma as the overall market adopts newer, cheaper, better, faster (disruptive) technologies. Many of us saw that happen at SAS over and over. Both product managers and engineers alike would insist on catering to the needs of the dwindling, aging customer base of "SUGI Sam", loudly proclaim success internally for "listening to customers", trying to get a promotion or recognition, all the while not paying attention to or being cognizant of the dwindling and the aging of this base. Many tried to go beyond, but failed for many other reasons these posts haven't really covered beyond the Peter Principle issues.

No matter the title, a person who only talks to customers but refuses to listen to them is a person who should be given the boot sooner rather than later. ESPECIALLY A HIGH LEVEL person.