I recently attended a fantastic conference on Business Continuity and Disaster Recovery put on by DRI Canada. The keynote speaker, Dr. Thomas Homer-Dixon was phenomenal. He founded the Cascade Institute, which studies crisis and the factors leading up to them. One key theme emerged from his talk. A crisis is almost never a random or unpredictable event. It’s a combination of both long-term systemic stressors (predictable) and a short-term trigger (unpredictable) that causes a crisis. Stressors x Trigger = Crisis. Think about large forest fires, the long-term stressor is a drought, while the trigger might be lightning from a storm.
I started thinking about the world of analytics. Modern Python and PySpark based analytics brings incredible cost, speed, scalability, functionality and productivity gains. But organizations, particularly governments, large banks, insurance companies and healthcare organizations, actively resist the change, along with the enormous benefits it brings. It’s human nature to avoid change, as much as we all like to claim otherwise. We only change when the change is forced upon us. Think about our collective lack of action on climate change. We’re only now starting to act because we’re in the middle of the crisis, not 50 years ago when we knew there was a problem. Let’s all just hope it’s not too-little-too-late.
The stresses on organizations to modernize their decades old analytics technology is constant and increasing, and yet most still refuse to do anything about it. Everyday, they loose competitiveness to others, and they lose the best and brightest minds to more forward thinking companies. These are not mere abstract concepts, I’ve witnessed them in action countless times recently. Some of my smartest friends left great jobs at large financial institutions because their technical skills were left to rot on the vine. But for most organizations, there has been no trigger and no crisis that would force them into action. And so they wait…
I would love to say that our business relies mostly upon the so called “forward thinkers”, who want to benefit from open-source innovation, but alas I cannot. The sad reality is, over half of all our customers come to us in a complete state of crisis. Most knew they were putting off their analytics modernization, but maybe they just didn’t think it would come back to bite them so hard.
I’ve talked with some organizations that have been thinking about modernizing for the better part of a decade, but just haven’t gotten to it yet. Some trusted that SAS’ closest SI partners would help them modernize – can you say conflict of interest. Some were so delusional to believe that SAS itself would help them modernize.
Let’s face it, SAS is legacy platform, born out of ideas from the earliest days of computing. It’s not going to miraculously become otherwise. It’s an almost inconceivable fact, but there’s no practical modernization path for SAS9 customers to migrate to SAS’ “modern” Viya platform. The only realistic option to modernize SAS9 workloads, is to convert them to the industry standard PySpark.
What about the triggers? Let’s break them down into two categories. First we have the customer level triggers, the ones that will affect an individual organization largely separate from others. The other category is the large scale industry wide triggers, the ones that will affect many customers simultaneously.
While our customers have told us about a great many triggers, there are few broad themes that emerge. For many, they hit a critical wall with respect to performance and scalability. With every increasing data volumes, SAS9’s inefficient, single-threaded execution model developed in the 1960’s is just not up to the tasks of modern data analytics. The nightmare inducing impacts of blowing out a overnight ETL runtime window, is an all too frequent problem.
For others, the trigger is more resource based, like the loss of key SAS based staff. They lost that one grey-haired guy who still knows how to fix SAS workflows. One customer I spoke to, lost their entire data analytics team after one linchpin resource retired, causing a cascade of departures. The team that took over were very highly skilled, but like all modern data workers they were Python based, knowing absolutely nothing about SAS.
Many times, the trigger is some kind of compliance or regulatory issue. The SAS9 platform is notoriously complex and insecure by modern standards, with breaches and data loss originating from SAS systems being commonplace. For these reasons, most in IT, especially those with a mind for security are no fans of SAS. Failure of IT security audits is a trigger, albeit one that is far less common than it really should be. Funny how much scrutiny is involved in getting new software approved in an enterprise, but nobody questions the complex and insecure decades old software already lurking around. Let’s also not forget that regulators play a large role in the SAS’ core Banking, Insurance and Healthcare sectors. Regulators are showing more and more signs of being fed up with how slow to respond SAS users are with regulatory requests.
Then we have the triggers directly caused by the vendor itself. Many companies have policies ensuring they have commercial support for their legacy tools. SAS knows this, and in and effort to get the best sale price in a company sale, they want pump up their modernization numbers. They desperately want customers to adopt its bizarre proprietary Viya plaform. As such, they have been declaring to customers that SAS9 is end-of-life and will no longer be supported, and they need to sign up for Viya. How exactly they will no longer support the product that virtually all of its customers are using? I’m not quite sure about how this one will play out.
The majority of customer triggers we hear about involve debilitating financial pain. Many of our customers report getting hit with massive increases in annual licencing costs, costs that were already a pain point. It all comes down to how the original deal was structured, and how desperate the salesperson is to hit their quota. While many customers report being hit with relatively moderate 20-30% increases, we’ve heard many talk about 400% increases or more. Those are some ransomware level extortion numbers that would make even Broadcom’s CEO grin.
Speaking of Broadcom, let’s talk about some of the large scale triggers. Like a earthquake or hurricane, these are triggers to a catastrophic crisis, one that might have SAS’ customers heading for the exits all at once.
We’ve already had some close calls, like a potential sale to Broadcom back in 2021. Rumor has it, what saved SAS customers, was just how messy SAS’ internal bookkeeping and accounting is. Ironic, coming from a company that literally sells accounting software. With VMWare, we now all see the Broadcom playbook clear as day. Broadcom likely walked away from SAS, because it struggled to understand how much more money it could extort from the SAS customer base. SAS customers may not be so lucky in the near future, as despite numerous delays, SAS is still publicly marching towards becoming “IPO Ready” – industry speak for “someone please make us an offer”.
Sale or no sale, at some point the inevitable will happen. It’s a fact, even greedy self-centered billionaires can’t escape death. I for one, take some small comfort in knowing that, while those with unlimited means can escape the law and taxes, they like the rest of us can not escape the grim reaper.
There’s no denying it, data analytics and AI is the most radically changing field on the planet right now. At 81, the majority owner and CEO is well past his best-before date for handling such challenges. But step-away he will not – as long as he’s owner and still breathing, he will still be CEO. Insiders describe the relationship as “He is SAS and SAS is him”. No public succession plan; even mentioning the word succession is taboo. After over a decade of declining real revenues, it’s not like there’s a pool of highly skilled executives to draw upon. But the fact remains, when the inevitable happens, a power vacuum will be the result. Other egoists waiting in the lurch will rush in to fill it. More than likely, the infighting of the vultures will push the company into a full scale existential crisis, no doubt publicly impacting customers.
The stressors and triggers of a legacy analytics crisis are foreseeable and preventable. When the crisis hits, industry leaders like Neilson, Capital One, TransUnion, AT&T, and Sun Life, who’ve done the real work to prevent their own crisis, will be the ones to thrive instead of dive. Learn how to avoid your own legacy analytics crisis – contact hello@wisewithdata.com today.
*All product or company names are trademarks of their respective holders. Use of them does not imply any affiliation with or endorsement by them. WiseWithData, nor any its directors or executives, own any interest in publicly traded companies mentioned in this post. Any ideas or thoughts mentioned in this post should not be construed as investment advice.