I wonder if the tall guy realizes that his success was based on SAS being the only game in town for so many years. Was his decision making less than stellar from the beginning? It would have been next to impossible to fail in those early years. So what’s left at this point? A customer base that a number of companies would like to have. Unfortunately, that customer base is not worth what others think it is.
3 replies (most recent on top)
Sure luck is involved in all billion dollar companies but to imply that Dr G is a buffoon who stumbled into it is stupid. Reeks of jealousy and little man syndrome.
SAS started as a platform for US agricultural statistics. If SAS had stayed in that niche, it might indeed have been “hard to fail”.
But to grow into a $3B company — for general data analysis, not just agriculture, and global, not just in the US — that was a large and difficult task. There were many ways to fail.
But we did not fail. We innovated, with MVA and other ideas, and we grew.
Those ideas were strong: they built a revenue stream that grew for twenty years. It was not until the early 2000s that growth slowed and began declining.
There’s always a lag effect. We had stopped innovating, but the lag effect took twenty years.
Meanwhile, the free market did what it always does. Others innovated, and got rewarded for advancing into what used to be our market.
This is deeply saddening, for those of us who spent careers at SAS. But those who make the innovations deserve the rewards.
SAS was a very legit innovation organization/company from its earliest days in the 1970s through the 1980s with the introduction of the MVA Architecture (host-specific runtimes for executing the portable Data Step and PROCs C code compiled for a particular host), making it relatively easy to port Base SAS along with various growing product suites ( basically collections of specialized PROCs) to new host platforms. During that era, SAS hired very smart people and cultivated top platform/host level programming and development talent with deep specialties, and unique skill sets (emphasis on “for that era”). This included building our own C compilers for specific host platforms when necessary It would seem, there were a lot of good business and technical decision making in those days, and there was also competition to contend with. Nothing was a gimme and had SAS not developed MVA, its success would’ve been much more limited in the 90s.
SAS succeeded by providing a highly standardized way of importing data, cleaning it up and doing analytics with the opportunity to save it in a highly optimized SAS proprietary format (this was before the days of open source were most products had proprietary file formats). SAS was known for very high-performance, the ability to handle lots of data, and an unmatched analytics capabilities. With MVA this could be delivered to a myriad of host/OS platforms with considerably less engineering effort than it would’ve otherwise required. This ability to deliver SAS to a huge variety of users, and use cases is how the first $billion in annual revenue was achieved.
By the late 1990s it became clear that future growth would require new ideas and products. this saw transition to employing SAS as a tool set to build enterprise computing products and ultimately a multi-tiered enterprise computing architecture/platform. The company had grown considerably and factions/political rivals formed (natural in any fast growing organization that reaches a certain plateau). Many things were done in haste, and in silos were there was a lot of redundancy,. In just a few short years, R&D morphed from a world-class innovation machine where top developers had reasonable autonomy, to a “little r big D” software manufacturing concern presided over by a small number of politically motivated leaders (in their defense, I believe JG-imposed we’re a big part of why the organization evolved like this — after you got $1 billion plus generating cash machine why mess up a good thing with thinking about how it could eventually be disrupted)
Yet, the company very much continued, with a “not invented here, mentality“ that often minimized attempts to investigate what was going on in the outside world. This was in full force by 2005. If not already achieved, the company was sneaking up on. $2 billion annually. Apart from a few Linux diehards, open source was mostly nascent in 2005. SAS was able to execute on a plan to grow to $3 billion annually — all while big West Coast Tech, the cloud and open source was slowly changing the game right underneath our noses. Sure, we back Hadoop — and where is that today?
Sadly, SAS surrendered the platform 20 years ago, it just took another 15 years for that to come to light.