Quick Thoughts on Broadcom/SAS
The latest rumor: Broadcom will buy SAS.
Don’t say I didn’t warn you.
Wall Street seems to like the idea. Broadcom stock (trading under the AVGO ticker) closed up 6% for the day. That’s a vote of confidence.
The reaction of many: who the hell is Broadcom?
It makes a lot of sense for an industry outsider to buy SAS. Why? Because any established analytics vendor will have channel conflicts that diminish the value of the company. SAS GSI partners will abandon ship if IBM acquires the company. Same for Microsoft, Oracle, or SAP. SAS has always positioned itself as platform-agnostic, providing customers with tools that work on all platforms and with all databases.
Assuming that this deal is real — a hell of an assumption, but work with me here — what will Broadcom do with SAS?
First, kiss the rock collection and the goat farm goodbye.
Next, expect new leadership to take a hard squint at that bloated product portfolio. They will ask questions like how much revenue do we get from this product? and how many developers support this product? They will quickly discover that there are a few products with legs and many that — ahem — do not have legs.
The same rule applies to headcount more generally. Computer Associates employed 10,500 people before Broadcom acquired it in 2018. After the sale, Broadcom’s headcount rose by about 4,000. So, about 6,500 jobs disappeared, and CA wasn’t exactly bloated.
Those fancy offices in castles and manors? With reduced headcount, overlapping office space and the WFH trend, the few that remain open will be sold and leased back. More cash. If you’re still working for SAS, you’ll love WeWork.
Interesting question: will new leadership keep Viya and toss legacy SAS, or the opposite? Adoption of Viya is very low among SAS customers, while development and marketing costs are very high. The new boss just might decide to keep the legacy software, milk it for cash, and let it decline. Or they could sell off one or the other.
Here’s what Broadcom won’t do: leverage their strengths in chips to develop advanced AI appliances. SAS R&D has never done very much with hardware. The company was very late to the GPU game, and generally works with industry standard hardware from Intel or HPE.
So acquiring SAS does not provide Broadcom with the talent it needs to make a hardware-based AI play. That’s not to say they won’t try that play, just that acquiring SAS doesn’t help them.
Anyway, those are my first thoughts on this. Will be interesting to see if this deal goes through.