The Broadcom deal with VMware has finally closed, which ends one element of confusion swirling around the companies. It doesn’t end all the elements, though. In fact, I think the finalization of the deal raises a lot of questions, not only for Broadcom and VMware, but also for the tech space overall. The big question? What kind of merger is this? The history of the tech space offers a lot of possibilities.
Cisco is a network company that’s done a lot of M&A over the years, and their motivations have evolved through two primary classifications—financially driven and sales-complaint-driven. Financially driven M&A is based on the difference in the P/E multiples of the buying and bought companies. If you have a P/E multiple of twenty and you buy a company that has a multiple of ten, their revenue creates a big upward push on your share price. This was a really popular strategy in the pre-Chambers days. Then we saw a shift, to the sales-complaint model.
This model starts with a hypothetical visit to Chambers from some favorite friend in sales. He says “Shucks, John, we’d have won the XYZ deal if we could have sold them Widgets”. Chambers ponders, then says, “Well, we’ll just go out and buy us Widgets Inc.” Behind this seemingly-trivial move is a deeper thinking that has now taken hold, which is that if a buyer wants a specific product that’s tied to the fringe of your own product line, you risk losing account control if you let them get it from a competitor.
The next phase of this account-control M&A drive, I think, is the “future strategy” driver. This driver has two dimensions; competition-risk and opportunity-growth. In both, the buying company isn’t addressing today-level problems or opportunities, but things that appear to be on the horizon. Those things might threaten their incumbency down the road, or they might open a new opportunity area that could create major growth. This driver has the greatest potential impact on both the companies involved in the M&A and on the market they represent.
There is an outlier concept on the table, too. Some companies in the last couple decades have aimed at being a tech conglomerate, creating administrative efficiencies by pulling a bunch of businesses with no real product affinity into a collection in order to leverage common resources for the task of running the company. You have common CFO, common payroll and personnel, a single stock, but separate product and sales processes.
One thing that makes analyzing Broadcom’s M&A here complicated is that in large deals (as this was) it’s likely that there are multiple drivers whizzing around in their own complex orbits, adding up to a new kind of planetary system. We have to look at all the factors to come to a useful conclusion.
One factor is what Broadcom currently offers, which is a wide range of stuff that really makes them a de facto conglomerate-builder. They’re likely best-known for their network white-box chips, but they’re a kind of tech conglomerate, having bought (in reverse order) VMware, Symantec, CA Technologies, Brocade, Emulex, PLX Technology, LSI CyOptics, and more, and is in fact the result of an acquisition by Avago Technologies. Many of the products they have today have come from those acquisitions, but there have been criticisms of Broadcom’s handling of all these acquisitions. That’s particularly true with the Symantec and CA deals; customers and financial analysts were equally disappointed, and Wall Street friends tell me that they don’t think that the conglomerate model works to create a good stock multiple. In fact, their biggest concern is that Broadcom’s biggest stock upside might be at risk if all Broadcom does with VMware is swept into the Mulligan stew that’s the current Broadcom structure.
That weights heavily on the factors that could justify the deal for VMware. If customers didn’t like the most recent of Broadcom’s M&A outcomes, then there would be a risk that would also be true with VMware. If there is a near-term customer revolt, then it’s unlikely that there will be any favorable outcomes from the other factors that could justify the deal. No short-term, no long-term.
So we’ll start with the short term. Factor number one is that financial considerations relating to the oldest buy-revenue M&A driver are paramount in the early days. It takes time for a buyer to realize the other benefits that an acquisition could create for it, and in the meantime there will be quarterly reports and meetings with Wall Street. It’s critical to stand tall in these, because to show a problem out of the gate is to potentially poison future benefits. That means that the first thing Broadcom has to do is ensure VMware’s acquisition creates favorable numbers almost immediately. This is the goal of the decision to establish four VMware divisions, already announced.
Broadcom is establishing two divisions, VMware Cloud Foundation (VCF) and Tanzu (TNZ) that represent the bulk of VMware’s operating revenues and their primary incumbency. I think that Broadcom will leverage these in a strategic sense but won’t interfere much with the product directions and sales tactics. This is aimed at preserving and even expanding the revenue stream for “incumbent VMware”. Retention of the VMware customer base, and the quick establishing of a positive attitude within that base, is absolutely critical. These divisions set the stage, and without that initiative nothing else will matter.
The other two divisions represent both what Broadcom thinks is the future opportunity and where Broadcom thinks positive symbiosis between VMware and Broadcom would exist. One is the Software-Defined Edge (SDE) and the other Application Networking and Security (ANS). Given that Broadcom is first and foremost a network chip company, it seems obvious that this is where they’d start in creating ties between their own business and that of VMware.
Inside this divisional structure, I think, is a strategic view that “hosting” is generalizing into the business of running stuff wherever it makes sense. Thus, VMware under Broadcom isn’t pushing an explicit data center strategy but is instead generalizing its platform tools to run in both the data center and the cloud. VCF is where I think that initiative will fall, and we should note that VMware has been moving in that direction for a couple years. Think of this as embodying the “hybrid cloud” model.
Why, then, have SDE? Because, I believe, SDE represents two things that VMware hasn’t pushed in the past. One is embedded control, which is what “edge computing” really means in today’s world. The other is edge middleware, a set of tools designed to provide the framework on which edge applications would be built.
ANS is where the immediate opportunities for symbiotic growth would develop, because that’s where VMware intersects with the core of Broadcom’s chip business. The concept of “application networking” is pretty vague; some articles have suggested it’s nothing more than load balancing. Surely, though, it could embrace the SD-WAN tools VMware has, the NSX virtual network technology, and likely SASE too. VMware’s SD-WAN has already started to integrate with Symantec’s Enterprise Cloud and this suggests that VMware will be integrated with some of Broadcom’s own assets in this area. And many of these things could be solidified into appliances, which would be white boxes that could be built on Broadcom chips.
VMware, then, might be nothing more than another element in Broadcom-the-conglomerate. Of course, anyone who’s followed my views on the Broadcom/VMware deal knows that I think the biggest potential impact would come from emphasizing the relationship between switching chips (from Broadcom) and platform/middleware tools (from VMware) to create a big white-box revolution. This would take a bit of navigating in terms of channel conflicts, but I think Broadcom would gain a lot more than they’d put at risk here. That would be particularly true if they didn’t actually sell switching products and network appliances, but instead sold the underlying chips and software.
Wall Street is generally favorable about the acquisition, which is more a testimony to the near-term financially linked leverage and symbiosis than on its strategic value. Some inside Broadcom, after my blogs about the potential white-box linkage, told me that this is indeed “discussed” internally, but they stopped short of saying it was an approved strategy, or indicating how high up support for the idea could be found. I think a white-box symbiosis is the important point here, because the VMware deal is much larger than Broadcom’s typical acquisitions, and I don’t think any of the traditional ways they could get a good payback on the investment will serve to validate it in the long run.
