How’s the “Everything is Software” Trend Going?

Software, so they say, is taking over the world. I actually agree with this, and with a related premise that “hardware” and “network” companies have to not only think like software companies, but actually become one. These points raise the obvious question of how well everyone is doing, and fortunately there’s some Street research that offers a clue. Wall Street is always interested in how companies will be doing in the future, for obvious reasons, and digesting the Street view gives us a chance to rate at least the various types of companies with respect to their “softwareness”. Of course, I’ll add in my own thoughts where appropriate.

We can divide tech companies into three loose categories—software companies, computer/hardware companies, and network companies. If we looked at these groups simply in terms of how successful they were in leveraging the software space, we’d expect to see they fall into the category order I just listed. The first thing I find interesting in Street research is that they don’t, in one important case.

Yes, software companies tend to be more likely to be rated as successfully exploiting software, but the computer/hardware companies are rated below networking companies in that regard. That same ranking can be seen in how the Street says it expects companies to perform in software, relative to their stated plans. Network vendors, then, are seen as more likely to exceed software expectations than computer/hardware vendors.

The Street is better at recognizing symptoms than at offering a proper diagnosis. I think the primary problem with computer/hardware vendors is that hardware is supposed to run software, meaning that both the Street and enterprises expect that a computer vendor is neutral with respect to what’s being run. If they offer “software” at all, it may be simply a matter of convenience rather than something that they’re promoting to facilitate differentiation or adding value.

Viewed in this light, the decision by Dell to let VMware go might make a lot of sense. If “software” is truly a generic layer of value on top of hardware plumbing, then linking VMware to Dell would likely risk contaminating VMware’s value story in association with non-Dell hardware. Interestingly, Dell has fewer Street views that their software business will simply match their plans; most think it will either beat or fall short. Competitor HPE is seen as having a much greater chance of “on-target” software performance, mostly because the Street doesn’t see any changes that generate a big upside surprise for HPE.

Networking, obviously, is a different situation. For most of the history of network devices, software and hardware were bundled. If you got a Cisco router, you got some flavor of IOS, and if you got Juniper equipment you got Junos. Today, network vendors are “disaggregating”, meaning they are breaking up the hardware/software bundle. That lets them charge for software, of course, and the shift in the paradigm opens a new software opportunity that’s reflected in Street assessments of their software potential.

Network vendors also typically offer management software, and increasingly security security software too. Since security software is at least among the top software opportunities, if not the top, that gives network vendors a potentially greater level of software potential and upside versus plans.

The downside, for network vendors, is that while the “disaggregation” theoretically opens up additional revenues, the opportunity isn’t open-ended because it’s unlikely that an enterprise would buy a vendor router and then not run their associated network operating system. In my own surveys, I don’t encounter any enterprises who report doing that. Same for management software; it’s almost always tied to the hardware choice. Security software, and even security devices, are more likely to cross vendor lines in procurement, and generally the Street likes the software fortunes of security-focused network vendors better than it does the software opportunities for traditional network vendors, even if they offer security products.

One thing this suggests is that the notion of “disaggregation” in hardware and software terms isn’t an automatic guarantee of lofty software numbers for the network vendors. Generally, network vendors are seen as having a slightly larger upside versus plans in the software space, but on the average about a third of Street analysis suggests a downside. That contrasts to the software space, where less than a fifth of analysis shows that, and the computer/hardware space where almost half of analysis shows a downside risk.

Another interesting insight from the Street view of security is that everyone is confused about it, from the Street to the vendors to the enterprise buyers. The Street recognizes somewhere between five and ten classes of security products. Enterprises report having somewhere between three and six security classes in place, and vendors are all over the place in how they position their stuff. The fact that security products are more prone to vendor crossover, where a buyer gets a security product from a non-incumbent vendor, illustrates the complexity of the space too.

It’s always interesting, and challenging, to relate Street data to my own surveys, modeling, and assessments. This is easiest to consider in the specific case of SASE, which is a “new product category” and thus gets a lot of ink. The Street is split about 50:50 in whether they see SASE and SD-WAN as being related in any way, and where they see a connection they see SASE as being the sum of SD-WAN and security. That view favors vendor presentations of SASE, which tend to try to protect current security incumbencies and products. My view is that SASE has a critical foundation in a proper implementation of SD-WAN, something that virtually no SD-WAN vendor has actually accomplished.

That this leads to even more confusion is obvious. There is absolutely no correlation between the Street projections for who might be a winner in the SASE space, my own data on who enterprises think are winners, and my views on who actually has the best product set. The Street seems to be valuing incumbency in either the security space, the SD-WAN space, or both over any consideration of the actual capabilities of the product set.

So where are we with software opportunity for non-software companies? My view is that both network vendors and computer/hardware vendors have under-realized their potential in software, largely because their commitment to it has been superficial. Vendors rarely have a software strategy; they have more of a software dream. Dream fulfillment is always sketchy, and that’s particularly true when enterprises tell me they’re crying out for rational strategic positioning from their vendors. If vendors actually had a strong software plan, backed up by a strong positioning, they would do considerably better. As it is, lack of both ingredients is encouraging buyers to stay the course, favoring incumbents.

This is all surprising to me, given that we’re actually facing the most potential for technology and vendor shifts in at least 20 years. The bad news is that vendors have been blowing kisses at the software opportunity rather than actually trying to maximize it. That’s left most of them far short of where software could take them. The good news is that there’s time to fix this, particularly for network vendors, and the price for doing so could be very significant.