Saturday, October 24, 2015

Thinking About HP Enterprise

As we come closer to the official splitting of HP into two companies, I've been listening again to the HP presentations, looking at some of the other big players and reviewing  my SWOT analysis of HP Enterprise.

In the three year turnaround of HP, CEO Meg Whitman characterizes FY 15 as being a year of "accelerating progress," in which the combined company invested in security, networking, and cloud activities and acquisitions, as well as on improving execution.

She noted that research and development as a percent of sales increased in each of the three turnaround years.  The company, she said, introduced new products and services, e.g. all flash memory, 3Par storage, HP OneView, software defined networking, a Gen 9 server, enterprise-wide implementation solution for Office 365, and HP Helion, i.e. the public cloud.

Post-split, HP Enterprise will have pro-forma revenue of about $50 billion, with operating margins of over 9% and operating income of about $5 billion.  Fifty percent of revenue will come from enterprise hardware, 37% from enterprise services, 7% from software, and 6% from financial services.

Notice the absence of any reference to Big Data, a major theme of several conference calls during the three year turnaround period.  IBM still talks about this theme, especially in its research publications, and in connection with Watson.  What happened? Autonomy.  The only way this acquisition made sense was as the analytical engine and product generator for a Big Data effort.  That investment has been vaporized, and whatever remaining products are in the market can't sustain a major presence in this theme. A definite weakness going forward.

We've written for more than three years about the weak, commodity service focus of Enterprise Services.  Despite the cost base restructuring, the fundamental weaknesses of this business compared to others like IBM and Accenture remain.

A $4 billion software business which has flat lined for several years is too small to be important to the needs of CIOs in the transition to the next generation data centers.  Another weakness that we've identified for years is still in the standalone company.

I would surmise that CEO Whitman, in her heart of hearts, sees the same landscape.  How else to justify a statement like more than 90% of the IT spend over the next several years will be what she calls "traditional IT."  This should be comforting to investors, presumably, because 50% of HPE revenue will come from traditional hardware boxes.  Also, traditional hardware sales have historically been accompanied by enterprise services contracts.  Actually, these statements, which may be correct, are reasons to be very much afraid if one were an HPE shareholder.

Traditional IT Spend Means Further Hardware Commoditization

A recent podcast on the Andreessen Horowitz website, entitled "Dell + EMC: Say Why?" gives some interesting views on how CIOs will behave over the next few years.  The first generation model of the data center is what they call the "Wall Street Data Center," made up of server racks from vendors like Cisco, HP, Dell, storage products from EMC, and Oracle software.  The next gen data center is represented by Facebook, Google, and Amazon.

In their deployments, server hardware has been truly commoditized, to the point that the best way for Facebook to achieve performance for its particular business needs is to produce its own customized server designs.  The software layer of the stack is the critical element in this type of data center.

Dell's recent merger with EMC positions it as a systems provider to some types of enterprises in helping them get the the next gen data center configuration.  Dell, which survived commoditization of its once core PC product, is better equipped to survive a commoditization of server hardware than is HP, so the panelists suggest.

What Kind of Cloud For Large Scale Enterprises?

The Andreessen Horowitz panelists agree that the future will largely be made of up of hybrid enterprise clouds, incorporating some legacy structures alongside newer, next gen designs.  Public clouds are a losing business, given the efficient, excess capacity of Amazon Web Services.  As we have written for some time, AWS, despite some recent big wins in the public sector, won't be the option of choice for the largest, multinational corporations in data intensive, higher risk industries like financial services, energy, and transportation/logistics.  

HP recently announced that it is shutting down its Helion public cloud business touted as an exiting new product introduction by CEO Meg Whitman just a few weeks before; HP is sending its customers to AWS.  This leaves HP pursuing the hybrid cloud structure, against stronger competition.

The panelists cite the fact that relatively few companies have the kind of technically savvy, global sales forces to succeed in selling and implementing a hybrid cloud computing offering to the biggest global companies.  They name IBM, Microsoft, and EMC as being the strongest sales forces, and HP are not mentioned.  So, Dell's interest in EMC beyond storage is in acquiring a world class sales force in order to help it become a systems provider and service partner for its customers.  

Most companies, the panelists rightly say, cannot wave a wand, spend billions, and risk their businesses on a straight implementation from a Wall Street data center model to a Facebook-type data center. Instead, they reckon that most CIOs will recognize that there is excess capacity in public clouds and huge price pressures on hardware vendors; the customers will exploit both to give them a multi-year, relatively lower cost glide path to the development of some Facebook-style new data center architectures alongside legacy public clouds.  

The winners in the private enterprise clouds they reckon will be, as we have suggested, Microsoft Azure and IBM.  

HP Enterprises path to generating earnings growth beyond year 1 of easy comparisons will be fraught with difficulty, and raising research and development expenditures significantly or making big acquisitions will both be greeted by consternation on the part of new shareholders. HPE may still suffer from the past sins of the combined HP entity, despite a workmanlike, three year turnaround. 

No comments: