WRFest 2Mar08(Technology): Small to Large - IT Industry Structure
Odd as it might seem we're still, at the end of this week, just catching up with last week's news
summaries. Here we want to focus on the Technology Industry with several interesting stories. In the process of several of these recent Tech focus summaries we've been wrapping the excepts with some charts that show how the industry is put together. Earlier we introduced a simple stack picture which showed all the elements from platform to middleware to application to interface that are necessary for any particular solution to be put in place. Think of that as the basic characterisitcs of the industry's ecology. Another dimenation that structures an ecology is the number of large and small players and what niches they go over. Which we try and capture with the accompanying chart.
Imagine that the stack describes the vertical axis (you have to picture four segmenting lines horizontally across the dumbbell please). So another dimension of the ecology is the size of the customer from very small to very large. What the dumbbell tries to capture is how well populated, or covered or served, are each of the major niches. So for example IBM and it's strategic partners/competitors, e.g. SAP, ORCL, et.al., have their roots in the large end of the scale. And from platform to application they purport to cover most of the requirements and sell to most of the customers. In fact they ARE in fact installed, one way or another, at most of the customers. Last time I looked 70% of the world's data processing was still occurring at the large end and IBM is very much the dominant player on the platform and middlewar layers. In fact it defines the market despite all the hype you've been reading for 20 years about PC's taking over. When you hear someone talking about Software as a Service (SaaS), the applications cloud, etc. oddly enough they're returning to the roots of almost forty years ago with time-sharing !
On the other end is the small user - businessess and consumers alike (they're some caveats here though - Dell for example really had it's home base selling PC's to sophisticated users in large businessess who understood what they were getting and could provide their own support). But by and large the small end is well served by the Wintel guild though not as completely covered as the large end. That's for two really critical reasons. First the needs of many small businesses and users are served by spreadsheets and other simple application/middlware software packages, e.g. QuikBooks (as good an accounting and small business package as their is). The interesting thing is the un-populated middle where mid-sized businesses have needs that are often nearly as complex for sophisticated applications as large businesses but haven't the resources to afford large installations. In particular they can't afford the staff or application integration resources required to install, adapt and maintain these applications which means they are woefully under-served as measured by the gap between needs and solutions. In other words there is, and has been for decades, a bigger opportunity to help mid-sized businesses solve their technology problems than in any other part of the ecology. It's rather like it must have been for the early settlers first bringing more complex capabilities to North America in the 1600s and 1700s. With many of the same sort of challenges. And all the vendors know it because they keep going after those markets. The problem is that the applications are very complex, big, cumbersome and very...very hard to use. NOT until somebody figures out how to make easy-to-operate business applications that can be sold thru distribution channels will that change. But there are many huge cultural barriers in the way.
The following story excerpts should be read with this ecology in mind. The first one is about the growing use of freeware and/or inexpensive solutions, e.g. Wiki technology, to bring some of the level of sophistication for team collaboration to small and medium businesses to market. Then are two pairs of articles. The first pair contrasts Dell's continuing struggles to re-invent itself and how Apple can increasingly look to Macs, which are growing marketshare more rapidly than any other platform, to sustain itself. It turns out that all the iPod/iPhone/retail store helps with introducing new generations of users to a more friendly, easy-to-use platform, otherwise known as marketing. Meanwhile Dell's forays into Retail were doomed to fail because it's entire functional capabilities were built around the assemble-to-order business model which doesn't fit well into Retail channels.
The second pair looks at the bottom slice of the dumbbell and compares & contrasts the demand for data centers. Everybody thinks of Google as a search, that is software, company. But what makes it work is the acres of data centers and thousands of PC-class servers tied together in giant sphaghetti furball that does all the constant searching and indexing. This might be one among several Google Achilles heels. It certainly means that they will be a capital intensive business. The greatly amusing thing for those of us with some experience is that very large servers (BTW - these are the guys actually running your life whether it's airline reservation systems, you bank or credit card company, your retailer's store orders of what have you) have been wrestling with these same scalability and management problems for decades. And very successfully we might add. IBM gets a huge amount of business from folks who take what're alledged to be a bunch of large SUN Unix servers and consolidating them to one easier-to-operate large mainframe, or a piece of it. And that's even more true of comparing server farms with thousands of PCs vs one mainframe. The staffing, power, reliability, etc. etc. are just not in the same class.
So as you evaluate the trends, directions and investment opportunities in these businesses you need to bear in mind what's really going on. Remember our discussion of Warren Buffett's maxim of "don't invest in anything you don't understand" ? Well it turns out that Warren spent decades of hard...hard work getting smart until he's built up a huge backlog. But he won't touch technology. As he says they'll still be drinking Coke, chewing gum and eating candy in 20 years. Guess what - they'll still be processing larger and larger amounts of data in 40. Who the successful players will be is another question but understanding the ecology will help you figure that out.
Technology Readings
Boosting teamwork with wikis At its simplest, a wiki is software that lets users work together to create and edit a collection of linked web pages. The online encyclopedia Wikipedia is the best-known example - its 85,000 contributors have written, edited, and policed the content of more than nine million entries. Like Wikipedia, all wikis benefit from the network effect: The more people who use it, the theory goes, the better the quality of the information. When FSB first wrote about business wikis nearly three years ago ("Workers of the World - Collaborate!" April 2005), they were largely untested, and early adopters were finding it hard to encourage employees to use them. Now the nearly universal familiarity with Wikipedia and a proliferation of services targeted at small business have helped make wikis commonplace. SocialText, one of the better-known providers, started hiring salespeople dedicated solely to the small-business market for the first time in 2007. Many firms, including Rosen's, are keeping secure wikis for purely internal affairs. But others are finding that opening them up to customers saves a surprising amount of time, energy, and money - and offers other benefits as well. Angel.com, a McLean, Va., maker of software for calling centers, which has 55 employees, created a wiki on SocialText last year that lets clients leave comments, share their likes and dislikes, and even submit suggestions for troubleshooting the software. Those features have slashed the company's tech-support workload and given Angel.com a 10% boost in productivity, estimates company CTO Sam Aparicio, 33.
Dell Recovery Plan Falters as Wal-Mart, Staples Shoppers Let PCs Languish Michael Dell gave himself until August to prove he can rescue his personal-computer company by selling machines through retailers. Shoppers at Wal-Mart Stores Inc. aren't buying it. Hewlett-Packard Co. and Sony Corp. are crowding out Dell Inc. computers at chains including Wal-Mart and Staples Inc., said Cowen & Co. analyst Louis Miscioscia in New York, who has had a ``neutral'' recommendation on the shares and an ``outperform'' assessment of Hewlett-Packard since August 2006. One year after Dell's return as chief executive officer, the turnaround he promised remains elusive. Dell, down 14 percent since January 2007, costs 13 times estimated profit, almost as much as Hewlett-Packard, which became the biggest PC maker in 2006. Dell will fall further behind with sales growth of 6.7 percent this year, compared with Hewlett-Packard's 10 percent, according to analysts' average estimates. Dell, 43, has remodeled his Round Rock, Texas-based company's direct-only sales strategy, attempting to fight Hewlett-Packard in retail outlets worldwide. Starting with Wal- Mart in June, he put Dell machines in 10,000 stores in the world's 10 largest markets in six months. ``I think 18 months is a good timeline to think about,'' Dell told Business Week in February 2007 when asked how long it would take to revive the company. It's been a struggle since then. Hewlett-Packard, the top seller of PCs for six straight quarters, is in 110,000 stores. Dell doesn't have enough models on display to satisfy shoppers looking to weigh their options, Miscioscia said. He said Dell had only three models at Best Buy Co. stores he visited and Wal-Mart displayed Dell's computers ``under glass in a cage, locked down tight.''
Mac has Apple's back The Mac might just be about to become even more important for Apple Inc.'s future growth plans. BMO Capital Markets analyst Keith Bachman on Monday cut his price target on Apple's stock to $140 a share from $160, and said in a research note that "Apple's three growth drivers [iPods, iPhones and Macs] has now turned to one," that being the company's line of Macintosh PCs. Bachman said that Apple's share gains in the PC market are continuing, "with strength in desktops in particular," and raised his Mac sales estimates to 9.4 million units from 8.2 million for Apple's current fiscal year, which ends in September. Bachman also raised his forecast for Mac sales for Apple's current quarter to 2.06 million units from 1.87 million. The analyst said the Mac's strength will take on more importance because the company is not likely to see the kind of growth from the iPhone that it has projected this year. Also, the iPod is starting to show some signs market saturation, leading to slower growth rates for the iconic digital media player.
A RAD Goal: Replicating Google These data centers are the engines that run Google's search software; Yahoo, Microsoft and Amazon have their own. A data center transforms thousands of cheap computers into a single massive Web server. The big search companies had to invent data centers at the dawn of the search era because the problem was far too big to run on a single machine, no matter its size. Today, the list of computer problems for which data centers might be useful is still growing. Yet, the facilities remain very difficult to design and operate; you basically have to be a Google or an Amazon to even try. Enter Dr. Fox and his team. Their goal is to make building and running data centers as easy an engineering feat as putting up a new building and turning on the lights. That's a time-honored job in computing: taking something recondite and esoteric, then making it simple and widely accessible. Over the decades, those forces worked to make possible the extraordinary computers now on every desktop. No one is suggesting there will ever be "personal data centers," but their use could extend outside the small circle of big Web companies.
IBM rolls out faster, cleaner mainframe IBM Corp. rolls out a new mainframe computer Tuesday boasting a 50% performance boost and dramatically lower energy costs than its predecessor. The new System z10, with a starting price at about $1 million, comes as IBM focuses on lowering the price tag for running its storied line of data-crunching workhorses. The Armonk, N.Y.-based company said it designed the new machine to help companies and government agencies that rely on mainframes - usually for critical data processing such as bank transactions or census statistics crunching - save money on energy bills and better handle a flood of Internet information. The size of IBM's investment - the company spent five years and $1.5 billion developing the new mainframe - also underscores its commitment to the long-term viability of the mainframe and efforts continue adapting the decades-old product line to the Internet age. For years some IT experts predicted the demise of the mainframe, bulky and expensive machines that face competition from smaller, less-expensive servers. But IBM says mainframe revenue is growing, rising in 5 out of the last 7 quarters, thanks in part to interest from emerging markets like Brazil, China, India and Russia.
A Highflier Loses Altitude as Google’s Clicks Go Flat Investors have focused with new intensity on Google’s so-called paid clicks, which grew at 30 percent in the fourth quarter, because the search and advertising giant earns the vast majority of its revenue from text ads, for which it is paid only when users click on them. Many analysts saw the comScore report as the clearest sign that Google, which does not give forecasts about its future performance, is not impervious to the slowdown that is buffeting the United States economy.