Marketing Memos

The technology business has more than its fair share of lose nuts.  Some are wealthy as well as insane, while the dot-communist are poor and insane. 

Perhaps the key to success in this business is being bonkers.

Sadly, some of the insanity is settling in Microsoft, as witnessed in a blog written by Sergey Solyanik.  Sergey recently escaped from Google and landed back at Microsoft.  Perhaps he had been on a spying mission all along.  Regardless, he seems to have mentally drifted in the direction of a room designed with very soft walls.

“I can’t write code for the sake of the technology alone — I need to know that the code is useful for others, and the only way to measure the usefulness is by the amount of money that the people are willing to part with to have access to my work.”

(emphasis mine) 

“Only” is a rather exclusive word.  It means alone; solely; exclusively;  It means no exceptions.  It means Sergey left a marble or two in his cubicle at Google.

The true measure of usefulness is how many people actively use something.  Google uses a lot of Linux, and basically pays nothing for it.  Has this company with $166B market capitalization not found use in Linux?  Have users who perform 62% of their web search using Google not found it useful despite not giving Google a dime of their own money?  Certainly advertisers — who don’t buy Google technology directly and yet they toss about $5B into Google coffers every quarter — find Google technology useful.

Sergey confuses revenues with value.  The two are not unrelated, but one can create technology that is fantastically useful and not make a dime on it.  Or they can create technology that is useful and widely used, and for which revenue is generated indirectly.  His defense of the traditional marketing model for software make sense if you are at Microsoft — that is the game they play best … having failed at most of the newer technology business models.

He goes on (and on) with more off target insights:

I was using Google software … and slick as it is, there’s just too much of it that is regularly broken. It seems like every week 10% of all the features are broken in one or the other browser. And it’s a different 10% every week - the old bugs are getting fixed, the new ones introduced. This across Blogger, Gmail, Google Docs, Maps, and more.

Aside from sounding like a normal software development side effect, Sergey ignores the history of how Open Source and Freeware evolves.  Linux was buggier than a night in a swamp when it was first launched, and for a few years thereafter.  Like a small child it was learning to walk and falling down a lot in the process.  But walk it did.  Then it ran.  Then it ran away with the market.

Google has the brain- and horsepower to do unguided R&D, putting tools onto the net in their formative years to field test what is and is not useful.  Companies like Microsoft that thoroughly study concepts before writing code are slow.  Companies that try, fail, try again are nimble and visionary.  Google will make mistakes but in the process will create technologies that never existed before, are insanely useful, and will indirectly earn them even more money.

Sergey missed this lesson.

Bookmark to:

You gotta love loath Larry.

Ellison’s Oracle has been on a buying spree, snapping up enterprise infrastructure and application companies at a rate that makes even aged Silicon Valley watchers blink. Nothing, not even the Federal Trade Commission, could sate Oracle’s acquisition gluttony. Only Microsoft was outside of Ellison’s financial grasp (or perhaps simply dislikes peddling highly defective products).

The array of acquired companies is broad in scope. BEA, Hyperion, Agile, Siebel, PeopleSoft, Innobase … and the list goes on. Two common themes appear in these acquisitions: the acquired products are either essential to the management of IT, or a set of acquired products comprises all the top-tier competitors. Larry has single-handedly consolidated much of the industry and captured most of the server stack above the operating system.

In other words, Larry has the average enterprise CIO by the short-and-curlys.

Throughout Larry’s market-vacuuming process, many competitors and many more customers expressed fears Oracle was showing monopolistic tendencies, with the inevitable outcome being that prices would rise given an artificially shrinking market.

So, it is entirely unsurprising that Oracle is now jacking-up prices 15-20% … without adding any additional value. Polite people would call this robbery. What impolite people call it is not printable.

Marketing lesson #1: When the switching cost for your customers is very high, then you can steal from them … for a while.

WebLogic, a product acquired by Oracle in their recent BEA plunder, has some unique features that make application servicing more effective. Once these features are used, it is expensive (if not impossible) to implement a different application server. This is why Oracle bumped-up WebLogic pricing 47%. Databases are even more fundamental in the server stack, and Oracle’s DBMS prices were lofted 20%.

Such vendor lock-in drives a lot of revenue in the IT technology market, and is the one gripe most often uttered by CIOs and CTOs. Surveys conducted here at Silicon Strategies Marketing indicate that a prime driver behind IT adoption of Open Source is the liberation from lock-in. However, the higher the switching costs the less likely it will ever occur. For example, much of the banking industry’s current technology is the ugly remnants of 1960’s IBM mainframes — billions of lines of COBOL code are just too expensive to migrate.

Marketing lesson #2: Things do change, sometimes slowly.

One of the reasons enterprises love Linux is that they have faced the vendor lock-in problem for decades, and now have tools for avoiding it in the future. As businesses die and are born, the newcomers learn from the suffering of those who came before. Google runs almost entirely on Open Source as does Yahoo and other new millenia firms. Indeed, their business models could not survive if they had been the groom at an Oracle shotgun wedding.

Larry’s larceny will work — his customers simply don’t have a short-term choice. But it does add pressure (especially in these troubled economic times) to investigate alternatives. Existing companies will avoid committing new projects to Oracle, and new companies will run from Oracle faster than a politician runs from campaign promises.

This will accelerate adoption of more Open Source projects, and in the process improve them. These projects will rapidly come to parity with the commercial products, or even surpass them in functionality (anyone looked at Firefox lately). Larry will make a(nother) lode of lucre in the short-term, but …

Marketing lesson #3: Arrogance (and overly aggressive pricing) will aid your competition over the long-run.

Buy Oracle stock, but be prepared to dump it a couple of years from now.

Bookmark to:

Well, that didn’t take long.

Assuming that rumors based on odd leaks are at all accurate, Microsoft may have learned that their Vistas are limited, and that certain competitive pressures cannot be ignored.

Microsoft has heard the market concerning XP’s expiration. It would be hard to ignore this considering that the market has been screaming at the top of its collective lungs. Though covering their mistakes in much the same way that cats do, Microsoft has agreed to let XP live on until 2010. Microsoft claims that only XP Home will be available on “low-cost” PCs. The definition of “low-cost” is vague and gives Microsoft both wiggle room for defining which computers are allowed to ship with XP preinstalled. It also allows for some leakage of this preferred OS into the gray markets and to force Vista onto beefy machines and protect what is left of their brand image in the PC OS market (an image that Apple has successfully trashed).

The new 2010 XP expiration date is more than coincidental. Windows 7, the next Microsoft OS, is expected to be released that same year (of course, given the delays in Vista’s arrival, we should assume 2010 is optimistic). Assuming that Microsoft is learning harsh lessons from their blunder (Vista), this timing allows them to extend XP availability and deter defections to Mac and Linux until a better Windows is available. After all, the switching cost to Mac or Linux is non-trivial.

One of the defects of Vista is its lack of modularity. The monolithic, out-sized kernel and the inbred stack keeps Microsoft from reacting quickly to market changes and providing real and valuable new features. Linux has a tiny kernel and since all functionality outside of the kernel is modular, you can add just as much extra capability to Linux as you need. That is why Linux is rapidly finding its way into embedded devices. Macs are based on UNIX and Apple took the same modular approach. We see Apple releasing updates to their OS much more frequently that Microsoft (major earthquakes come slightly more frequently than Microsoft operating system releases, which means there must be some mystic connections between all forms of disasters).

To understand the marketing downside of the non-modular Vista approach, consider two markets: minimal laptops and servers. Vista’s only claim to fame is the new GUI (which a lot of people hate). The GUI requires a lot of overhead and most of it seems to be baked into the kernel. Thus Vista is not good for minimal laptops and forces laptop buyers to purchase a much more powerful and pricey system than they would otherwise. As for server administrators, they don’t need, want, or crave a sleek and consumer focused GUI — they would gladly rip the Windows Presentation Foundation out of the box … if they could.

Contrast the recommended systems requirements between Vista Premium and Linux. 1GHz vs. 0.4GHz processor, 40GB disc vs. 7GB, 1GB vs. 0.25GB memory.

Microsoft evidently is feeling the pressure. Rumors about Windows 7 indicate they are moving toward a more modular architecture, nicknamed MinWin. Some of the rumors indicate the kernel might consume as little as 40MB of memory, which is hefty compared to some competing operating systems, but surprisingly lithe compared to Vista and well within the specification of low-end machines.

More importantly, it allows Windows 7 to swap non-kernel components to meet the specific needs of different users (imaging swapping the entire GUI without a reboot). Recall that Linux is rapidly becoming the OS of choice for embedded systems. Little wonder considering you can make a Linux footprint incredibly tiny and bolt-in only the services and applications necessary. A Vista kernel can’t compete and this forces Microsoft to offer a completely different operating system for the embedded market.

The Windows 7 kernel switch allows one Microsoft OS to be embedded, run without a GUI on servers, on low-powered micro laptops, desktops and high-end game machines … and all on commodity hardware.

But that’s not the god news for Microsoft.

Linux runs on everything. On ARM-based cell phones, x86 boxes, x64 machines, PowerPC servers, and IBM mainframes. Vista runs on WinTel chips. This dichotomy is a tribute to the Linux micro kernel approach.

Could Windows 7 run on a mainframe? On an IBM mini? On a SPARC machine? On an ARM?

Bookmark to:

Is desperation the right word to apply to Microsoft?

I avoid using terms of human psychology to describe an entire corporation, but desperation may be accurate. If a person or organization ignores basic principles, and instead chases odd and tangent opportunities, they do so in order to fool themselves, avoiding painful realities and hard work.

Which explains why Microsoft wants to bribe you to use their search engine.

Having failed to produce a competitive solution against Google, Microsoft wants to pay people to use their search engine. In a scheme slightly more complex than the U.S. tax code, Microsoft will reward customers when they buy a product located through LiveSearch. Retailers benefit because they only have to cough-up money to Microsoft when something is actually sold to a consumer (pay per acquisition) and not for the mere lead (pay per click). The system is an affiliate program for sellers.

To make all this work, retailers need to contract with Microsoft. To their credit, Microsoft lined up an impressive array of top-shelf retailers like eBay, Barnes & Noble.com, Overstock.com, Sears, and others.

And it will succeed as well as the iWon search engine (what, you aren’t using iWon?).

Microsoft ignores why people use search engines to begin with, which is to intelligently discover things. When they search for products, these products fall into two broad categories: commodity and specialty. Microsoft’s scheme doesn’t work for specialty products because the customer is often searching for special features from (typically) small vendors.

This leaves commodity products, which creates a problem with Microsoft’s rebate plan. An iPod is exactly the same regardless of it you buy it from Circuit City, Best Buy, and Jersey Frank’s “fell of the back of a truck” Emporium. When the product is identical between retailers, there are only a few significant differentiators, namely price and trust in the retailer. Take it from me, don’t trust Frank.

Which is why a lot of people skip product searches and run straight to Amazon.com, a company that has a solid customer service history and regularly has the best price available.

There have been several studies of search habits for consumers. People use search to:

1. Learn more about the originally desired product.
2. Discover competing products.
3. Find the best deal for the selected product.

Quality of results drives all three items. The quality of search results will most rapidly educate a consumer about the product. A few extra keywords will lead a consumer to competing products and a wealth of comparison sites and customer reviews. And adding the phrase “price compare” after the product name will lead a customer to the best price, typically through many price comparison sites like Price Grabber.

Microsoft is ignoring the fundamental need of the customer. Applying what is essentially a rewards program to search does not improve search, where Google is the master. Instead of adding value where users need and want said value, Microsoft is desperately chasing nothing.

LiveSearch, the next iWon.

Bookmark to:

Carly Fiorina’s time at Hewlett Packard and Louis Gerstner time at IBM proved that decentralized management of product lines in a large company is the only sane management strategy.

Lou took an ailing and heavily centralized IBM, and saved it from extinction by decentralizing its product categories. Carly took a healthy, decentralized HP and tried to kill it through heavy centralization.

Now AOL, the perennial poster child for developmentally disabled technology, has seen the light and is decentralizing (while some of their executives face jail time) .

AOL — which lost the ISP wars by failing to add value in broadband, and whose Time Warner appendage lost the portal wars by failing to provide appetizing content — is now decentralizing their content components to the point of debranding — jettisoning the AOL name. Given the tarnished AOL brand image, debranding their targeted web properties is essential lest the fractured pieces of the empire carry the lingering stench of failure.

AOL is creating new — and re-branding old — web properties that are largely devoid of any reference to the centralized AOL moniker. By making each web property (product) a separately branded entity, each site can create its own identity and appeal to highly targeted markets. This amplifies the appeal of each, making the properties stronger.

(This is a lesson Yahoo may take to heart - weakness in one business may create a weak brand image for all other components. Hence Yahoo’s reluctance to be acquired by Microsoft lest Yahoo products suffer even more brand debasement.)

Herein is where branding in large organizations is problematic. If one or more products/divisions are strong, there is the potential for that brand strength to carry into the other divisions. For example, HP’s strength in printers has led many people to buy their often unreliable laptops.

But the reverse is true. I used to manage large HP data centers, and had problems with their proprietary MPE operating system — for a while it crashed too frequently for mission-critical work. During that period I rushed to visit a relative in intensive care, and I was not happy to see her hooked-up to a number of HP medical monitoring devices. The weakened MPE brand created in me a negative prejudice for their medical products.

In web properties — where the subject matters presented on a web site may be audience specific (say teenage boys vs. geriatric women) — decentralized branding is essential. Contrary to common belief, segregating the brands does not eliminate potential synergies.

Take SourceForce.com (formerly VA Software) who has a number of technology industry web sites, on which they sell a ton of advertising. They reap significant synergies from different web properties that have at best sporadic cross traffic. The branding needs for their IT manager’s web site is very different than their Linux junkie web site.

Take two of AOLs newly branded properties — Asylum, a site for young men, and WalletPop, a site devoted to personal finance (a topic with which young men are largely unfamiliar). Pulling both of those under the torn AOL umbrella would do neither property any favors. But allow each to tailor the entire brand to their members is favorable … and thus smart.

The marketing issue here is multifaceted. First, you must decide if there is any real cross-over between the product areas. If so, a unified brand might do you some good. If there is no cross-over, then odds are the unified branding will not help in the slightest and may actually hurt.

Second, does the parent brand offer any specific advantage? HP has a strong overall technology brand, and it generally helps all of their product categories. AOL does not, and thus the AOL brand may hurt the individual web products.

Finally, top management has to have the strength to let go. Decentralizing and allowing subordinates make both strategic and tactical decisions about the decentralized products and brands takes guts. But the bigger and more diverse your company is, the more important decentralising will become.

Bookmark to: