Apple is pushing people to populate their phones with installed applications while Google, IBM and Microsoft are urging folk to remove apps from desktops.

This is not nearly bizarre as it sounds.

The success of Apple Apps for iPhones is slightly more phenomenal than the second coming. The universe seems consumed by the desire to have useful and useless apps installed onto their handsets. Sure, most of the iPhone app rush comes from the rush of playing with a new toy, proving once again the only difference between men and boys is the price of their data plan.

Yet this month shows that the desktop is slowing turning into little more than a SaaS suckling tool, whereby apps are delivered online. Google may have led the pack with early availability of desktop apps-on-tap, but now IBM and Microsoft have tap danced onto the stage.

(The mental visual of Steve Balmer and Sam Palmisano doing a vaudeville soft-shoe act is highly amusing)

IBM’s offering seems to offer what the market would seriously not consider. Big Blue and Ubuntu (say that ten times real fast) have teamed to push applications from a Linux cloud to a Linux desktop via a Lotus leaf. Ignoring the narrow niche of combined technologies, the requisite once-and-future Linux desktop is a momentum killer. Swapping XP for Windows7 will be hideously painful, but slightly less so than a migration to a completely new desktop OS and application package.

Which is unimportant. $3 per user per month is the real news story.

Marketing clouds parted when the rental cost of the solution set was mentioned. Good, bad or indifferent, SaaS apps have caught the attention of application vendors for a number of reasons, the most motivating of which is that steady, predictable revenues streams are far preferable to the old model. Sure, support revenues were the underlying motivation for many software plays, but the agony of product release and upgrade support programs coupled with associated spikes in revenues and expenses made maddening money flows.

Which is one reason Microsoft is chasing the same market.

Though unavailable until next years Office 2010 release, Microsoft is readying a web version of Office. Though pricing and options are not yet known, Microsoft has in the past licensed server-side solutions in creative ways (for example, many ISPs supply SharePoint on a per head rental basis). If Microsoft’s web apps are sufficiently adept, many enterprises might opt for the technical and budgetary convenience of serving applications via a browser than installing every bit on every lap-and-desktop in an organization. Microsoft would certainly approve of monthly/quarterly/annual rental fees since a steady flow of greenbacks makes wallpapering Bill Gate’s den a simpler process.

The question is if enterprises will bite.

Part of technology marketing is knowing how IT technoids work, or would work if they had the nerve to assassinate their end users. After costs considerations are sacrificed, IT’s primary motivation is pain avoidance. The two greatest sources of pain for IT staffs are end user stability and systems stability. Ignoring that no end user is mentally stable, the topic turns to their computing environment stability. Web sourced apps have a number of end user stability advantages, including the inability for end users to augment the program with unlicensed software (a.k.a. malware) and the uniformity of having all users executing the same program and revision.

There will be fewer bald and bleary-eyed IT admins if web apps become the norm.

But no enterprise shop can go 100% web application. One Forrester Research analyst noted that “Our own research shows that a good portion of information workers rarely use all of the tools in their Office arsenal.” True, but there are power users and road warriors aplenty, and they likely can/will not switch to web apps. Web office applications are (likely) less feature soaked than their binary buddies, but are unlikely executable from a back seat somewhere between Baoji and Baoshan.

This creates a conundrum. Since enterprises have invested mega money into desk-lap-top maintenance systems, would they double their drudgery in order to rent apps? Since upgrade cycles for lap-desk-flat-top applications is about five years, and that the average upgrade cost for a bulk buyer of Microsoft Office suites is around $150, the breakeven costs (ignoring maintenance infrastructure and ulcer medications) is about $2.50 a month per user.

Which is where IBM got their $3/head price.

Web hosted applications may be the wave of the future, but current moment will retard any inevitability. There is a lack of marketing connection at play – a lack of urgency, a dearth of desirability, and a shortage of savings.

It is insane to declare web apps DOA, but one wonders why outfits like IBM and Microsoft sense getting products to market is necessary. I think their main motivation may be less demonstrated enterprise demand and more Google’s goading.