Before WWI cars were hand-built and driven by chauffeur mechanics; likewise GIS late last century was an industry driven by experts. Before WWII automobile mass-production diplaced the artisanal, and owner-drivers replaced chauffeurs; today's changes in GIS firms also look like a shake-up, as GIS now exits the back office. After WWII automobiles became user-friendly with automatic gearboxes and air-conditioning in America at least; I argue that GIS won't reach the corner office until geomatics go under the hood.
Why did I put my money where my mouth is, by joining a webGIS start-up? Because I think that the web allows to do today, what couldn't be achieved 20-odd years ago at my first attempt: to subsume the geospatial to user needs, and provide what augments rather than replaces current workflows. Back to automobiles, I said earlier that we needn't kow how ABS brakes work in order to drive a car. Likewise webGIS helps separate the user interface, from from the underlying engine.
And there as many ways to achieve that as there are players. Server, desktop and web iterations needn't be repeated here. But who and how this will evolve into, is as good a question as automobile manufacturers asked themselves between the two World Wars (eerily they also faced financial meltdown, clash of fundamentalisms and global terrorism, even if those terms hadn't been coined yet). And while user-friendliness is well documented in GIS, what about mass-production?
released under GNU FDL, plate seen in traffic