– Submitted by: Kas Thomas, Analyst
Economic downturns tend to accelerate change in the IT world: People with budgetary authority find themselves taking a fresh look at what they’re spending money on, how and whether IT investments are paying off, why bad investments are not paying off, and what to do differently going forward.
Given the situation we’re in, now might be as good a time as any for potential buyers of software systems — and licensees whose contract renewals are coming up — to declare war on per-seat pricing.
Seat-based pricing has been declining in popularity for some time (everyone I know hates it), but like a weed growing out of a crack in the sidewalk, it never seems to go away entirely. In theory, per-user pricing is rational because it allows costs to scale in a predictable (and fair) fashion, according to an organization’s size. But in reality, there are many problems with the “headcount equals usage” notion, the most obvious being that if nine people touch the system once a month and one person uses it eight hours a day, seven days a week, you still need 10 seat-licenses even though nine out of ten users are offline at any given time.
Some vendors try to address the idle-user problem by taking a “concurrent users” approach in which you pay according to how many simultaneous sessions you think the system will handle in times of peak demand. In the previous example, you might be able to get away with a two-concurrent-user license if the nine infrequent users could space out their sessions. But there are still problems. What if you can’t anticipate what your “peak demand” will be? What if the demand is not uniform, but occurs in “spikey” fashion?