Quote:
Originally Posted by Vincent Ferrari
What I said was pricing your application in the $0.99 tier or the $0.99+ tier isn't an automatic predictor of your sales.
|
Actually, that was your second point, and I didn't disagree with that (hence my not quoting it). Certainly a useless app at $0.99 won't (or shouldn't) sell well.
The part that I disagreed with (as did Marc) was your statement that "The takeaway here is obvious: pricing your app doesn't, in a tangible way, affect the sales of that application." That's clearly wrong, and not what the study tried to show.
Maybe what you meant (and what the statistics tried to prove) was "the pricing of
other apps doesn't have much affect on sales of
your app," but that's not what you said.
Quote:
Originally Posted by Vincent Ferrari
Hockenberry's point was that $0.99 apps are hurting sales of more expensive apps by virtue of their price. That simply isn't the case and almost all paid apps in almost all categories are distributed equally between pricing tiers. If Hockenberry's premise was true, we'd see a huge disparity in application sales between $0.99 apps and $0.99+ apps.
|
Except, as both Marc and I pointed out, "$0.99+" apps include both "no-brainer" (good phrase, Marc) prices (like $1.00 or $1.99) and apps that are actually comparatively expensive (like $9.99 or more).
Quote:
Originally Posted by Vincent Ferrari
It would also seem obvious that cheap apps would sell the hell out of paid apps and, not counting the free ones, that isn't even close to happening.
|
I haven't looked at the App Store and its apps, but maybe the cheap apps (in general) aren't things people would really want anyway, and the expensive apps (again, in general) are things people really do want. There are lots ways to justify those statistics. Remember the famous saying, "There are lies, damn lies and statistics."
As somebody who took upper level classes in both probability and statistics in college (and has a degree in math), I know that doing a proper statistical analysis isn't easy (look at how polls work, for example). Did the authors of that study do a proper statistical analysis? I don't think so, and my original post pointed out several reasons why (the two biggest being that the sampliing brackets weren't designed well and the study didn't compare similar applications well enough). It was
interesting, certainly, but that doesn't mean it was accurate.
Quote:
Originally Posted by Vincent Ferrari
That was the premise of Hockenberry's "open letter," so they took his claim and analyzed it. You should probably read it when you have a chance. |
OK, I read it (or parts of it). It seems like he backed his reasoning with numbers that applied to
his situation. The statistical analysis looked at all applications. Again, as both Marc and I pointed out, that wasn't comparing apples to apples.
In fact, it seems like Marc and I both had many of the same conclusions. Maybe my original post wasn't just picking on you.
So, what would I suggest? I think case studies of similar apps with widely differing price points would be more likely to refute Hockenberry's broader premise, especially if the apps competed with his. However, I don't see how you can claim the broad study refutes his narrower premise (dealing with his own applications), especially when he provided detailed numbers.
Steve
Finding Fault
Linear Mode
