Widespread 3.0 Upgrade By T+30 Days?

When do you start building apps that rely on capabilities in 3.0?

I’ve been thinking about the answer to this question while putting together the class for the Mobile Orchard Workshops, our forthcoming iPhone development training seminars that’ll premier in May.

At first blush, it looks like an easy question to answer: build for 2.X unless you need some of the fancy-pants features. Thing is, many of the standard-pants enhancements in 3.0 represent leaps forward that’ll impact building all but the most basic apps. The most obvious example of this is Core Data; most apps need data persistence and Core Data beats the pants off of using raw SQLite.

So, when do you start building for 3.0? There are really two factors: when Apple releases 3.0 to the public, and how quickly it’s adopted.

For the former, other than to say “summer,” Apple is characteristically vague about about when to expect 3.0.

For the latter, AdMob has some guidance. The chart shown in thumbnail size above and full size at their site shows that somewhere north of 70% of iPhone owners will upgrade to major releases within the first 30 days, with the number creeping to between 80-90% within two months.

There’s some nuance in AdMob’s numbers:

1. They show that, through mid-last month, only 65% of the people had upgraded to 2.2.1, while an additional 26% were on 2.2.0. IIRC, Apple and AT&T don’t push the third-dot point releases as much as they do second-dot releases, so I think it’s safe to combine those two numbers to make predictions.

2. By a similar token, the 7% still on 2.1 seem least likely to upgrade to 3.0 when it arrives.

3. While nobody (save Apple) knows the relationship between upgrade installation and third-party app usage, it’s reasonable to expect a correlation.

So, when will you start building for 3.0? As always, we’d love your perspective. For folks working on apps to ship in early summer: are you using 3.0 features expecting an early summer release and risking sitting on your app?