I recently tested a mobile recruiting mobile site for a friend over in the states, his company career mobile site was sold to him as supporting Android smartphones. I was not personally impressed with the mobile site, it tried a little too hard to behave as an iPhone app, instead of a website. I find such an approach slightly bizarre as it expects the Android user to be familiar with the iPhone UX. To achieve an App like experience the mSite included a significant amount of logic and 'ajax'. All this client side code ran ok on a iPhone 4S but on less powerful Android devices using the site was slow and difficult to use - bordering on pointless.
The problems were not limited to the performance and usability on older less powerful Android devices. At the time of writing around 50% of Android devices are version 2.3 (about 3 generations old) or less of which the majority are using the Android native web browser. The msite in question had not been tested on Android 2.3 with the native Android browser, there were many HTML problems making the mSite unusable on the Android native browser.
Developing mobile sites for Android is fraught with trouble.
- There are thousands of different models, each with different performance capabilities.
- There is a wide spread of different Android versions in the hands of users.
- There are a number of browsers.
- There are many different screen sizes and resolutions
To put it simply - the Android market is FRAGMENTED.
When procuring a mobile site be sure to talk to your supplier about their Android support and their test procedures. Question them about the different versions, ask to see screen shots of previous clients sites on the various devices they test on. Failing to drill down on what exactly is meant by 'Android Support' may leave you paying a bill for something far less.