Android Device Makers: Ask, And Ye Shall Receive

Use the Community to Help Build a Better Product

The T-Mobile G1 has had storm clouds over it since shortly after its introduction, mostly stemming from on-board storage. Initially, the complaints were targeted at lack of space for installing third-party applications. Now, the complaints are lack of space for the Android OS itself, which at some point will mean no more Android updates for that device.

Those complaints are no doubt annoying to HTC, T-Mobile, and others. The community at present is incapable of understanding the design trade-offs that go into making a device, because the community lacks all of the information. There may have been very good reasons for limiting the on-board flash, from cost constraints to parts availability to simple miscalculations.

One lesson that device makers should take from this episode is that the community can be rather noisy. It is not uncommon for memes to be started in blogs and jump to more mainstream media, extending their visibility. This is far from unique to Android, of course.

Some firms may take the angle of trying to avoid the public. This works, so long as you are not trying to sell anything. Since most firms sell things, eventually the public winds up face to face with the products, if not the firm, and the complaints will simply begin then.

An alternative approach is to try to engage the community early on. Give the community a sense participation, and you may be able to stem some of these complaints. Unfortunately, all too often, firms try to get input too late, or from too few people, or with too many pages' worth of NDA limiting the community's chances for input. But it doesn't have to be that way – after all, if you can crowdsource finding veins of gold ore, you can probably find ways to crowdsource major device features.

For example, offer up a budget-style survey, where participants can allocate points among a series of device criteria. Some criteria may be expensive (e.g., 100 points apiece for “extra memory for firmware, to perhaps extend the number of OS upgrades” and “extra memory for third-party applications”), and some may be cheap (e.g., 20 points for “ships with useful case”, 5 points for “artistic laser engraving on battery cover”). Voters only get so many points, insufficient to vote for everything, so they have to make decisions. This serves two purposes:

1. You gain useful opinions, to gauge interest in various capabilities, which might steer product decisions
2. Those who participate in the survey start thinking about implementation trade-offs, hopefully realizing that for everything they want added (e.g., more OS updates), something else has to be traded away (e.g., more space for apps)

The community is oblivious to the design decisions that go into device manufacturing. If you want better criticism of your decisions, give us structured ways to provide such criticism, early and often. We are your consumers and your biggest source of fans. We are not the enemy, even if some days we might seem like it.

• Google
• Wireless / Mobile
• Android
• Android devices
• T-Mobile G1