Recent discussion of lessons learnt taken from the famous Microsoft ‘Bob‘ project (from around 1995).
They include (details and context given in the original post):
1. Never underdeliver against expectations.
2. Consumers don’t care about strategy.
3. A small marketing budget can work wonders.
4. If you start to get feedback from customers that your product is anything but great, don’t forget that you only get one chance to make a first impression.
5. Don’t be afraid to take risks.
6. Place bets on smart people who push the envelope.
7. Never forget the crucial role influentials play.
8. If it doesn’t work the first time, be open to the idea that it might work down the line.
9. Don’t be afraid to poke fun at yourself.
Taken at face value these observations apply to many product development projects – they’re totally generic and to that extent also fairly predictable. The interesting part is how these observations were taken into account (assuming they were) for future projects, this is the tricky and messy part – see also here.
Also, the benefits of proposed changes emerge over long timescales and this needs to be proactively managed or the desired results are unlikely to appear (at least in the way intended) see eg here.
The other issue that is customarily not emphasised is the role of politics, even though this may often be the over-riding constructive or (more usually) destructive factor. To establish real change you need convincing and sustained management understanding and commitment supported by process improvement and solid knowledge management practices. In my experience, getting the former is harder than getting the latter!
Update: detailed background and discussion of the project at the Technologizer via The Bob Chronicles.