GD is so far my favourite end-user BI tool and eclipses all other cloud tools, while also enabling use of GoodData on top of a corporate 'gold standard' platform such as Business Objects, or just an Oracle based Enterprise Data Warehouse (EDW).
If you want to succeed with BI you need to embrace "Laisser Faire BI", in other words put a great product in people's hands and get yourself out of the way. Some formal structure may be necessary in some orgs, but the real adoption curve is driven by demand, not supply.
GoodData is most popular among "data hound" category of business user. Anyone who is capable of running and refining large Excel pivots, and who can extract SQL statement results into Excel, is easily able to build beautiful analytics and visualizations in GoodData.
Some less proactive GD users are merely consuming the analytics views created by their colleagues.
In one customer demo, we used online banking data (yes, for real) and observed as GoodData created different dashboards for different people depending on what it found in the downloaded data. It was very cool. Download 1 full year of online banking transactions into CSV, then create new project using this as base data. Where debit categories were clearly showing a pattern, the dashboard put this item first and showed a pretty useful pattern. where people had variable income and from various sources, it showed that first and created great interest in seasonal variances by source.
We let it go viral among the smartest and most data-hungry and analytical people in the organization. That generated some me-too following, and before we knew it we had a Mexican wave. The stragglers too were enthusiastic, especially when they could just use metrics created by their colleagues
GoodData is still a small, hungry and very capable organization. I do hope while they grow and scale up their operations they will maintain this hungry yet humble culture that is always there for their customers.