Ask HN: Brainstorming Analytics. What's the best way to share ideas with others?
At this point I have seen hundreds of co-hort and funnel analysis. I know all about designing star/snowflake/EAV schemas across dozens of db dialects (be that BigQuery, Redshift, Vertica, etc). I have performed dozens of implementations of Facebook/Google/Segment/(you name it) analytical funnels.
All of the above is great when I am approached to architect analytics for yet another e-commerce start-up or a young SaaS business. I am starting to feel like 50%-80% of the work is repeatable (albeit, still requires custom modeling/SQL/ETL).
The problem I keep running into is the other 50%/20%. And that is the stuff I feel that actually makes companies money.
Some say that Data Science is the answer, but from my personal experience with a Data Science gig back in 2013, I beg to differ. Many companies just don't know what those long-running drivers of growth are, so there is no point in optimizing with statistics or machine learning.
All of this is also new tech, so no one really knows what the long-running drivers of growth should be. People working at the same company often have contrary opinions. In this environment, Data Scientists just get the most credibility and time to try stuff out. But statistics won't in itself give you ideas.
With that I am starting to feel more and more the need for an analytics community. Some place where I could ask people about problems with the hope that someone can recognize similar patterns they've encountered.
Am I the only one?