"Does anyone give a shit about your dumb idea?" should be a bumper sticker
I dislike this style of communicating a point. There has been an epidemic of it on the web over the last year or two (that and its close cousin, the "X Reasons Why Your Y Sucks" post). Apart from being crude and overused, I think it sends the wrong microsignals. Many good ideas seem dumb at first. To me it indicates a deep misunderstanding of the creative process to blast anything in the delicate, incipient stage with this kind of harsh language. The most consistently creative people I know don't do it.
This may be a merely stylistic point, a matter of taste, but I suspect it's more. The way one puts things has effects. In particular, it affects what ideas and possibilities open up next.
I actually think it gives exactly the right signals. The people you'll have to sell to (customers, investors, future hires) will probably not be very understanding of the creative process. They'll think your idea is dumb, and they'll say "no thanks," or just hit the back button, without a second thought. The presentation is about convincing them otherwise, so it makes sense to put the problem statement front and center.
David does a great job of pointing out a mistake people make when interpreting the Lean Startup philosophy. It isn't about optimization, its about validating basic assumptions.
This presentation can basically be summed up as a verbal kick up the ass - "JFDI". For that, it's useful - beyond that, I feel it's much too shallow for the HN audience (we know this stuff... we've read Steve Blank, we're looking for more nitty gritty details).
I have a question does the churn rate he talks about (less than .8% of users stay more than a year) apply to startups aimed at business critical applications like inventory control etc.?
I do know that most business "software" is horrendously underused, but does anyone know the actual churn rates? Also does it vary for desktop applications? If so what types and why?
Those 3 chart types are the only data visualizations I recommend using. I should have left out the pie chart, since I don't use those.
I wasted many years of my life at Compete obsessed about "data visualizations" only to learn that no one except me and my statistical team cared about them. Please don't waste as much time as I did playing with exotic visualizations.
Ohh I am sorry I thought you were emphasising the data the charts were measuring rather than the actual types of chart themselves!
If you only mean that we should use only line, bar or pie charts, then I would say that scatter graphs at time have a good use of showing outliers, such as some guy who spends say ten minutes on the site, or visually able to show for example how many people spend ten minutes and how many 3. Communicate in other words much more information than a line, bar or pie chart can at times. I mean, people did not invent these other charts for no reason :)
Nice presentation, keeps it very simple. Oddly enough I signed up for his other startup performable the other day, seems his whole startup is oriented around helping you do this lol.
The graphic from the homepage of www.ABtest.com is also pretty self-explanatory about A/B.