This posting is spectacular in quality compared to what we usually get to see here.
1) It uses proper graphs for information display. Note the box plots with quantiles, and properly scaled axes. And the first plot which denotes the confidence intervals.
Yes, it makes it possible to clearly see all the errors.
Someone remarked that the numbers add up to more than hundred percent for the first year, and the confusion is because the y axis is mislabelled. It should say "Cost for team as percentage of revenue".
Also, the first two years shouldn't be plotted as part of the line as they are just two data points in total for these years and since there is considerable spread in better represented data we know any single data point is probably not representative. At least not if you insist on using decimals.
It's also a bit unclear what the line for "more than seven years since funding" comes from, since the provided summary table only shows up to year 7. It seems the data table truncates, but there are two companies that cover age 16-24 in the spreadsheet, once again giving a very small sample.
On that first graph, for year one the percent of revenue spent on the team (using the means) exceeds 100%...? The mean for R&D is ~50%; for S&M it's 90%. How does that work? Is this due to fundraising?
This is one case where it would be nice to see the number of votes on a comment--I upvoted you because I agree immensely, and even though I also upvoted the submission there is no way for the community to know why things are getting up voted so we have to write follow-on comments like this one.
Many thanks to OP for an excellent write-up and I, too, would like to see more of this quality content on HN!
1) It uses proper graphs for information display. Note the box plots with quantiles, and properly scaled axes. And the first plot which denotes the confidence intervals.
2) Actual numeric data is shown in tables.
3) The source data is avialable.
Well done! More of this, fewer "Infographs" plz.