Ugh, I hate these graphs. Cumulative graphs make even moderate or bumpy growth look predictable.
I guarantee the company KPI here is acquired users / month, not total users. Look closely: January 2013 was probably a kind of scary month. October 2012 was awesome ... and they didn't repeat it until 2 years later.
This is great for investors and the public, but you can't drive with this graph.
The chart is a cumulative chart to hide the fact that the source data is only eight data points from various blogs. That's very sleazy and borderline misleading because it creates an extrapolation as mentioned in the comment above.
If you are interested in making this a series, use robust data, or don't do it at all.
Additionally, an area chart is the incorrect type of chart for this kind of graph, since the area under the curve does not represent the total number of users. (You'd use an area chart of the y-axis is number of new users.) A line chart would be more accurate.
Great point on an area graph being misleading - changed.
What data would you propose using for a private company like this? There is very little public data and this is the only KPI that is consistently reported. Something is at least a little better than nothing to understand how a business has grown over time.
This was one of the primary reasons I stopped doing analysis on startup metric data. Most of it is deliberately incomplete and it would be dishonest to do analysis on incomplete data.
In this case, "grown" is a loaded question from your metric data, since "number of developers" does not distinguish between the number of developers who have joined and who have left; it's only the total, but the former metric is what's important.
4 comments
[ 3.2 ms ] story [ 23.5 ms ] threadI guarantee the company KPI here is acquired users / month, not total users. Look closely: January 2013 was probably a kind of scary month. October 2012 was awesome ... and they didn't repeat it until 2 years later.
This is great for investors and the public, but you can't drive with this graph.
If you are interested in making this a series, use robust data, or don't do it at all.
Additionally, an area chart is the incorrect type of chart for this kind of graph, since the area under the curve does not represent the total number of users. (You'd use an area chart of the y-axis is number of new users.) A line chart would be more accurate.
What data would you propose using for a private company like this? There is very little public data and this is the only KPI that is consistently reported. Something is at least a little better than nothing to understand how a business has grown over time.
In this case, "grown" is a loaded question from your metric data, since "number of developers" does not distinguish between the number of developers who have joined and who have left; it's only the total, but the former metric is what's important.