Part 1 of this article provided some graphs showing the rate of growth of Fotopedia. The main conclusion was that the growth was steady (“linear” for geeks), but not enough to propel Fotopedia into the realm of the really well-known sites. Part 2 contains additional data as well as presenting the data in better ways.
The left column shows the amount of photos in Fotopedia. From top to bottom:
- The number of photos in the entire Fotopedia database
- The number of photos in 3 Fotopedia Projects: Heritage, Japan, Amsterdam.
- The fraction (in %) of all Fotopedia photo’s within these projects (#1/#2 or for those who speak MatLab… #1./#2).
The middle column shows the number of articles. From top to bottom:
- The number of articles in the entire Fotopedia database
- The number of articles in the same 3 projects
- the relative fraction (in %) of all Fotopedia articles within these projects (#1/#2)
The right column shows the average number of photos per article:
- The average photos/article in the entire Fotopedia database
- The average photos/article in the 3 projects
Apart from some anomalies (spikes due to unusual contributor activity or to database cleanups), the conclusions from the previous article are still valid: there is constant growth, but at a relatively slow rate.
Both “constant” and “relatively slow” worry me: the site should exhibit exponential growth in this phase of its lifetime. Thus, for example, the release of version 2.0 of the Fotopedia Heritage app for the iPad (which happened a few weeks ago) hasn’t led to increased community activity – although it undoubtedly temporarily attracted some traffic to the servers.
PS: software tooling and more detailed view
I switched to using fancier software tooling to generate the graphs. The previous posting simply used Excel. Although this is party meant as an exercise to brush up on these scripting skills, the use of MatLab scripts allows me to automate a bit more and get more control over the output.
Thus, for example, the next diagram shows the daily increases in the number of photos and the number of articles. The vertical scale is logarithmic because of the large ratios between the various plots.
These plots confirm the constant growth: although a bit deformed by the logarithmic scale, the black and the blue plots shows no real increase or decrease over more than 2 months of monitoring.
Note that sample points with <= zero units/day are represented as 0.1 (10^-1, the baseline in the plots). Note I have no access to data for the Japan and Amsterdam Projects before November 15th.