Marketplace Statistics Documentation
Last update: .
How revenue is estimated ?
We fetch number of installations and users of app from Atlassian Marketplace. We take totalInstalls, totalUsers and addon pricing. Those are only data available and everything else we have to model based on transactions from our addons. We use Gamma distribution to model users number per installation. We have average number of users per installation totalUsers/totalInstalls and we shape the distribution curve based on our applications. We assume only half of installations have active license and only half of this number is taken into estimation. The same goes with the users number. Each pricing plan (server, cloud annual, cloud monthly and datacenter) has its own distribution shape value. The following table presents current coefficients:
|
payment plan |
installations |
users |
distribution |
|---|---|---|---|---|
addon with free cloud tier |
server annually |
0.012 |
0.050 |
0.030 |
cloud annually |
0.050 |
0.150 |
0.180 |
|
cloud monthly |
0.425 |
0.125 |
0.050 |
|
datacenter annually |
0.013 |
0.175 |
0.300 |
|
total |
0.500 |
0.500 |
- |
|
addon without free cloud tier |
server annually |
0.050 |
0.050 |
0.500 |
cloud annually |
0.050 |
0.050 |
0.500 |
|
cloud monthly |
0.350 |
0.200 |
0.500 |
|
datacenter annually |
0.050 |
0.200 |
0.500 |
|
total |
0.500 |
0.500 |
- |
How top rating is calculated ?
We wanted not only average stars to be taken into account when rating the users favorite addons/vendors but also the number of reviews. We believe that for example addon with 3.9 stars and 200 reviews is much more appreciated than addon with 4.0 starts from 20 reviewers. In order to accomplish such rating we had to merge two scales: average stars and number of reviews into one. Our rating “adds” up to 0.25 stars for addons with thousands of reviews, is neutral for addons with 128 reviews and drastically discriminate addons with only dozens of reviews. The exact formula is:average_stars + 0.25 - ( 32 / ( reviews_count + 1 )).