First, I noticed that the discount selected in "Name your discount" is not the discount that Purse gives. What Purse does, is it takes the price of the item, takes off the discount, but then adds on their fee. The discount you name, is the discount before their fee. So I decided to plot the Purse Discount against the Actual Discount (where the purse fee is subtracted and the discount recalculated). You can see that selecting 50% discount only gives just over a 35% discount after fees.
I also noticed that the fee that Purse charges is 1% (up to 10% selected discount), and then increases exponentially. So I plotted this in a graph too. You can see that the high discounts result in a high fee. Selecting 50% discount results in a slightly above 14% fee.
Of course, the higher the fee you select, the lower the probability of your order being fulfilled. They give an estimate on their website, so I decided to plot that in a graph too. I expected this graph to go down in a smooth line (probably exponentially), but it appears to go down linearly in a stepping fashion. Initially, there is a low impact on success for small increases in selected discount - but there is a tipping point at around the 12% mark.
Last but not least, I was curious what the purse estimate graph for delivery times looked like. So I plotted their estimated delivery times. Where they gave a range, I chose the upper number for this range. In the end, they give '20+' estimates with no upper value, so I just took that value. You can see that it increases quite linearly up until 44% and then increases significantly.
It would be interesting to plot the data from the merchant section of Purse, to get a feel for at what discount level people are placing their orders. I guess I could check that list every so often to get an understanding of how long it takes for an order to be removed from that list. Maybe another day! I thought it would be fun to share this data here.
