Through the Seattle Times, here is a clip on the history of Amazon Prime. Below, I discuss two points on scaling Prime services.
Two Points on Scaling:
Scaling Volume: Prime was clearly a way to scale revenues through the volume of transactions, even at a huge setup cost, by introducing a subscription plan. An issue with scaling revenues this way is the stickiness of prices. It took a whole nine years for Amazon to go from $79 to a more profitable fee of $99. (I thought they would be raised to $108 at $9 a month – closer to NetFlix rates – but the fees were stickier than I had thought).
Scaling Speed: It is much harder to scale on the speed of delivery because localized demand tends to dis-aggregate and un-bundle orders. Storing inventory for fast deliveries also suffers from diseconomies of scale (i.e., pooling benefits diminish).
In other words, faster deliveries tend to be smaller orders, making it harder: To fulfill demand quickly, Amazon has to store chunks of inventory in several locations close to customers.
However, the penetration and clustering of Prime customers in geographical locations (once a certain scale was reached), provides quick identification of test markets for same-day shipping. I suspect that Manhattan borough of NYC, given its high population density and high income, has a high volume of Prime customers, so Amazon tends to see more predictable same-day shipping orders in the localized region, (compared to a less-dense place with fewer Prime members). Hence, it is easier to introduce same-day shipping, as long as Amazon operates fulfillment centers near NYC (Amazon does.).
Note that the service competition has pushed for scaling in Speed, which gets harder as time windows narrow. For instance, one-hour delivery at $7.99 per order, is much harder than a 2-hour delivery window, which is harder than same-day delivery, and so on.
Previous Last-Mile Posts: Last Mile #1, Last Mile #2, Last Mile #3