In a nutshell, their approach to this daunting task combines an economic model underpinned by detailed data on online transactions with the geographical distribution of consumers relative to Amazon’s fulfilment centre network.
For context, in 2006, Amazon had eight US fulfilment centres. By 2016, it had 90, and by 2018, it will have over 100 facilities.*
The authors estimate Amazon’s revenues, shipping costs, fixed costs and profit per dollar of goods sold from 2006 to 2018 based on Amazon’s actual fulfilment centre network each year over that period and compare it to the counterfactual of Amazon’s network remaining the same as it was in 2006 for every year to 2018. The purpose is to isolate the impact of the fulfilment centre expansion over the timeframe, holding factors such as prices and demand growth constant.
The impact on revenue is seemingly strange. Under Amazon’s actual network in 2018 (based on planned investments), sales are estimated as $9.6bn lower than they would have been under the company’s network in 2006. But this result is largely explained by a tax effect. Generally, Amazon must only collect sales tax if it has a physical presence in the customer’s state. When it expands its network to a new state, it must then charge sales taxes where it may previously have not, hence the clear reduction in revenues.
Unsurprisingly, fixed costs are far higher under the actual network given the higher number of facilities, while shipping costs are considerably lower as the average distance to the customer falls. The key question is whether Amazon’s network strategy has improved profitability? The results are very clear: for its actual network in 2018, Amazon’s overall margin for every dollar sold is put at between 21.5 cents and 22.4 cents. Under the 2006 network in 2018, its margin is estimated at 18.4 cents to 21.3 cents, between 5% to 14% lower.
This profit gain is driven by a staggering 51% reduction in overall shipping costs, on the back of the average shipping distance from fulfilment centre to consumer decreasing by around 180 miles by the end of 2018.
Clearly then, the results suggest that there are economies of density (savings due to the spatial distribution of its fulfilment centres) associated with Amazon’s network, and that is even in the face of sales taxes dragging down revenues.
Amazon’s expansion appears to have been good for the consumer too. The extended network has the added bonus of enabling Amazon to offer more rapid shipping options. Furthermore, the authors state that Amazon’s prices have fallen by about 40% over the sample period, suggesting that “a significant share of the cost savings have been passed on to consumers.”
An interesting thought experiment is to pose similar questions for Amazon’s European network. A clear factor in Europe’s favour for maximising economies of density is the absence of a tax effect like in the US. Complicating factors include the presence of borders and country-specific factors which likely distort Amazon’s network from maximising economies of density, just as US tax law has.
It may be difficult to overcome such barriers, but that does not mean nothing can be done. According to this paper, lengthy transit times for longer distances make online consumers reluctant to purchase goods outside their home country in the EU, with cross-border e-commerce in the EU less developed in terms of transit times than interstate e-commerce in the US. In 2014, US online retail sales were €224bn, 43% greater than the EU, despite the EU’s GDP being 6% higher. While the land area of the EU is just 45% that of the US, it has similar or longer transit times due to so-called ‘border effects’. The key finding is that European cross-border e-commerce would benefit from improved express delivery services. Amazon’s economies of density and profits may well too.
Source: Transport Intelligence, May 9, 2017
Author: David Buckby
The report provides market analysis first from an industry-wide perspective, and then delves further to examine the supply chains of major e-retailers and the logistics providers which support them. The report includes Ti’s bespoke market size and forecasts at a global, regional and country level in addition to data and analysis of e-commerce logistics costs as a % of sales for 20 online retailers.