Unfortunately, after New York, the potential to deliver goods economically in other cities falls quickly. Most major cities offer around $1 million per square mile in sales each year — equal to less than $3,000 per day. And that's the total for all online sales of physical goods, not just the fraction that people want delivered instantaneously. Worse yet, that fraction may have to be shared by several local deliverers.
High Last-Mile Costs
Not only are the sales spread across lots of "last miles," our experience suggests that it costs a lot of money to get there. To demonstrate the economics, we conducted some top-down analyses of one low-cost player and also built a bottom-up cost model for a local delivery business. Extracting from publicly available information, we estimate that Kozmo.com, absent overhead costs such as advertising, currently spends around $10 to make a delivery. Even this cost estimate reflects a heavy bias for its original Manhattan-based operations — the densest delivery area in the country, and thus probably its lowest-cost market. With an average order size of around $15, it's little surprise that Kozmo.com is losing money.
As any New Economy entrepreneur knows, scale offers the obvious solution. But unfortunately, physical delivery does not benefit from the network effect that supports other types of information-economy businesses. As our cost model demonstrates, variable labor costs drive local-deliverer economics, with drivers delivering packages comprising the bulk of the cost. (See Exhibit 3.) A van-based deliverer gets some economies by fully utilizing the vehicle space, but a bicycle courier can carry only a limited number of items. More customers simply means more bicycle trips.
Exhibit 3: Breakdown of Total Distribution Costs
Of course, the impact of the delivery cost depends on the value of the package being delivered. According to Goldman Sachs, an online purchase of $100 incurs an average charge of 8 percent shipping and handling. For $50 orders the research reported an 11 percent charge. As shown in Exhibit 4, increasing shipment density slightly reduces the absolute cost per delivery — but changing the value of the package dramatically changes the relative cost per delivery. For a $100 package, a local deliverer comes close to breaking even with the current average cost for shipping Internet goods. That means a local deliverer that charges standard shipping and handling fees could offer same-day delivery for the same cost as the delayed shipment from a category killer — clearly an opportunity to gain an advantage.
Exhibit 4: Delivery Costs as a Percentage of Shipment Value
Unfortunately, average online orders typically run in the $50 to $100 range, not higher. In fact, larger orders seem antithetical to the notion of instant gratification, which is more about impulse videotape rentals than about, say, the purchase of VCRs. And our research on the current delivery firms shows the companies with the greatest emphasis on rapid delivery tend to have the smallest order size.
New Tradeoff: Speed Vs. Variety
More fundamentally, under current models, the local deliverer resolves the issue of instant gratification at the expense of limitless offerings. To achieve fast response, the local deliverer must hold product locally, rather than in large national distribution centers, as category killers and large catalog retailers do. So the speed advantage gained by Kozmo.com and Urbanfetch.com Inc. means a variety loss. Kozmo offers about 15,000 items in total, versus more than 10 million total items at Amazon.
Even applying the Pareto principle — that 80 percent of the sales dollars come from 20 percent of the items stocked — the local deliverer will be able to steal only a fraction of the total sales volume from a category killer: that represented by high-volume items. Typically, though, high-volume items provide the least profit per unit because of heavy price competition. Furthermore, at least today, the category killer's broader product line provides leverage in negotiating with suppliers. So even though the local deliverer addresses the instant gratification need, it presents a new trade-off of a limited selection. Nothing comes free.