Last-Mile Deflation Part 2: Autonomy And Free Distribution

The decline in ecommerce delivery prices outlined in Part 1 of this post, Retail, Ownership, and Deflation in the Last Mile, has been driven mostly by business model innovation (UPS/FedEx, subscriptions, marketplaces) rather than a change in the underlying cost of transportation. But that may change soon. Just car ownership grew from a few million (predominantly wealthy) households in 1917 to mass adoption over the 20th century, the few hundred thousand electric vehicles of 2017 with partial autonomy capabilities are multiplying. This could produce a second transportation revolution, no less dramatic in scale than the one that took place a century ago. And just as the advent of the automobile brought massive, unpredictable changes in both consumer behavior and retail geography, so too will autonomous vehicles.

In that vein, I think it would be entertaining to extend the insights from the last post to the future of commerce and transportation. Some second-order effects will be more straightforward than others. For example, the cost of transporting goods will decline and a portion of the savings could be passed onto consumers through lower prices. However, my focus here is on the more counter-intuitive set of predictions (and therefore the ones that are more likely to be wrong). Rather than attempt precise forecasts, the purpose of this post is to have fun imagining what commerce could look like in a world where product distribution undergoes dramatic deflation.

While some predict slower adoption, RethinkX published the AV bull case, predicting that autonomous electric vehicles will launch by 2021 and bring the cost per mile down to 16 cents and, eventually, 10 cents by 2030. This will be primarily driven by 500,000-mile vehicle lifetimes, far lower maintenance, energy, and insurance costs, as well up to ten times higher utilization as vehicles will transition from individual ownership to fleet ownership by Transportation-as-a-Service providers. For comparison, the equivalent cost per mile today is around 55 cents, while the lion’s share of delivery expense is labor-related. We can leave the debate on the specific numbers for a future post, but, for now, let’s ponder the seemingly strange behaviors and business models are likely to emerge from this stunning collapse in transportation costs.

From Store to Storage

While the last post looked at the decision to purchase online versus offline by comparing the marginal costs of same-day delivery and driving to the store, we saw that this is unlikely to remain a relevant distinction for the consumer of the future. As transportation costs collapse, consumers will instead compare the marginal cost of access (e.g. delivery) to that of ownership (such as storage, maintenance, depreciation).

To see why let’s run a thought experiment that will be familiar to an urban renter reading this post. A two-bedroom NYC apartment rents for $2,490 per month (much more, if you live in Manhattan). At an average of 850 square feet, we can assume a monthly price per square foot of $2.93. In fact, renters are already willing to pay more than $6 per square foot in NYC for on-demand services like Clutter to store unused goods away from their apartments while preserving some level of access. Let’s say your camping gear occupies about a square foot of your New York apartment. Even if you plan to camp three times this year, you’re effectively paying $11.72 in storage costs per use. That’s already a good deal higher than most delivery options you have available and, in 2021, it will be almost 15 times higher than having an autonomous vehicle drive five miles to deliver the equipment each time you need it.

But, of course, delivery vehicles will be driving more than just one package at a time. Let’s assume each vehicle takes a 20-mile round trip to a distribution center (delivery and return) with just four parcels per trip, which works out to almost 54 parcels per day at the 274 miles that RethinkX estimates autonomous cars will drive each day (50,0000 miles over a five-year useful life for each AV). By comparison, UPS trucks can deliver almost ten times that at over 500 parcels per day on an urban route. This implies a total cost per delivery and return of 79.5 cents in 2021 and 50 cents in 2030. Assuming 2017 rental prices for NYC, delivery for each use will be 24% cheaper than storage for bulky products that are used even as often as once a week. Even in Charlotte, where 2 bedroom apartments currently rent for $1100, goods used every other week (e.g. a vacuum cleaner) will cost more per use to store than to deliver.

But these delivery figures are estimates of the costs to the delivery business, not prices to the consumer. As we saw in the last post, businesses are perfectly happy to make losses on delivery in exchange for a deeper relationship with the consumer and increased product sales. These businesses will not just be paying five orders of magnitude less to operate delivery vehicles, but will also be freed of the largest delivery expense, the human driver. And not all delivery will occur by car, a combination of drones, droids, mobile lockers, and other technologies are already under development to make delivery even cheaper and more convenient. Furthermore, as some have suggested, autonomous vehicles have a multitude of other potential sources of revenue beyond charging for transportation. The last century of retail history gives us every reason to believe that instant delivery will be entirely free to the consumer in an autonomous future.

The Rolling Stores

One interpretation of the 20th century history of commerce could tempt us to conclude an inverse relationship between the cost of transportation and the distance between the home and the store. But the opposite view of retail topology could explain more: As transportation costs declined, the store got closer. Sure, the business of the local store moved farther out to the big box, but its function, to hold local inventory that consumers can quickly access, moved to our homes. Our local grocer was replaced by our fridges and pantries and our local hardware store moved into our garages and sheds. Meanwhile, warehouses, in the form of big box stores and, later, ecommerce fulfillment centers, started selling goods directly to consumers. What conclusions does this relationship lead us to, assuming an imminent collapse in transportation costs?

For one, local inventory may stop being stationary. Retailers spend most of their waking hours worrying about maximizing turns on their inventory. Stationary inventory generates economic costs: space isn’t free, things are liable to break, decay, or get lost, and everything we own is massively underutilized. One reason why so many of our things sit around collecting dust is because it costs even more to move them to more productive use. In a future with radically lower transportation costs, more local inventory will be mobile. Many things could be placed in constant motion, quickly available when requested, but free to move along when no longer needed. This is, in many ways, an extension of the original model of UberEATS, where food is pre-loaded into the trunk as the car drives around, waiting to be plucked by a hungry user. Many functions of the local store could move into what we today know as the delivery truck.

Lifting The Weight of Ownership

We’ve already seen how the value of ownership eroding for the myriad of bulky, rarely-used goods that we hoard in our homes and apartments. Camping gear already costs significantly more per use in storage than delivery for a New York renter. But that’s just a thought experiment — t the real problem is that campers pay hundreds of dollars for everything from tents, sleeping bags, tents, and coolers to only use their gear a few times a year. For many, it already makes little sense to buy such products. As Seth Miller points out, autonomous transportation could have the same effect on physical goods as the internet did on music and movies by making the delivery of digital goods entirely free. As distribution becomes free and access frictionless, ownership of durable goods could go away altogether, even for smaller, more regularly-used products.

The transition to access threatens what has been the core business of retail trade since antiquity: selling goods for profit. Even the most differentiated retail and consumer goods categories are commoditized under access-based business models. Much like the last transportation revolution, many businesses, including many of today’s most dominant retailers and brands, will struggle to adapt to the radical shift in consumer behavior. However, before autonomous vehicles even enters the picture, last-mile deflation is already moving us closer to free instant delivery, mobile inventory, and entirely access-driven consumption of durable goods. Whether these ends-states are fully realized should matter less to incumbents as the core business model of retail is already starting to be challenged. One way forward for businesses to survive product commoditization will be the bundling of products with services, such as education, inspiration, personalization, and experiences. Pioneers of the access economy, like Airbnb through Experiences and Joymode, are already directly selling curated experiences bundled with products. Retailers and brands should take note.