Let’s unpack that.
There is a war going on right now for the last mile deliveries space with players as diverse as Uber, to DoorDash, to Walmart. But what is last mile delivery (LMD)? It’s the part of the delivery process that ends at your doorstep, and it’s also the part that costs the most per package delivered.
Cost was not an issue in an era where customer expectations were lower, they were willing to pay for the delivery, or the items they ordered were valuable enough that the sellers were willing to pay for the delivery on behalf of them.
Today, LMD is increasingly dominated by on demand small ticket high frequency orders with shorter expected delivery times. Instead of a dress from a catalog that cost $50 and could be delivered in a few days, think of groceries that cost $20 and need to be delivered within a day, or a burger that costs $5 and needs to be delivered in 30 minutes.
Even something as common as ice cream can be differentiated by offering unique flavours or just more flavours (Baskin Robbins) or claiming to use high quality ingredients that make you feel good (Haagen-Dazs).
How do you differentiate a door step delivery and claim a higher price for it?
The gig model was useful to scale delivery fleets quickly. Venture capital kept the party going with driver subsidies initially, but as companies try to move towards profitability, the real costs to your friendly Deliveroo guy are becoming apparent. Workers are demanding to be recognized as employees and for better pay and conditions.
How can you offer free delivery, make a profit and also pay a living wage to your workers?
The sellers/merchants who sell the stuff you deliver are facing intense competition themselves and are constantly looking for a better bargain. All while their expectations also increase in lock step with their customers’.
Can you compete for business on dimensions other than cost?
As the space matures, companies find themselves squeezed on all sides.
As the space matures, companies find themselves squeezed on all sides, with what was once impressive now becoming table stakes. Expectations are becoming asymptotic with no room for anyone who fails to meet them at the highest levels. This also commoditizes the space, leading to cost as a major dimension to compete on.
I predict that the existing crop of companies will respond to this in one of two ways.
1 They will exit the LMD space. Either by shutting down or by decoupling from their delivery ops and moving upstream, leveraging the experience they gained so far to become (or go back to being) a merchant/seller and focus on offering unique value to their customers, leaving the commoditized delivery to the the experts.
As an abstract example, may be one of the food delivery companies that is losing market share (and consequently also losing scale) will decide to focus on developing a dark kitchen network & better food packaging and use their erstwhile competitors’ delivery networks, instead of fighting to stay relevant in a space that is slipping out of their hands.
2 They will double down on deliveries. This is a hard but potentially rewarding path, with monopoly/duopoly tendencies and massive scale.
Anyone can continue operations indefinitely with VC funding. The only catch is that VC funding is not indefinite.
To survive long term in the space on their own merit, companies have to extract value at every micro-step, optimizing every second of the journey to be able to meet high expectations and also make a (very tiny) fee.
The low revenue per transaction means that massive scale is the only way to make a profit.
But that still needs at least a tiny profit to be made on every transaction, which I think can be achieved by early investment into optimization — Particularly into maximizing the volume of deliveries per worker per hour.
Driving this metric depends on delivering many orders simultaneously and continuously setting up the next step for the workers so that no time is wasted, using highly accurate predictions.
Finally, the key to making these predictions is the ability to ingest and process massive streams of real time data along with deep knowledge of the delivery environment including the streets, routes, traffic conditions, which elevator to take to get to the customer’s door, how quickly you can expect the concierge to open the door for your worker, and more.
An example of this is Uber. Uber has invested heavily into real time awareness and knowledge tools and is also open sourcing some of them. Their engineering blog is another indicator of their level of investment. They created their own geospatial indexing system, a visualization tool, a DB, their own maps & routing, and more because the products available to everyone in the open market just aren’t sufficient to build the kind of efficient, predictive knowledge based systems they need to survive in the long term.
If done well, this would lead to a virtuous cycle for the winner.
Efficiency allows them to offer the best costs and attract more deliveries (scale). Each delivery adds to deep knowledge. Knowledge increases the accuracy of the predictions. Accurate predictions help to keep the costs down or even decrease them, attracting more business.
In the end game, early movers will widen their advantage over time and edge out late entrants with insurmountable scale and deep experience gained over billions of deliveries.
Automation is being projected as a way to drive down costs, but it is costly to develop, fraught with political & regulatory issues and frankly the role of automation is over hyped.
I see some use of drones to expand reach to rural areas, but the core of the demand in dense urban areas will continue to be served by humans.
I can write more on automation, and “LMD: Renaissance” that answers the questions I raised in this article if you are interested in reading them. 😃
Let me know in the comments.