The Delivery Economy and its New Building Blocks — Part I
The growing online economy which now delivers services not just products to users’ homes has necessitated that companies also run logistic functions internally. Simply put, it has now become impossible to separate the quality of delivery from the actual product/service in user’s approval ratings or NPS scores. This need now applies to everyone from e-commerce, healthcare, waste management, groceries, food, concierge to on-demand home services.
Global receipts from home deliveries or parcels had been $250bn just last year at the cost of $80bn that excludes pickup, line-haul, storage and sorting. This means the cost of only moving the good/service from a seller/warehouse to consumer’s doorstep, colloquially called the last mile, contributed one-third to the total alone. No part of this cost was directly borne by the end-user.
The premise is that building a delivery function is becoming increasingly expensive alongside retailing. Even in the hands of fulfilment specialists who’d be required to do anything from an on-demand, same-day to scheduled deliveries for the same retailer, a 5–6% net margin and only a clerical oversight to cut/control fixed costs, the road to serving 3bn deliveries by 2020 alone in India (and $18bn globally) looks bumpy.
So what makes it so incredibly difficult and expensive to keep making deliveries to growing set of online consumers.
- User Location: They are spread across 500+ towns alone in India (and 10,000+ globally) and it is difficult to map them accurately. Even with repeated orders, the quality of location access to users, at least in India has been very poor. Access means high rate of first time successful attempts that gets your delivery runners more productive hours and hence more deliveries. With 400 million deliveries done alone in India in 2015, lack of proper access meant employing an army-strength cadre on ground trained to remember customer addresses like postmen. Manageable?
- The Dispatch: Almost all deliveries are manually planned at the source, sub-optimally assigned to available runners with little/no understanding of faster delivery routes in a city. This either builds incorrect route memory in runners on customer accessibility or necessitates the need for too many distribution centres that require more such planners! Every wasteful manual planning hour saved at the source translates to 10% extra time doing deliveries.
- The Delivery Run: From Pick&Deliver kinds fulfilment to physical warehouses aggregating parcels, the only leg of the infra that can optimised dynamically is the delivery run. The delivery output of any runner is a function of the efficiency of the route he could take, intending to maximise the number of deliveries in a single day/trip like a traveling salesman.
The math could differ between sectors but the 400 million deliveries in e-commerce in 2015 would have costed $3/delivery, 1 million man hours wasted just planning them and required 15% more workforce than necessary. Would an operations stack just for last mile that can reduce fulfilment costs by 50%, bring down man hours to one-third the original and reduce biker count to 20% make sense? We are using AI and data science in creating these new building blocks for last mile. Check more @ www.transporter.city.