Why Late Cutoff Times Eat Into your Margin
Updated: Aug 28, 2022
The Last Order
We have a handful of Tarot Routing users in specific industries where the Last Order Cutoff is less than 30 minutes before the drivers depart.
Specifically, this means:
Orders (to deliver something) can be placed until 10am, for example
then the drivers leave the warehouse to start delivering at 10:30am.
This is great for their customers.
Some of our users even flaunt it as part of their competitive advantage.
And in some industries, like automotive parts delivery, the end customers expect this level of service.
However, there are two import downsides to having such a late Last Order Cutoff.
The most obvious downside is the scenario when you deliver to the same customer (or nearby customers) several times in the same day.
Again - this is fantastic for them, but it means you are multiplying your cost of delivery.
For example, if you deliver 5 parcels each day to a customer:
How much does it cost you if you deliver all 5 at once?
How much does it cost you if you deliver them across 3 visits in the same day?
Are you adequately compensated for this?
Or are you eating into your margin by providing this level of service?
Optimisation and The Loading Dock
This one isn't so obvious.
In fact, many of our customers hadn't considered it.
In the industry, everyone knows that you can substantially reduce your delivery costs by using route optimisation to plan your delivery routes.
But one of the major benefits of route optimisation is that an algorithm decides which vehicle delivers which parcel.
Yet, somehow, these parcels need to get from your warehouse shelves, onto a loading dock, and eventually into the delivery vehicle.
Imagine this timeline:
10:00am is your Last Order Cutoff
10:05am you run your route optimisation algorithm (you gave them 5 mins leeway after cutoff!)
10:08am routes are planned and dispatched to drivers and warehouse pickers
10:08am → 10:28am you only have 20 minutes to get all the parcels into the right vehicles. You couldn't start before now because the Allocation of parcels to vehicles wasn't decided yet.
10:30am drivers are supposed to leave.
Is it feasible to get all your parcels into the vehicles in 20 minutes, given that the warehouse pickers don't know which loading dock to put each parcel on ahead of time?
Let's call this problem the Loading Time Problem.
This is probably one of my least favourite words. Yet workarounds seem to hold the world of logistics together. And it seems like after a few short years, these workarounds quickly become "that's the way we've always done it".
Workarounds quickly become "that's the way we've always done it"
Nonetheless, here are two good workarounds we've seen to the Loading Time Problem:
Instead of letting the optimisation algorithm have free reign over the Allocation, you could partially restrict the Allocation so that each Parcel could only ever be delivered by one of 3 drivers.
This means that those 3 drivers would all be loading parcels from the same pile (instead of having a separate rack for each vehicle).
This seems to be a pretty good compromise:
pickers can put all the parcels into the appropriate 3-run pile before the cutoff.
drivers still have to sort through parcels while loading, but it's a much smaller pile. Probably feasible in the 20 minutes.
the algorithm has enough freedom to find a pretty good solution, even if it isn't as good as it could be without this restriction.
Instead of optimising routes after receiving all the orders, you could optimise early, say 30 minutes before the Last Order Cutoff.
Following the timeline example above, this would give you 50 minutes (instead of 20) to sort your parcels into the right docks. This is likely enough time.
Then, for the remaining orders that come in after optimising, your optimisation algorithm can probably insert them optimally into the already planned routes. (Well, you can if you use Tarot Routing's Insertion Optimisation feature!)
So again, not perfectly optimal, but a pretty good compromise between optimality and the Loading Time Problem.
You have to keep in mind that this approach assumes the remaining orders won't change the planned routes that much, since most orders have already been placed.
But this is a big assumption.
And experience shows us that it often isn't true.
It's just human nature. Most of your customers are going to place their orders right before the Last Order Cutoff. We're all procrastinators. That's just how it is.
The right solution is to move your Last Order Cutoff to at least 60 minutes before your vehicles depart.
Yes yes, I know your customers want to get their stuff now!
And yes, I know that your competitive advantage is that you deliver so soon after the cutoff!
But allowing an Optimisation Algorithm to Sequence and Allocate your deliveries will reduce your delivery costs by 30%.
Are you really willing to increase you largest operating expense by 30% just to provide a slightly quicker delivery service? What would your customers say if you passed on some of the saving to them?
It's a no brainer.