In my last blog post, I explained how consolidating orders to create routing and scheduling plans is much more complicated than widely assumed.
So complicated, in fact, that doing the calculations manually often results in costly errors.
That’s why automating the process makes a lot of sense.
Still, some freight managers understandably find it difficult to accept that paying for an automated version of such a routine operation saves money and improves customer service.
To help logistics professionals understand the case for automation, I’ve put together a hypothetical case example. The example shows the pitfalls of crunching the numbers manually as well as the consequences of this error-prone approach to order consolidation. Here goes.
Consider the following order set and shipper requirements:
When mapped out, the set of orders looks like this:
You might be thinking that intuitively the solution is a slam dunk! There is a close geographic spread between the orders and the heavy ones are at the end of the run, so the solution is to put all of the orders on a truck.
On closer inspection, however, the slam dunk turns out to be a bust, because it exceeds the FTL pallet limit. We have 46 pallets—one more than the maximum allowance of 45.
That’s not all. Did you notice that the Kentucky stop has a relatively fast transit time and a single-day drop window? As a result, delivery would be late by a full day based on this solution. In addition, were the out-of-route miles and loading/unloading times at each stop allowed for when the Cleveland stop was considered? That one missed its drop appointment as well. Ouch!
This solution would cause two production lines to be shut down, and probably result in an order being left behind because it wouldn’t fit. Not good.
OK, let’s put our heads together and try again.
We’ll start with the obvious; combining the Buffalo orders. That has to be a good start.
Wait a second, the LTL limit is 10,000 lb. This shipment has to ride FTL now. Is it cheaper to ride FTL? Well, the direct LTL charges would have been around $720 for the two smaller orders and $1,126 for the larger one, bringing the total to $2,561.62. This is based on the Czarlite base rate 1/1/1995 with a 50% discount and FAK of 70. The FTL rate per mile is $1.75 with a total of 1,910 miles for a cost of $3,342.50.
So, consolidation actually cost us almost $800. That’s a tough sell, to say the least!
Should we keep the orders together or break them apart? That depends on whether or not we can fill the truck a little more. What do we do? That $800 deficit will be hard to overcome.
At this point we know that Kentucky must ride FTL due to time restrictions, and Cleveland has a short transit time and is over the LTL weight limit. Also, if the Buffalo orders ride independently of other stops they should be split to some extent; probably the 6,000 lb. order should be combined with one of the lighter orders to increase weight without exceeding the threshold.
But there are only seven orders. How hard can it possibly be to assess every combination to see which one works best?
Well, we could ship them all individually, put two together and let the rest ride separately, put three together and ship the remaining orders individually, combine two groups of two together and let the other three ride separately.
Now the exercise is getting more complicated. Which begs the question: How many different ways are there to combine these orders?
Well, actually, there are 7,563 different ways. And for each option we need to determine the cost, and whether or not any constraints are violated.
How long would it take to rate and find the best solution 7,563 times? Probably a significant chunk of a career. And that’s assuming we have a rating engine capable of accurately rating a hypothetical, multi-stop load with stop-off charges and out-of-route miles.
Back to the drawing board.
Let’s group by geography based on what we know. Cleveland has limited opportunity to add stops beforehand and is right on route with Buffalo. We can put these orders together. Whew! That actually saves $680. Can we go further and add the Kentucky stop? Success again, and it actually saves an additional $2,400 (considering that the baseline for Kentucky is FTL).
Now let’s consider the two remaining Tennessee orders. They are in close proximity to each other, so we can try the multi-stop option. No constraints are violated but the lightweight order comes after the heavier order. The baseline cost for the Nashville order is only $79. Therefore, with the $75 stop-off and a couple out-of-route miles, we are over the baseline.
At last we have the final answer: Multi-stop all of the non-Tennessee orders, and ship the Tennessee orders separately—one LTL, and one FTL.
Can we guarantee that this is the best possible solution?
No, not without going through the 7,563 combinations of orders and assigning best mode, carrier, and consider expedited, etc. for each configuration.
In other words, despite all our hard work, it’s impossible to know for sure whether we’ve come up with the perfect solution. All we do know is that we’ve saved money and didn’t violate any constraints.
Now imagine doing this calculation for a set of 60 or 750 orders—or 1,500 orders for that matter.
Could you undertake such an exercise without violating any constraints and in such a way that every consolidation saves money? How long would it take? And even if you were able to attempt this mammoth task, the risk of exceeding critical truck dimensions such as a weight limitation or shutting down a production line owing to an error is uncomfortably high.
As I explained in my first post on the subject, sometimes there are valid reasons for going the manual route when consolidating orders. But as this example shows, as the problem increases in complexity, so do the merits of automating the process.
At the very least, I hope I’ve given you some food for thought.