The Scaling Paradox in Logistics

Why I resisted spending a million dollars on an automatic sorting system

The year was 2016. I had merged ALM Group into iCommerce. From a CEO, I became the Director of Technology. We had built our own last mile operating system and it worked well. My CEO came to me that he was thinking of investing a million dollars in automation, so that we could raise more funding. Somehow, this is what VCs wanted.

I told him no. First of all, this was money we did not have. Second of all, it was not time. We were then doing about 3000 volume a day. Sorting took about 1 hour. He wanted to cut the time needed to sort. I get it. But we are a startup. And startups need to do more with what they have.

The funny thing about Logistics is that things like scanning in many 3PLs are still done manually. It took NinjaVan for example close to a decade to invest in belt sorting machines, and they were the largest 3PL in our region.

What is the Scaling Paradox?

The Scaling Paradox is the state of growing quickly without the tools, like automation, to manage the growth. Growing quickly without ways to manage the growth, may actually slow you down. For example, you might take on more projects than you can manage. In order to handle them all, you spread yourself thin, and are not able to produce the quality at the right speed.

But if we don't invest in automation, when we enter peak season, and we will, we would have a situation of 3 hour sortation times or more. We need to put bodies into the problem.

How might we reduce the time to sort, without the use of automation?

We decided to look at our processes, and figure out how can we change our processes such that we can sort as humanly possible, yet be as fast as automation.

So we held meetings and interviews with our ground team, in collaboration with our engineering team, and came out with a few ideas.

1. We wanted to reduce all the scans we needed to sort a parcel, to just 1 scan per parcel.

2. We wanted to introduce mobility, without introducing mobile devices.

3. We wanted to be able to sort, even if there was only 1 man on the ground, and he must be able to sort parcels in a single wave in less than 45 minutes.

The team came up with great ideas during the brainstorming session. This is what we did.

1. We would introduce mobility by using trolleys to put our laptops on. We would also use roll cages to do an intermediate sort to a range of postal codes. All our sorting bins were already arranged in sequence. So the roll cages were marked 01-20, 21-40, 41-60, 61-82, (these were prefixes of the postal codes), CX/PS (Customer Experience/Problem Solving - where problem parcels and international would be sorted to)

2. We got rid of the secondary scan, because once we introduced the roll cages, we did not need it anymore. Sorters would simply roll to the sorting bins and drop the parcel in each bin.

3. From this change in processes, we got rid of all other scans, and the entire sorting process took 1 scan.

We then made the changes to the software, and updated the SOPs. We arranged to release the code on Tuesday, and tested our hypothesis.

Based on a 1 man team, we were able to scan and sort our average daily load in 40 minutes. With a 3 man team, we were able to do it within 25 minutes.

The best win - We did not spend a million dollars to realise a gain in our productivity. And everyone got to keep their jobs.

Photo by Tiger Lily

The Epson Problem for Multicoloured Pens
How might we replace the ink of one colour without buying a new multi-coloured pen?