Eight years in, the lesson is still process

Eight years is long enough to get bored of your own advice. It's long enough to watch software get prettier, 3PLs get more regional, and "visibility" get sold as a product with a capital V. It's long enough to watch the same meeting happen with better coffee. Someone wants a shortcut. Someone wants a platform. Someone wants SEA to behave. The floor, in every year I have been on it, still wants to know what to do when the scan fails.

I started writing this stuff down when I was building a cross-border company out of Singapore. I kept writing it when that company wasn't mine anymore. I kept writing it while running ugly lanes inside a 3PL across six markets, and while keeping a small advisory practice on the side because the questions didn't stop just because my day job had a different logo. The logos changed. The lesson didn't.

Eight years in, the lesson is still process. Everything else is a way of repeating it, or a way of avoiding it.

What changed

The tools got better. Of course they did. A decent WMS in 2025 is a different animal from the things people were wrestling in 2017. Marketplaces got more precise about punishing you. Customers got used to a scan that means something. Air capacity went from impossible to merely expensive and back again. Indonesia rewrote rules, then rewrote them again. None of that is news to anyone who moved a parcel in this region.

Head office language got smoother. APAC. Control tower. End to end. Partner ecosystem. I can give you those sentences. They are not useless. They are just not the work. The work is still: who owns the step, what "done" looks like, and what happens when it isn't done.

Volume got bigger for a lot of brands, then smaller, then bigger in a different channel. 500 a day to 5,000 is still the same cliff. Heroes still drown on that cliff. They just drown with a nicer dashboard now.

What did not change

People still buy software to avoid an argument about definitions. They still treat a 3PL as a brain they didn't have to hire. They still launch a region because a map looked dense. They still call a trust problem a visibility problem. They still leave the Ferrari in the carpark and tour it for visitors.

Customs is still a process with a government attached. Last mile is still where the slide meets a street with no name. Exceptions are still the job. I can write those sentences in my sleep. I wish I couldn't.

The other thing that didn't change is the operator who does the boring work. They walk the order. They name the pile in the corner. They brief the partner like a grown-up. They look slightly unfashionable in the meeting, and their customers complain less. That person is still the whole game. I would like a more interesting conclusion. I don't have one.

The temptation to get clever

After enough years you start wanting a new thesis, because process sounds like something you should have graduated from. You didn't. I didn't. The region didn't. Clever is how you skip the whiteboard and go straight to architecture. Architecture is a lovely place to hide.

I've been tempted. A new tool, a new model, a hybrid that was actually a muddle, a regional contract that was going to save everyone from having to think per market. Sometimes a tool helps. It helps after the sentences exist. Never as a replacement for them. That is not wisdom. That is scar tissue.

If a new idea cannot survive a failed delivery on a wet Tuesday in a market you named, it is not an operating idea. It is content.

What I would still do on Monday

Write the process as it runs, not as the SOP wishes. Pick one market. Define the events that a human can trust. Put defaults on the 8 percent. Then choose the labour, the software, the 3PL, the in-house floor. In that order. I have not found a better order. I have found many more expensive ones.

Keep a brain even when you rent the hands. Especially then. A partner will faithfully run the brief, including the broken part. That sentence has paid the rent on a lot of QBRs I would rather not sit in again.

And stop calling it APAC when you mean a job. I will probably fail this one in a meeting next month. The pull of the blob is strong. The dock does not care.

The goal was never a new lesson

The goal was an operation that still worked when the logo on the polo changed, when the volume doubled, and when the person who remembered the folklore was gone.

Eight years in, I still start with the process. If that sounds repetitive, good. Repetition is what operations is. The clever stuff can wait until the scan means something, the hold has an owner, and the customer is not being told a story. Then you can have a new thesis. I am still waiting for one that beats that.