Picking Optimisation: Clone Your Best Picker | Socius24
The Human Algorithm: How to Clone Your Best Picker
If you’re not already a Dispatcher WMS user, there’s probably a person working in your warehouse (because there’s one in almost every non-Dispatcher warehouse) who moves just a little bit differently from everybody else.
You might have noticed them without quite being able to put the right words to what you’re seeing.
They’re faster, but they’re not rushing. They’re more efficient, but they’re not cutting corners. It’s more subtle than that. They move through your warehouse in a manner that feels practically choreographed, hitting locations in an order that the system certainly didn’t tell them to, and covering an amount of ground that somehow seems considerably less than the ground that everyone else is covering.
They’ve been there for years. They know where everything is. And at some point, without anyone asking them to do it, they basically stopped following the route that they’d been handed and started doing it the way that they thought it should be done.
The WMS lets them do their own thing, too. It hands out a route, then your picker does something else entirely, and the system neither notices nor cares.
Just FYI, Dispatcher WMS both notices things like this, and it cares about them, too.
What your best picker is actually doing
What your best picker is doing, in that moment between looking at their list and starting to move, is running an optimisation.
Not consciously; they’re not sitting down with a spreadsheet. They’re clustering orders in their head, they’re identifying adjacent locations, and batching picks in a sequence that cuts out any need for backtracking… things like that.
And they’re doing all of that in roughly the time that it takes everyone else to reach their first location.
They can do this because they’ve done it ten thousand times before. Because they’ve got a spatial map of the warehouse that’s been burned into their memory so thoroughly that they could probably navigate those aisles in the dark.
They are, in the most literal sense, operating as a human algorithm. A brilliant one. But an expensive one to train, given that their education has probably been accumulated over years of showing up and paying attention.
The problem with Genius
The only problem with having a human algorithm on your team is everything that comes along with having a human algorithm on your team.
They can only be in one place at a time. They have a finite number of shifts. They’ll get ill, they’ll go on holiday, and eventually, inevitably, they’ll leave.
And when they do leave, they’ll take their algorithm with them. It can’t be transferred. It can’t be scaled. It doesn’t train the next person, except in that vague and rather unreliable way that proximity to brilliance can occasionally pass on a skill via osmosis.
Meanwhile, the rest of your team is still following the route that your WMS generated, which was configured some years ago, for an order profile that no longer exists, and it’s sending them all around the warehouse in a way that your best picker would consider, to put it diplomatically, ‘suboptimal’.
'Suboptimal' performance
Operations teams talk a lot about the differences between their best and worst performers.
They talk about it in terms of training, attitude, experience, and effort. And all of these things matter. But no-one talks nearly enough about how much of that performance variation is a systems problem rather than a people problem.
Your best picker isn’t faster just because they’re trying harder. They’re faster because they’ve built a better routing model than the one that your WMS is handing out. So, the gap between them and a new starter isn’t only a training gap. It’s an information gap, too. It’s the difference between donkey’s years of optimised spatial data and a route that was set up years ago, in a configuration review that nobody’s looked at since.
And that gap has a tangible financial cost.
It shows up in your labour efficiency numbers, in your cost per pick, in the performance curve of new starters, in how long it takes someone to become genuinely useful, instead of being there in body, but not in mind.
And of course, it all compounds at Peak. Because at Peak, you’re not scaling a team of veterans… you’re scaling a team of people who’re following a map, possibly for the first time, that your best people stopped trusting forever ago.
Directed picking: what if that expertise didn't have to evaporate?
There are two things that need to happen here.
First off, your workers need to follow your system. Dispatcher WMS can be configured to direct putaway and picking. Which means that the routes it offers aren’t simply a polite suggestion that your team can take or leave. The system directs what’s going on and the work completed gets recorded against it accurately.
Incidentally, Dispatcher WMS can also be (and often is) configured to interleave all kinds of warehouse tasks, like picking, relocations, replenishments, putaways, stock checks… the list goes on. You choose what you want your people to do and how you want them to do it.
Yes, there might be changes, because warehouses are always going to warehouse.
But those changes will get recorded. And that adjustment to how things get done, just on its own, will start to close the gap where your best picker (and anyone who might have been copying them) might wander off into their own private routing model.
Secondly, that directed route needs to be the right one to follow. Because, let’s face it, a system directed path, based on a configuration that nobody’s reviewed since the last order profile was updated, is just a more efficient manner of sending people the wrong way.
Which is where Optioryx Pulse comes in.
Pulse is an AI-powered picking optimisation engine that works directly on Dispatcher WMS’s move tasks. It analyses pick walks in real time. It clusters orders, it minimises travel, and it keeps adapting as your order profiles and stock positions change. Most importantly, it doesn’t require a dozen years of institutional knowledge in order to be able to do any of this.
Instead, what it offers is the ability to give every picker on your team access to the kind of routing intelligence that currently only exists in the heads of your best people. Not an approximation of it, and not a static version of it, either. A live, continuously updated version, in real time, that keeps on getting better as it learns more about your operation.
It’s also an engine with numbers to its name: Bleckmann cut picker walking distance by 20% with Pulse, and CEVA lifted picking productivity by 15 to 20%.
The best bit is that using Dispatcher WMS and Pulse together means that you won’t have to wait the however many years it might have taken for everyone else to catch up with your top people.
Bonus.
Want to see it in action? Book an obligation-free discovery call now.
FAQ: Picking optimisation
Working out the smartest sequence and shortest route for every pick walk, rather than sending pickers around the warehouse in whatever order the original configuration dictated. Done well, it means that your picker covers less ground and gets more done, without having to rush or by cutting corners.
Partly training, partly attitude and partly experience. But also, partly something that no-one really talks about enough: the best pickers have built a better routing model in their heads than the one your WMS is handing out. And that makes much of the performance gap a systems problem rather than simply a people problem, which means that it’s fixable at the system level.
The system decides the route and the sequence, the picker follows it, and the completed work gets recorded against it accurately. Dispatcher WMS can be configured to direct putaway and picking, and to interleave tasks like relocations, replenishments and stock checks along the way.
Pulse is an AI-powered picking optimisation engine that works directly on Dispatcher WMS’s move tasks. It analyses pick walks in real time, clusters orders, minimises travel, and keeps adapting as order profiles and stock positions change. In deployments so far, Bleckmann cut walking distance by 20% and CEVA lifted picking productivity by 15 to 20%.
Request a Demo Now
If you’ve enjoyed this blog, get your free subscription to our LinkedIn weekly Newsletter
– The World of WMS –
for more of the same great information!
