The Invisible Man in Your Warehouse

H.G. Wells wrote about The Invisible Man long before we did. His story is about a man that no-one can see, doing stuff, getting results, but that no-one even knows is there. It’s great fiction, but it’s also, somewhat inconveniently, the reality of what’s playing out in most warehouses every day.

Because while a lot of those warehouses can tell you which of their pickers clocked in this morning at 6am sharp, far fewer of them know which one of those pickers has moved the most cases by the time you’re reading this article.

Possibly, and somewhat controversially, today, we’re going to be talking about the fact that not every metric is the right metric to pay attention to.

Hours are easy

Hours can be counted by using something simple, like Excel. Output is harder, so output gets estimated, averaged, or, well… skipped. Using a spreadsheet to keep track of who’s in the building means that you’re aware of punctuality and presence. And that’s great if you’re running a primary school, gold stars all round, but it’s somewhat suboptimal for a distribution centre.

We will be talking about how to easily get the right information out of your Dispatcher WMS later on in this article. But for now, just know that it’s there, being recorded, in real time, at transaction level, all of the time, and that you can get it out easily enough, if you know how.

But back to the point of the metrics (and this is where this kind of article usually goes wrong) because a lot of folks like to talk about how having output metrics lets you find the slackers in your crew. We’re not going to do that, because, firstly, and quite frankly, it’s a bit grim, but more importantly, it’s mostly not true.

Your strongest performers and the people who are having a harder week of it can look identical in the surface data, and even when you start to dig into it, you might want to hold off having any chats with your low performers until you understand why they’re working the way that they are.

So, this isn’t about squeezing people. It’s about knowing what’s going on well enough to actually help them get better results.

Because the BEST way to use these metrics isn’t necessarily the most obvious one

The most useful thing that you can use good output data for is to spot where the system is letting people down.

For example: someone’s pick rate is half what it was last month.

There are lots of reasons this might be happening, but the two most likely are: they’ve checked out for some reason, or something has changed in their work environment.

• Maybe the SKU mix shifted and now they’re walking three aisles more every time they do a pick.
• Maybe their scanner is timing out, but nobody’s logged a ticket (even though they’ve asked about it repeatedly).
• Maybe they’ve been moved to a zone that they weren’t trained on.
• Or… maybe they pulled their back at the weekend and haven’t said anything because they need the shift.

Your spreadsheet doesn’t know any of the above. And neither do you, until you’ve got data that’s fine-grained enough to tell you if it’s the route, or the person.

We’d suggest that this is the actual case for getting better output metrics. It’s a blunt instrument… if you just want to use it as a stick to beat people with, well, we suppose you could. But if you’re willing to use it instead as a diagnostic tool, you’ll end up making fewer demands of your people, not more, because you’ll be able to find and fix the things that have been making the work harder than it needed to be. And in this labour market, we’re convinced that THAT is worth thinking about.

Moving the needle

Luckily, there are tools that will help you both find the right metrics and improve them. If, for example, you discover that you need to optimise your picking, Optioryx Pulse sits directly on Dispatcher WMS’s move tasks, and it can rewrite your pick path on the fly.

Here’s a worked example to show you how that might improve your metrics. Let’s be conservative and say that it saves you twelve seconds per pick. Not much, perhaps something you’d currently consider to be rounding-error territory. But let’s do the maths before we dismiss it, shall we?

For an operation that does 4,000 picks a shift. Three shifts a day. 250 working days a year. That’s about 10,000 hours of picking time that you can recover annually.

• Picks per year: 4,000 × 3 × 250 = 3,000,000 picks
• Time saved: 3,000,000 × 12 seconds = 36,000,000 seconds
• Convert that to hours: 36,000,000 ÷ 3,600 = 10,000 hours

Which is roughly four full-time pickers’ worth of capacity that you’ve just uncovered, without hiring anyone new, without changing your floor layout, and without buying a single new piece of equipment.

The numbers will obviously be different on your site. But they’ll probably still be larger than you’d think. So perhaps it’s worth running on your own volumes before deciding that those twelve seconds aren’t worth bothering with? We’d be happy to help you do that, if you’re interested.

An important point: implementing Optioryx Pulse is a system change, NOT a people change. Pulse is about optimising your routes rather than your workers. Nobody is being asked to walk faster or skip their break. Instead, working on the floor starts to become less punishing, your numbers go up, and the people who are doing the actual work will most likely need to buy fewer pairs of shoes.

Identifying the numbers that really matter

So, Pulse moves your output, but the big problem is actually seeing it when you need to see it. So this next little bit is going to explain how you can get the exact metrics that you wanted to see in the first place.

Most ops directors are running their floor on yesterday’s data.

And that’s because the reporting cycle is slower than what’s happening on the floor. Typically, the pattern goes like this: floor manager has a question, raises a ticket, IT pulls the data, someone writes the report, and it lands back in their inbox the following afternoon. However, by the following afternoon, the issue has either resolved itself, or else, it’s compounded into a totally different one. Either way, that report is an artefact, not a useful tool.

AskUSP eliminates that gap

If, instead, you implement AskUSP, our natural language AI, your floor manager could ask, e.g. “how many picks were completed by each team yesterday morning?” in plain English (or whatever their language is) and rapidly get a clean answer back. The same question can be asked at 6.15am, 11.30am, and again at the end of the shift. And all without filing a single ticket or learning SQL. Plus, your IT team will get some of their week back, and that is, quite honestly, a win in itself.

The difference that AskUSP brings to the table is small to describe but significant to live with. You go from “I’ll find out by Wednesday” to “I’ll find out before lunch” and having that information at the tip of your fingers can change how your floor gets run. People stop saving questions up. They start spotting dips before they become a pattern. And you can catch the broken scanner on the day that it breaks, rather than the day that someone finally complains to the right person to get it fixed.

AskUSP, incidentally, is run from within our own Blue Yonder approved micro-services, User Services Portal (USP). It queries information directly from Dispatcher WMS transactions and delivers it via natural language.

Compounding effects

Output metrics work best when they’re reviewed on a continuous loop: Measure. Make changes. Measure again.

And when that loop runs every working day, the questions you’ll start to ask about it will inevitably get sharper.

• What’s the gap between your top performers and the rest?
• Is it the SKU mix, the slotting, the shift pattern, the training, the equipment?
• Where’s the recoverable time, and is it sitting in the route or is it stuck inside the way the work is being assigned?

The answers tend to be unglamorous. We’re not going to pretend otherwise.

It might be a poorly slotted aisle. Or a handoff between zones that’s wasting a minute every time that it happens. It might simply be a training gap that nobody’s had time to address. Nice, boring, and (typically) very fixable things. The sort of things that don’t often make it into a strategy meeting but will chew through hours without the right people noticing.

Fix enough of them and your floor will no longer feel as if it’s running on the goodwill of a few people. Your Super Workers (the people who’ve been compensating for whatever everyone else didn’t do last week, and, whatever ‘The Invisible Man’ might suggest, not all men) get some of their time and effort back, or at least it gets noticed and hopefully acknowledged. And the people who were having a hard time stop having a hard time, because the thing that was making it hard has been fixed.

The tech stack that you’ll need to see all of this

It hangs together like this: Dispatcher WMS runs your warehouse. Pulse moves the needle. AskUSP shows you what you need to know.

And then you get to do it all over again tomorrow.

The point of this isn’t to find out who’s slow. It’s to find out what’s slowing them down. Most of the answers are duller than you’d hope and bigger than you’d expect. Book an obligation-free discovery call and we’ll show you what’s actually happening in your warehouse.

H.G. Wells’s character spends the whole novel trying to be seen, and never quite manages it. We believe that your best pickers really shouldn’t have to suffer the same fate, what about you?