There's a peculiar habit in software demos. Everyone shows you the order that behaves itself. One customer, one warehouse, one price, one neat little box. It's the retail equivalent of a show home: lovely to walk around, and nobody actually lives there.

Real orders are far more interesting. A product that another warehouse fulfils, sitting on the same order as everything Helm picks and packs itself. A price that depends entirely on who's buying, not on what's printed on the label. These aren't edge cases in the way people think. In most businesses, they're Tuesday.

What struck me about this week's questions is that none of them were really about a feature. They were about a shape: the order that doesn't quite fit the model a platform was built around in the first place. It's a bit like designing a kitchen for someone who only ever makes toast, then discovering they host Christmas dinner for fourteen.

The good news is that most of these don't break the system. And where something genuinely can't be fully solved yet, we'd rather say so plainly than dress it up. Oddly, I've found that admitting what you can't do is one of the most reassuring things a vendor can say. It makes everything else more believable.

Because every vendor can demo the standard case. The real question is what happens the first time an order doesn't look like it.

So below are a few of this week's questions that I thought were worth sharing. Hopefully you'll find them as interesting as I did.

Can Oversized Products Trigger a Different Pick Method?

Yes: Helm holds size, weight and dimensions against every inventory item, and inventory categories or tags can trigger a different pick method, such as a two-tote pick, for anything oversized.

Automatic tote-size allocation purely from those volumetrics isn't built yet. For now, a two-tote pick has to be triggered by category or tag rather than Helm calculating it from the numbers directly, a sensible stopgap rather than a finished answer, and exactly the kind of gap the open API exists to let you close yourself rather than wait for a roadmap.

Does Helm Optimise Pick Routes or Rely on a Fixed Path?

Helm provides AI-optimised pick routes rather than a fixed A-to-Z path, minimising steps and backtracking across a pick.

A priority order can also be set to follow a specific physical route, such as an S-shape or U-shape down the aisles, so pick sheets list items in the most efficient walking order for that layout, rather

than whatever order they happen to sit in on the shelf. The point of optimising the route at all is the same one worth repeating from a few weeks back: picking speed is mostly a function of walking distance, and a fixed A-to-Z path treats every warehouse as if it were built in alphabetical order, which none of them are.

How Do You Stop Externally Fulfilled Items Creating Ghost Stock?

Yes: Helm lets you set an externally fulfilled SKU up as a "Service" type product rather than physical inventory, so it doesn't require stock, doesn't appear on a pick sheet, and won't block the rest of the order.

That lets an item fulfilled from a separate chilled or fresh warehouse pass through an order safely alongside everything Helm does pick and pack, without someone manually removing it from every order it appears on. The same trick is used elsewhere for parent or bundle SKUs that show up on an order but were never meant to be individually picked in the first place, which says something about how often the real problem is a product that shouldn't behave like stock in the first place, not a missing feature.

Can You Hold Your Own Packaging Stock in Helm?

Yes: packaging, such as printed boxes, can be set up in Helm to hold its own stock, deducted automatically at despatch, with low-stock warnings sent by email.

The alternative is the warehouse remembering to flag it, which works until the week nobody does, and a print run of branded boxes runs out mid-despatch with no warning until someone's already packing into a plain one. Packaging isn't inventory in the way products are, but it behaves exactly like inventory the moment you run out of it, so it might as well be tracked like it.

Is There a Mobile App for Warehouse Stock Counts?

Yes: staff search for the product or scan its barcode and enter the counted quantity; Helm updates immediately. The app currently works on Android and IOS version.

A count that updates immediately, rather than getting typed up later from a clipboard, is the difference between a stock discrepancy getting caught the same day and getting caught whenever someone finally gets round to transcribing a stack of paper counts.

Does Helm Cope with Prices That Vary by Customer or Account Type?

Yes: Helm's reporting on revenue and orders comes from the actual orders as they arrive, not a fixed list price, so the price held on the product import itself isn't critical.

It can be left blank or set to an average without breaking anything, since the number that actually matters is whatever price each order came in at, not whatever figure sits on a product record. Worth knowing before importing a catalogue and worrying over a single "correct" price for a product whose real price depends on who's buying it, not what's printed on a label.

Can Warehouse Staff Time on Ad-Hoc Services Be Tracked for Billing?

Yes: Helm supports time and task tracking for manual or ad-hoc services, such as kitting, specifically for billing purposes.

That distinction matters for anyone billing clients for warehouse labour rather than just shipped units: a kitting job or other ad-hoc service that eats into someone's time doesn't show up anywhere in a per-order or per-parcel billing model on its own, so without time tracked against the actual task, that work either goes unbilled or gets estimated after the fact from memory. Tracking it at the point the work happens is the difference between billing what actually took the time and guessing at it later.

Can Shopify Tags Be Used as Filters in Pick Rules?

Yes: Helm reads tags set in Shopify, and they can be used as filters or triggers inside pick rules.

A tag that already exists for merchandising or marketing reasons doesn't have to stay siloed in Shopify; the same label can decide how an order gets picked once it reaches the warehouse, rather than warehouse logic needing its own separate tagging system duplicating work already done upstream.

Does Order & Tracking Status Push Back to Shopify?

Yes: Helm pushes order and despatch information back to Shopify when an order is printed or dispatched, and tracking status updates feed Shopify's own tracking display too.

Helm's own statuses are more granular than what Shopify shows natively, which matters more than it sounds: Shopify's view is the simplified summary, and Helm's is the detail that summary was built from. A customer checking their own order sees the simple version; whoever's actually running the warehouse needs the fuller one, and the integration's job is making sure neither side has to choose which version to trust.

Does Click & Drop Stay a Separate System Once You're on Helm?

No: once a sales channel is connected to Helm and Click & Drop is linked via an API key, orders flow into Helm and labels get generated from within Helm directly, with no need to log into Click & Drop separately.

Click & Drop still exists underneath; it simply reflects the order as dispatched once Helm has booked it, rather than staying a second screen someone has to keep open all day. That's the difference between an integration and a login shortcut: whether the old tool quietly disappears from your actual workflow, or just gets a single sign-on bolted onto the front of it.

Ten questions, and not one of them was about the standard order going through the standard way; every single one was about the order, the product or the price that didn't fit the shape the system expects. That's not a weakness in the FAQ; it's the actual test of a platform: not whether the demo

order works (any platform can manage that), but whether the next edge case gets a workaround and a plan, rather than a shrug.

If your own operation has an edge case that doesn't fit neatly into anyone's demo, bring it to a call rather than assuming it isn't supported: talk to us.