Home/ScanTrack Systems
ScanTrack Systems · our own software, from 2008Our own hosted stock system, written in 2008 and run for customers counting, receiving and picking every day. Two businesses are still using it. Nothing has been added to it since 2018, and this page says why plainly rather than quietly retiring the name.
If you are here because you used to log into it, or still do, the short answer is at the bottom: the counting half has been rewritten for Android and is available now, and the rest is being written.
Ordinary now. Not in 2008, when the alternative for a mid-sized warehouse was a spreadsheet and a clipboard, or a six-figure module bolted onto an ERP.
Symbol and Motorola MC9100 and MC9200 guns, the same machines we sold, fitted and repaired. Scan in, scan out, order picking, and a stocktake when it came round. The device held the work and sent it up.
Stock visibility without walking to a shelf, on a dedicated server in North Ryde that ran from about $800 to $1,000 a month to keep going. Customers paid a monthly subscription and used it daily, which is a different thing from software that gets opened at audit time.
The peak was before 2017. A good number of businesses on monthly subscriptions, using it as the system of record for what was on their shelves. It is the reason this business can say it has been counting other people's stock for decades and mean the software as well as the scanners.
Two names, and it is worth being straight about which was which. ScanTrack Systems was the name over all of it — everything this business did as hosted software from 2008 — and it was a real company, registered as ScanTrack Systems Pty Ltd. Not one invoice ever went out under it. Every one came from Barcode Datalink Pty Ltd, trading since 1991, because somebody handing their stock records to a hosted system would rather deal with a business that had already been answering the phone for seventeen years.
And Stocktakes Online was not the thing that came after. It was the stocktaking product inside ScanTrack Systems, with its own name and its own logo, and every subscription included it. The invoices settle it: a line reading a line reading “Stocktakes Online” went out in October 2010. What you can buy today is that product rebuilt, not a new idea wearing an old name.
ScanTrack sold as a stock system, and twice it became the thing somebody else's application was written on top of. Neither of those was a stocktake.
Every client had their own address and their own database. mycompany.scantracksystems.com, and behind it a database holding nothing but their stock. The portal was written to be multi-tenant from the start — and between 2008 and 2024 it never once ran that way. Every customer stayed on a database of their own.
That reads like a design nobody used. It bought two real things. If a customer wanted their data we zipped the database and emailed it — not an export, not a report with the columns we thought they wanted, the whole thing. And one client’s app and portal could be changed to suit them without any risk to anybody else, which is how a stock system ended up running a bakery and a set of ticket gates.
And in 2026 it finally runs that way. The portal behind the new software is multi-tenant — one system with every company on it, rather than a database each. That is what the 2008 design was drawn for and never once got to do. Eighteen years is a long wait for an idea to come good.
The talking was done in 2010 and the year went on building it — the handheld application and the web portal behind it, both written to suit how this business actually worked rather than how a stock system usually does.
Eight or nine stores, and a kitchen that had to decide overnight what to bake for each of them. Too much and it went in the bin; too little and the case was empty by lunchtime. We wrote them their own version on top of ScanTrack — an online product list, every store set up as a site, and a par level per item per store, because a store in one suburb sells a different number of éclairs from a store in another.
From January. At close, staff walked the display with an MC65, scanned or keyed each item and counted what was left. Chocolate éclairs, par level ten, three still in the case. The kitchen ran the report at 3am and made seven.
Scanning was only one of the functions on the device — receiving and wastage were on the same menu, because the same walk answers more than one question. Every store talked live over wifi to the same hosted system, and the owner and his chefs ran the reports from a browser. We supplied the Zebra printer and the label software with it, and the access points that made the wifi reach the back of each shop.
It did not stay at seven. The business kept opening stores over the years that followed, and each new one needed a handheld, so the fleet grew with it — eleven MC65s across the years, and by 2017 an MC67. That is the part a single install never shows: it was still being added to nearly a decade after it went live.
ScanTrack went from its first version to its second, and their application came with it. Four years in, with every store depending on it at close of trade, that is the kind of upgrade that has to be invisible.
December 2019 was the last month — nine years all told. The handhelds were bought outright and kept being bought as the business opened stores, along with a printer, access points and spare batteries. The software was a subscription billed monthly across every store together, and it was rebuilt once in those nine years without anybody stopping work. Labels, ribbons, repairs and onsite support ran alongside both.
The subscription is what made the handhelds worth owning. A scanner that is not talking to anything is a torch. What was being paid for monthly was nine years of par levels held per item per store, a report that ran at 3am and was right, and a system rebuilt underneath them without a day’s interruption.
It did not end because it stopped working. Their point-of-sale vendor started offering stock management inside the till system, and one system beats two. We would have made the same call in their chair.
And a third, which starts with a phone call rather than a sale.
A large manufacturer had a project to track production orders through the floor. It went somewhere else. Some months later the phone rang and we were asked whether we would take it on.
We wrote a new mobile application and modified ScanTrack around how they actually worked, rather than the other way round. That is what the first invoice is, and it is why it is the size it is.
A label was printed and then scanned, which sounds obvious until you look at what it needs: two printers out on the floor with wireless print servers, so labels are made where the work is instead of carried from an office. And long range 2D imagers, because on a production floor the thing you are reading is rarely at arm’s length.
Twelve more handhelds and twelve more cradles, in one order, twelve months after the first four. A pilot that works does not get extended, it gets repeated.
The software grew with it and the changes are the tell: orders exported as active or processed, a picker dispatch role for the people on the floor, and a date range filter on dispatch tracking for whoever was watching it from a desk. That is not a stock system. It is a production line with a barcode on it.
The monthly subscription grew with the fleet too, and tripled over the same stretch.
The subscription ended in 2018. They kept buying printers, ribbons and labels from us, and the last line on the account is a single charging cradle in June 2021 — three years after the software went away, for a device bought seven years earlier and evidently still working.
And a fourth, which is the one to read if you have an ERP. It is not a stock system story at all. It is the story of an application that never needed a fleet, because the fleet was already there.
A manufacturer bought seven gun terminals and an access point to make the wifi reach — one first to prove it, six more four months later, with the chargers and the three year warranties specified around them on the same day.
They were not bought for our software. They were running a terminal emulation session into Pronto — a green screen on a handheld, talking to the ERP that already ran the business. That is what a warehouse fleet is normally bought for.
Then for two years we sold them nothing but ribbons.
Every unit leaving the warehouse had to be marked so it could be traced if it was stolen. One operator, one bench. Each item out of its carton, marked in six places — some easy to reach, some deliberately not — and back in the box with a registration card, so whoever bought it could register it online afterwards.
No ERP does that. It is not receiving, it is not picking, and it is not stock. It is a job that exists for eighteen months and then becomes normal, and no ERP vendor is going to build it for one customer.
We wrote it as a ScanTrack application and put it on one of those seven scanners — because one operator was doing the marking, so one device was the whole requirement. It sat alongside the ERP session on the same handheld.
Scan the model. Scan the serial. Scan the tag. The portal held the link between them, with who was logged on and the date and time of every scan — and handed it out over an API to the customer’s own web developers, because the half a buyer sees was somebody else’s system and always should have been.
Fourteen months and the marking programme stopped — their decision, and we never found out why. The subscription did not stop. The month it ended, a different application started on the same fleet.
Goods in, this time. A receiving purchase order keyed in, then the tag on each coil scanned, then the location it was put away to. The portal showed management what had arrived that day, where every coil was, and the total weight sitting in stock. Scanning one out to a job consumed it and took the whole weight off.
That second application ran until 30 November 2020. Long enough that the scanners underneath it were replaced — the original CE5 guns gave way to a newer generation in 2015, including a long range one, and the application came across with them.
Ten years, two applications, and not one device bought for either of them.
This is the argument, and it is not a software argument. A warehouse that runs an ERP already has scanners in it, already paid for, already on the wifi, already in somebody’s hand for eight hours — whether that ERP is Pronto or anything else. The gap is never the hardware. The gap is the job the ERP was never going to cover — the one that is specific to how that business works, and too small for the vendor to care about.
Both of those applications were modifications of ScanTrack rather than new products. That is the only reason either was affordable, and it is why the second one could start the month the first one finished.
Most clients bought their hardware, usually from us, and used the system every working day. People who are in a thing daily do not need somebody standing behind them, so we did not manage them. We hosted it, we fixed it, and otherwise we stayed out of the way.
A stocktake is not a daily habit. Nobody should be expected to remember in November what they worked out last November, so for the rental clients we did it: deliver the scanners and the access point, install, test, stay while it started, come back for the kit afterwards.
One of them did that with us every year from 2010 to 2024 — fifteen counts, eight scanners and an access point hired each time, and the software at the same price for thirteen of those years without an increase. What hiring the hardware actually involves.
Every subscription invoice was numbered as it went out, so these are not estimates. They are what was written on the invoice at the time.
The hardware went out in June 2014 on one order — a handheld, its cradle, a desktop label printer, a licence for the label software and a couple of rolls of ribbon. Nothing else has been bought since.
The monthly subscription started eleven months later, in May 2015, and has been billed every month since without a gap. Eleven years and four months, and the invoice for this month has gone out like all the others.
A different business, starting 2010. The final invoice went out in September 2024 and says so on it — month 133, and then nothing.
Nearly half of what that account was worth over fourteen years was the subscription rather than the hardware. That is what a system looks like when it is used rather than installed.
This one arrived through a supplier who resold it, billed once a year in advance, and it ran that way from 2014 to 2022. Then the reselling stopped and nobody told anybody.
Two years of it went unbilled. In August 2024 we invoiced the customer directly for both of them, gave them the fortnight in between for nothing to line the dates up, and moved them to monthly. They did not negotiate. They have paid every month since.
Nobody plans for a hundred and forty-seven months. The software was written in 2008, frozen in 2018, and it has been quietly running a business every working day of the eight years since. That is not a boast about the code — plenty of it would be written differently now. It is the only argument that matters about a stock system, which is whether it is still there in the morning.
Eleven years on one account, and in that time everything underneath it was replaced — twice.
It started with a single handheld and a cradle, and a version of ScanTrack written around how a freight operation actually books things in.
The reports were built and billed as six monthly instalments — the invoices literally read set of reports, 1 of 6 through to 6 of 6. Nobody paid for six months of work before seeing any of it, and nobody had to argue about scope at the end. We have not done that often enough since.
Four years in, one order replaced the lot: sixteen Windows Mobile 6.5 gun terminals with 2D imagers, 53 key. The application did not change. It had been written for the job rather than for the device, so the device could be swapped underneath it.
From March to September 2020 they hired extra scanners on top of the fleet they owned — four and five week blocks, back to back, month after month.
Freight in 2020 did not slow down. This is what that looks like on an invoice, and it is the clearest example on this site of why hiring exists: nobody buys a fleet for a surge they hope is temporary.
The last years brought Android touch computers and a wearable ring scanner onto the same system — hands free picking, which is about as far from a 2008 Windows Mobile gun as this hardware has travelled.
The account ran to 2025. Eleven years, three generations of hardware, and one system that never had to be replaced to keep up with any of them.
The same platform, a completely different job. Ticket numbers loaded into ScanTrack, and a fleet of MC65s turned into gate scanners — none of it sold. Every device, cradle, access point and laptop went out on hire.
Scan a ticket. Green if it was valid and had not been scanned. Red if it was not valid at all. And the one that mattered most — a warning if the ticket had already been scanned in, saying which scanner had done it, and when.
That third state is the whole product. A gate can cope with a bad ticket. What it cannot cope with is not knowing whether it has already let somebody in on that number.
A festival on Sydney Harbour in November 2010, six devices. A New Year's Eve event on a Sydney beach with tens of thousands waiting to get in, eighteen. A football match at a ground the following February, four — MC9090s that time, big and bulky to hold all day and perfectly fine for a few hours at a gate. A national swimming event in Adelaide that July, four again — and one of us on a plane, because the invoice still has the flights, the hotel and the airport taxi on it.
Laptops went with them running the same application, and the whole lot came home in a suitcase by freight. About nine months of work, all of it hire and onsite support rather than a sale — which is what a rental fleet is actually for.
The beach is the one worth telling. The technician on site from 5pm until 3am was the person who answers the phone here. The invoice bills ten hours and it was not a quiet ten: eighteen devices had never been tested together at that volume, because there is no way to rehearse tens of thousands of people arriving at once. Nothing about it felt settled until all eighteen could be heard beeping and working.
Tens of thousands of people queuing to get in on the last night of the year, and the answer coming back on some of them was that the ticket had been scanned a few minutes earlier. Those people turned round and left. They had believed a number could be used twice, and for the length of that queue we were the only reason it could not.
It is also the honest measure of what this kind of work is. Nobody notices a gate that works. It is only ever visible on the night it does not, and that is why somebody flies to Adelaide for four scanners.
Microsoft stopped supporting the operating system the handhelds ran on, and the devices themselves stopped being made. The right move was to rewrite for Android. We did not do it then. That was about eight years later than it should have been, and there is no version of this story where that was somebody else's fault.
There is a difference and it matters to anybody still using it. ScanTrack was not switched off, broken or abandoned mid-job. It was frozen: no new features, no new customers, and no pretending otherwise to the ones already on it.
That held for seven years. In 2026 development started again, and this time it is written in house rather than contracted out. The two websites came first — this one and Stocktakes Online, both rebuilt and live — and the portal behind the new software is hosted and managed by us.
Worth being exact about what that means: nothing new has been added to the 2008 ScanTrack and nothing will be. What is being built is a new one, and the old one keeps running underneath while that happens.
We still host it. It has changed data centre more than once since 2008 and the bill has moved with it, and the two businesses still on it are still invoiced every month. Software that old, still doing the job every day, is not a failure
Checked against the ledgers rather than remembered: one is on month 137 and the other on month 25, and both were invoiced this month. The month numbers are written on the invoices themselves.
— but it is also not something we would sell to somebody new, and we do not.This is the part worth reading carefully if you used ScanTrack, because what you can have today is narrower than what you had.
Stocktakes Online is the counting half, rebuilt for current Zebra Android handhelds instead of Windows Mobile ones — a wireless scanner and a portal, hosted in Sydney. It is sold per stocktake rather than per year, because most businesses count once or twice and should not pay for the other eleven months.
ScanTrack did these every day and Stocktakes Online does not. It counts, and that is all it claims. The rest is being built as a new ScanTrack — close to finished, and meant to do a range of mobile functions rather than only a count.
Counting is available today. The wider system is not, yet. Those are two different answers and they are worth not blurring. Stocktakes Online is finished and working, and can be bought or hired now on the terms set out on its own page. The new ScanTrack is close, and there will be more to say once it is done — no date is claimed here, because a date on an unfinished thing is a promise somebody else has to keep.
If you ran ScanTrack and want to know what your setup would look like now, that is a conversation rather than a quote. The counting we can do this month. For the rest, tell us what you used it for and you will hear first when it lands.
Counting, receiving, picking or all three — the answer decides whether there is something for you today or later this year, and we would rather tell you which.
