Everything you own is somewhere. That is not the same as knowing where.
Asset tracking projects usually fail for the same reason: they answer where something is, when the business needed to know whether it is being used, whether it is due, and who had it last.
The pitch for asset tracking is always location. Put a tag on it, see it on a map. It demonstrates well, and it is the least useful part.
Ask what actually costs money and the answers are rarely about coordinates. A second bin hired because nobody could find the first one. A trailer that missed its certification because the reminder was attached to a person who left. A crew that drove to a site without the piece of equipment they needed, and lost the morning. Two of those three are not location problems at all.
The questions worth instrumenting
Before anyone buys a tag, it is worth being specific about which question is being answered, because the answer changes what the system has to be.
Where is it right now? Genuinely useful when items are mobile, valuable and frequently misplaced, and when someone will actually look at the screen. Much less useful for things that live in one place.
Is it being used, and how much? Utilisation is the question that changes purchasing decisions. Six items at forty percent utilisation is a different conversation from six items at ninety five, and only one of those justifies a seventh.
Who had it last? For consumables and hand tools this is usually the entire value. Not surveillance, accountability, and mostly it just needs to be easy to record.
Is it due for something? Service intervals, certifications, calibration, warrant of fitness. This is a scheduling problem attached to an asset register, and it is the one that carries real consequences when it slips.
Did it leave when it should have? Movement across a boundary, at a time nobody expected. This is the security case and it works best as an exception alert rather than a live map.
Most businesses want two of those. Systems get sold on the first one.
Why the technology choice comes second
The technology follows from the question, and the differences matter more than the brochures suggest.
Barcodes and QR codes are cheap, robust, and require a person to scan. That is a weakness if you wanted passive tracking and a strength if you wanted a deliberate record of a handover. Passive RFID reads in bulk at a gate or a doorway without anyone doing anything, and tells you an item passed a point rather than where it is now. Bluetooth tags and gateways give you rough zone level presence in a yard or a building at a per item cost that scales to hundreds of items. GPS trackers give real position outdoors and cost meaningfully more per unit and per month, which is fine for a trailer and absurd for a pallet jack.
The failure mode is picking the most capable option for everything. A yard full of GPS units is expensive, and half of them are on items that never leave.
The part that actually decides whether it works
Two things, and neither is technical.
The data has to land where decisions are made. A tracking system with its own login, its own map and its own report is a system people check when they already suspect a problem. The same data as a field on the asset in the ERP, or a line on the maintenance schedule, gets used without anyone deciding to use it. Every tracking deployment we would recommend includes the integration; it is not an optional phase two.
Somebody has to own the register. Tags fall off. Items get retired. New gear arrives and does not get tagged because the person who does the tagging was on leave. Six months of that and the system is wrong, and a system that is wrong is worse than no system, because people stop believing it and then quietly go back to walking around the yard.
That is the honest reason these projects underdeliver. Not the hardware. Nobody was given the job of keeping it true.
A reasonable way to start
Pick the one question that is currently costing you the most, instrument the smallest set of items that answers it, and connect it to a system people already look at. Run it for a season. See whether anyone acts on it.
If they do, extend it. If they do not, you have learned something cheap about whether the problem was ever visibility.