Why hard drives got expensive, and whether to buy one this quarter
Consumer drive prices are up roughly half since September 2025, and the usual advice to wait for the next price drop assumes a shortage that snaps back. This one is an allocation, not an accident: the drives are contracted to AI datacentres years ahead. What that changes about buying, waiting, and which files were ever worth the expensive shelf.
The storage forums have spent this summer writing one post in different words. Somebody reports what they just paid for a drive, phrased as a confession rather than a review. Somebody else asks how many people are deleting things to avoid buying one. The symptom gets described accurately and the cause almost never does, which matters, because the cause decides whether waiting is a strategy or just hoping.
The drives are not scarce. They are allocated.
The numbers are not in dispute. Consumer hard drive prices are up roughly 46 to 50 percent since September 2025. Trade coverage in February 2026 had a 4TB desktop drive that sold for under $80 in mid-2025 sitting near $100, and the second quarter of 2026 added about another 10 percent on top of that. Lead times on some capacities have stretched toward a year.
The reason changes what you should do about it. Western Digital has said its hard drive production is effectively sold out for calendar 2026, with long-term agreements running into 2027 and 2028, and Seagate's nearline capacity is similarly committed. The buyers are the companies building AI datacentres, and they do not shop. They contract, years ahead, for everything a factory can make.
Underneath that sits the fact the price threads miss. Cloud and nearline drives are around 87 to 89 percent of these manufacturers' drive business, and consumer retail is roughly five percent of Western Digital's revenue. A household shopping for one drive is not competing with other households. It is the residual claimant on a production line whose real customer signs multi-year purchase orders, and five percent of a business does not set its priorities.
That is why the usual advice does not transfer. The last shortage people remember, when flooding shut Thailand's drive factories, was destroyed supply, and prices fell back once the factories returned. This one is contracted demand. Nothing broke, so there is no repair to wait for, and analysts expect availability to stay tight through the end of 2026, with increases slowing rather than reversing. "Wait for the price drop" is a bet on a schedule nobody has published.
An assumption in everyone's storage plan just broke
Household storage advice written any time in the last fifteen years quietly assumes the price per terabyte falls, and that assumption is load-bearing. It is why "keep everything and buy bigger next time" was rational, and why extra copies felt close to free. Reverse the direction and every additional full copy of a library has a real price, payable now, at this quarter's rate. Which is why the same forums that traded capacity tips are asking what to stop keeping.
The useful move is not a stricter rule, it is a smaller unit of decision. Protection is worth choosing per collection, never as one policy over everything you own. Most of what sits on a home drive also exists somewhere else: media that could be re-acquired, files that are also in a cloud account, things that would be annoying rather than painful to lose. The set that exists only in your hands, the photos and scanned documents and recordings nobody else has, is usually a small fraction of the library, and that fraction is what deserves the expensive shelf. Costing out protection for everything at once is what makes the number look impossible.
Buy now, wait, or neither
One case for buying now is clear, and it has nothing to do with price forecasting. If something irreplaceable exists on exactly one drive, and that drive is old or has started reporting reallocated sectors in its SMART data, the risk is not financial. A drive that dies while you wait for a better deal costs more than the deal was worth. Buy the drive.
Otherwise, waiting is defensible and often correct. If what you keep fits on what you already own, this is a fine quarter to subtract rather than expand, and the hardware already in the house counts: the desktop that got replaced, the laptop in a drawer, the drive pulled from a dead machine. One caution on the salvage route, since the forums are full of it right now. A drive with unknown power-on hours is fine as a second copy and is not a plan for a first one.
The third route needs no drive at all. If you are paying for a cloud plan mainly because your files outgrew somebody's free tier, pulling a copy down onto storage you already own is the one move that gets cheaper as drives get more expensive, because that capacity was bought at last year's prices. It is also the only route here that lowers a bill rather than raising one.
Being straight about the obvious: Harbor Dock sells a box that needs a drive, so a post from us concluding "buy drives now" would be an advertisement wearing a chart. It is not the conclusion. For a household whose files are fine where they are, the honest move this quarter is to wait, reuse, or reclaim first, and to spend only on the thing genuinely at risk.
The honest tradeoffs
Waiting is not free. Prices may not come back to where they were, the lead times are real, and the thing you are deferring is a risk rather than a purchase. If the wait lasts a year, the irreplaceable set needs a second home before then, even a free-tier-shaped one. Cheap drives are also cheap for reasons: used and shucked drives carry unknown history and thinner warranty cover, and that discount narrows in a shortage. And deleting to save money is the one decision here you cannot reverse when prices settle.
Pulling files home has a limit as well. It changes who holds a copy, not how many places it exists, and if a cloud copy disappears when the paid plan lapses, that should be a decision rather than a discovery. The point was never to leave the clouds you like. It is to stop renting the copies you care about most.
Harbor is subject to the same market as everyone reading this. The dock does not include a drive, so its buyers pay this quarter's prices like anyone else. What the shortage changes is which argument matters. In a year of cheap storage, owning your own is mostly a preference. In a year when a terabyte costs this much, the capacity you already have is the cheapest you will see for a while, and the useful question stops being how much more to buy. It becomes which of your files were ever worth the expensive shelf.