It's ten at night on a Sunday, you're one number eight wood screw short of finishing the shelf, and the hardware store closed forty-five minutes ago. That feeling. Daniel's been living it.
Oh, that's the worst. That exact screw. You can see it in your head — the one you definitely bought last month and definitely put somewhere. You remember the little plastic bag it came in. You remember thinking "I should put this somewhere I'll remember." And now you're standing in the garage at ten PM, opening every drawer, every jar, every coffee can, and it's just not there.
He's sent us a whole thing about building a deep inventory for home workshops, scaled down from the aircraft MRO systems we talked about last time. Two specs. First, a sixty-four drawer small-parts bin — the Akro-Mills kind, each drawer about ten by five centimeters. That's the dedicated-but-basic tier. Second, scaling up to roughly three hundred drawers for the serious DIYer who's also tinkering with computer hardware. He wants drawer allocations per category, realistic cost estimates, and then the part he says he hasn't found a good answer for anywhere — how do you actually procure all these fasteners in one go without spending a week clicking through Amazon listings or paying a fortune in per-unit markup?
That procurement question is the real one. Anyone can list what to buy. The how is where it falls apart. I've seen people spend three evenings just trying to source a complete set of metric machine screws in stainless, and by the end they've got twelve browser tabs open and a spreadsheet that looks like it was designed by a committee of angry raccoons.
So let's start with what a deep inventory system actually is, and why the same logic that keeps a seven forty-seven flying can keep your IKEA build from stalling at eleven PM on a Sunday.
In the MRO context, a deep inventory isn't just having parts. It's having the right distribution so the probability of a stockout during any maintenance event is near zero. The metric isn't part cost — it's flow preservation. If a thirty-cent O-ring grounds a two-hundred-million-dollar aircraft for six hours, the O-ring wasn't the cost. The six hours was.
And at home, the cost of a stockout isn't the fifteen-cent screw either. It's the forty-five-minute round trip to the hardware store, the lost momentum, the project that sits half-done for a week because you couldn't get back to it until Saturday. And the psychological cost is almost worse than the time cost. You were in the zone, you had your workflow, you were making progress — and then one missing fastener pulls the plug on the whole thing.
Same inventory math at a different scale. The two specs Daniel's proposing map to two different ambitions. The sixty-four drawer unit is "I want to stop running to the store for common stuff." The three-hundred-drawer setup is "I want to never have to think about whether I have a fastener." That's a qualitative jump — you're not just covering the top of the curve, you're pushing down into the long tail.
Walk me through the sixty-four. What actually goes in there?
Alright. The goal is coverage of the eighty-twenty curve — the twenty percent of fastener types that account for eighty percent of home DIY needs. So let's allocate drawer by drawer. Twelve drawers to wood screws. You want number six, number eight, and number ten in lengths from half an inch to two inches. Number six is your trim and small hardware screw. Number eight is the workhorse — cabinet screws, general construction. Number ten for heavier stuff. Flat head and pan head. Zinc-coated, mostly.
Twelve drawers feels like a lot for one category.
It's the category you reach for most. Think about how many times you've opened a drawer looking for a wood screw versus how many times you've needed a cotter pin. The ratio is probably fifty to one. Ten drawers to machine screws — M3, M4, M5, M6 on the metric side, and six thirty-two, eight thirty-two, ten twenty-four, ten thirty-two, and quarter-twenty on the imperial side. Common lengths, ten to thirty millimeters or three-eighths to one inch. Pan head and flat head. Eight drawers to drywall anchors and concrete screws — Tapcon-style and plastic expansion variety. Six drawers to nuts matching the machine screw sizes. Six drawers to flat and lock washers, same size range. Six drawers to hex-head bolts — quarter-inch, five-sixteenths, three-eighths in lengths up to three inches. Four drawers to self-tapping screws for sheet metal. Four drawers to specialty items — small hooks, eye screws, cup hooks. Four drawers to picture hangers and lighter-duty wall anchors. And four drawers to miscellaneous — cotter pins, hitch pins, small springs, the stuff you need twice a year but really need when you need it.
That's sixty-four exactly. And quantities per drawer?
Modest. Twenty to fifty pieces for screws, ten to twenty for bolts, thirty to fifty for washers. You're not stocking a hardware store. You're stocking enough that you don't run out mid-project. Refill when a drawer gets low, not when it's empty. There's a psychological trap here where people want to fill every drawer to capacity on day one, and that's how you end up with three hundred dollars in fasteners you'll never use. Start lean. The inventory grows organically as you do projects.
Cost on this?
The cabinet itself — Akro-Mills or similar — runs forty to sixty dollars. The fasteners, bought in bulk from a supplier like Bolt Depot or McMaster-Carr, roughly eighty to a hundred twenty dollars for reasonable quantities across all categories. Total, around a hundred fifty to two hundred dollars. That's the dedicated-but-basic tier. About what you'd spend on two or three decent power tools.
And that covers, what, most common household projects?
It covers the stuff that stops you dead. Hanging shelves, assembling furniture, fixing a loose hinge, mounting a TV bracket. The fasteners you know you need but somehow never have. Where it falls short is when you're doing something weird — repairing a laptop, working with aluminum extrusion, anything involving tiny metric threads or specialty heads.
Which is where the three hundred drawers come in.
So that's the sixty-four-drawer setup — a solid foundation. But what happens when you want to cover not just the common stuff but the weird stuff too? That's where the three-hundred-drawer spec comes in. And this is where the inventory philosophy shifts from "cover my bases" to "I am building a personal parts department."
And the configuration here is what — four of the sixty-four drawer units plus a forty-four drawer unit stacked in a metal frame?
That's the standard way to do it. Four sixty-fours and a forty-four gives you three hundred drawers exactly. Now we can get granular. Forty drawers to wood screws — we're adding number four for very small trim work, number twelve for lag alternatives, and three-inch lengths. Fifty drawers to machine screws. This is where we go deep — M1.6, M2, M2.5 for electronics, plus four-forty and two-fifty-six for small hardware. Then the standard M3 through M8 and quarter-twenty through three-eighths. Multiple lengths per diameter, multiple head types.
Fifty drawers for machine screws alone. That's nearly the entire sixty-four drawer unit just in one category.
And it's justified. Machine screws are the category where being one size off means the thread doesn't engage or it strips. You can't fake a machine screw the way you can sometimes use a slightly longer wood screw and just drive it deeper. The thread pitch is either right or it's wrong, and if it's wrong, you're done. Thirty drawers to bolts — hex head, but now also carriage bolts, lag bolts, shoulder bolts. Twenty-five drawers to nuts — nylock, wing nuts, cap nuts, jam nuts. Twenty drawers to washers — fender washers, split lock washers, finishing washers. Twenty-five drawers to drywall and concrete anchors — toggle bolts, molly bolts, sleeve anchors, the stuff rated for actual load.
You're at a hundred ninety drawers and we haven't touched computer hardware yet.
Right. Twenty drawers to self-tapping and sheet metal screws — including tek screws with drill points for metal framing. Then twenty drawers to computer hardware specifically. M1.6 and M2 are the standard for laptops and SSDs. M2.5 for NVMe drives. M3 for three-point-five-inch drives and case fans. Motherboard standoffs — those are number six thirty-two UNC in most ATX cases, but some use M3. Nylon standoffs for mounting PCBs without shorts. Jumper pins, drive screws, the tiny screws that hold a WiFi card in place.
The screw you lose when you open a laptop and it pings across the room into the carpet.
That screw. And you can't finish reassembling without it. I once spent forty minutes on my hands and knees with a flashlight looking for an M2 laptop screw that had ricocheted off the desk. Found it inside a shoe. Fifteen drawers to specialty items — threaded inserts, rivet nuts, well nuts, rubber grommets, cable clips. Fifteen drawers to adhesives and small tools — threadlocker, epoxy syringes, precision tweezers, hex keys in the sizes you always lose. Ten drawers to electrical — wire nuts, ring terminals, spade connectors, butt splices. Ten drawers to plumbing — O-rings, small washers for faucets, hose barbs. And ten drawers left empty for project overflow — whatever the current build needs that doesn't have a permanent home yet.
Three hundred. That's a parts department, not a parts drawer.
It's a serious setup. Cost-wise, the cabinets — four sixty-four drawer units plus a forty-four — run two fifty to three fifty. The fasteners in bulk, filling three hundred drawers with twenty to a hundred pieces each depending on size, roughly four hundred to six hundred dollars. Total, six fifty to nine fifty. That's the serious DIYer tier where you're buying from industrial suppliers, not the hardware store.
And the per-unit savings versus buying as you go?
Massive. A box of ten machine screws at Home Depot runs about a dollar fifty. The same screw bought in a bag of a hundred from Bolt Depot is maybe eight cents each. Over three hundred drawers, that difference pays for the cabinets several times over. Let me give you a concrete example. Say you're building a workbench and you need forty number-eight by one-and-a-quarter wood screws. At Home Depot, that's four boxes of ten at a dollar fifty each — six dollars. From Bolt Depot, a bag of a hundred is about four dollars, and you use less than half the bag. You just saved two dollars on one project, and you've got sixty screws left for the next one. That math repeats across every category.
Alright. We've talked about what to stock and how much it costs. But Daniel raised a question I don't think we've fully answered yet. How do you actually buy all this stuff without spending a week on Amazon?
This is the procurement problem, and it's the part he said he hasn't found a good answer for. And he's right — there isn't an obvious one. The reason MRO-style kits don't exist for home users is that the fastener industry is fragmented in about four different directions at once. Metric versus imperial. Zinc versus stainless versus brass versus nylon. Phillips versus hex versus Torx versus Robertson. Coarse thread versus fine thread. No single kit covers all those combinations, and no retailer wants to stock a kit that only covers ten percent of customers.
So the kit approach is dead.
The kit approach is dead. But the solution isn't buying one by one either. The solution is what I'd call a covering order — you use a supplier that lets you buy by the piece or by the bag, and you build a master parts list from a template. Several Redit communities — the workshop sub, the DIY sub, the tools sub — have shared spreadsheets of starter fastener inventories. You download one, adjust for what's available at your local suppliers, and place one bulk order.
And which supplier?
Bolt Depot and McMaster-Carr are the two that come up constantly. Bolt Depot lets you buy by the piece with no minimum — you can order exactly seventeen M4 twelve-millimeter pan heads if that's what you need. McMaster has per-bag pricing that's slightly higher per unit but their selection is unmatched and their shipping is next-day for most of the country. The key insight is to buy the most common sizes in bags of a hundred, not boxes of ten. The per-unit cost drops sixty to eighty percent, and you will use them eventually.
What about the assortment boxes at Home Depot or Lowe's — the Hillman and Everbilt ones?
Convenient but expensive. They're priced for the person who needs one screw right now, not for inventory building. Per-unit cost is two to three times what you'd pay from a distributor. A Hillman drawer with twenty-five M4 screws might be four dollars. A bag of a hundred from Bolt Depot is about the same price. You're paying for the packaging and the retail shelf space. The plastic box costs more to manufacture than the screws inside it.
So the hybrid approach is — identify the ten to fifteen most common sizes in each category, order those in bulk from a distributor, then fill in the oddballs from the local hardware store bins as needed.
That gets you eighty percent coverage in one order without the premium pricing. The ten to fifteen sizes thing is important — you don't need every length of every diameter. For wood screws, number eight by one-and-a-quarter, one-and-a-half, and two inches does most of the work. For machine screws, M4 by twelve, sixteen, and twenty millimeters covers a surprising amount. You stock the workhorses deep, and the weird stuff shallow.
Daniel mentioned he likes to keep the very small computer screws in separate storage. Makes sense — M1.6 and M2 are easy to lose in a drawer full of M6 bolts.
And they need different organization logic. Computer screws you want sorted by the device they go to, almost — laptop screws, drive screws, motherboard standoffs. Because when you're fixing a specific machine, you want all the fasteners for that machine together. It's a different retrieval pattern than "I need a three-inch deck screw." The deck screw drawer you open once a month. The laptop screw drawer you open once, but when you do, you need all of them at once, in one place, with no hunting.
Before we wrap up, Hilbert's been making a face for the last five minutes. Hilbert, you've got something to add?
Hilbert: You're both overcomplicating the procurement problem.
Alright.
Hilbert: The reason there's no MRO kit for home users isn't the fragmentation. It's that nobody wants to pay the shipping on fifty pounds of fasteners. A single covering order from McMaster-Carr for three hundred drawers of mixed fasteners weighs about forty pounds and costs eighty-five dollars in ground shipping. That's the hidden cost nobody talks about. The solution isn't a better kit. It's finding a local fastener distributor who'll let you walk in with a list and buy by the pound.
By the pound.
Hilbert: There's one in every mid-sized city. They charge half what McMaster does because they don't have to ship it. You walk in, hand them a list, they weigh it out. The place smells like grease and has bins of Grade Eight bolts in the back. I spent six months in twenty-nineteen as a parts runner for a custom motorcycle shop in Austin. We had a four-hundred-drawer McMaster-Carr wall — every fastener you could imagine, organized by thread pitch and head type. The owner had a rule. If you had to walk to the hardware store more than once in a week, you were fired. So I learned real fast what a deep inventory looks like when the cost of a stockout is your paycheck.
How do you find one of these places? What do you search for?
Hilbert: Industrial fastener supply plus your city name. If the results show a place with a loading dock and no website, you're in the right spot. If they have a retail counter with a guy named Mike who's been there thirty years and can identify a thread pitch by squinting at it, even better.
The by-the-pound thing is interesting because it changes the economics completely. If you're paying eight cents a screw from McMaster plus shipping, versus four cents a screw from a local distributor with no shipping, on three hundred drawers that's a couple hundred dollars difference.
Hilbert: And they'll sell you the weird stuff too. McMaster makes you buy a bag of fifty when you need three. Local place will hand you three and charge you thirty cents. They don't care. They're selling to contractors who buy ten thousand at a time.
What's the minimum order at a place like that?
Hilbert: There isn't one, usually. They might give you a look if you walk in for two screws. But if you show up with a spreadsheet and you're buying across thirty categories, they'll take you seriously. Bring coffee. Mike likes coffee.
The other thing Hilbert's point surfaces is that a lot of the procurement difficulty is really a shipping problem disguised as a selection problem. You can find everything you need online. You just can't get it to your door for a reasonable price unless you're ordering enough to amortize the freight.
Hilbert: Forty pounds of fasteners ships in a flat-rate box if you're smart about it. But most people aren't. They order from five different places and pay shipping five times.
So the covering order from one distributor is the move, but the distributor should be local if possible.
Hilbert: Local if you can. McMaster if you can't. Never Amazon for fasteners — half of it's mislabeled and the thread pitch is wrong.
And you know this from the motorcycle shop.
Hilbert: I know it from the three times I ordered what I thought were M6 one-point-zero and got M6 one-point-two-five and had to explain to the owner why the engine case wouldn't torque down. He was not sympathetic.
That's a hard lesson. Metric coarse versus fine thread — M6 coarse is one-point-zero pitch, fine is point-seven-five. They look identical until you try to thread them.
Hilbert: They don't look identical to Mike.
Of course they don't.
Hilbert: Anyway. Local distributor. By the pound. It's not complicated.
The thing I keep coming back to is — we've designed these two setups assuming the person building them knows what they need. But what about the beginner who doesn't know what an M2.5 screw looks like? Is there a starter kit approach that works for someone who hasn't developed fastener literacy yet?
That's the open question, isn't it. You can hand someone a three-hundred-drawer inventory list, but if they can't tell a machine screw from a wood screw by looking at it, they're going to put things in the wrong drawers and the system breaks down immediately. I'm thinking of my neighbor who tried to use drywall screws for a deck project because "they looked strong enough." That deck lasted about eight months before the screws started snapping.
The reference box idea I've mentioned before helps here. A small box with one labeled sample of each fastener type — the actual screw, the actual nut, with a label saying what it is and where it lives. You build that first, before you fill the drawers. It becomes your Rosetta Stone. When you're not sure what you're holding, you match it to the sample. It's like those paint chip cards at the hardware store, but for fasteners. You don't need to know the name, you just need to find the match.
And the other thing that's shifting under this whole conversation is three-printing and CNC. As those become more common in home workshops, the fastener inventory problem changes. Printed parts often use heat-set inserts and specialty screws that don't appear in traditional fastener assortments. M3 heat-set inserts, for example — they're not in any Hillman assortment at Home Depot, but they're essential if you're bolting into a three-printed part.
The deep inventory concept may need to evolve to include those new categories. We're already seeing hybrid workshops where you've got traditional wood screws next to M3 standoffs for printed enclosures. The three-hundred-drawer setup could absorb that — those ten overflow drawers I mentioned — but the procurement channels for heat-set inserts are different from the ones for zinc wood screws. You're not buying those from the industrial fastener place on the loading dock. You're buying them from electronics suppliers. So now you've got two supply chains to manage, not one.
The inventory system has to accommodate multiple supply chains. That's actually a harder problem than just having enough drawers.
It is. And it's where the spreadsheet approach breaks down, because you're managing reorder points across suppliers with different lead times and minimum orders. At some point you need actual inventory management software, not just a list. Or at least a more sophisticated system than "this drawer looks low."
Which is maybe where we loop back to the MRO analogy. Delta's not managing four hundred fifty thousand part numbers on a spreadsheet.
No, they're running full ERP systems with automated reorder triggers. But for a home workshop, the "drawer looks low, add it to the list" method works fine until you cross about five hundred part numbers. Below that, visual inspection beats software every time. The human eye is actually a pretty good inventory sensor at that scale. You can glance at a drawer and know instantly whether it's at twenty percent or eighty percent capacity. Software requires data entry, and data entry is where home inventory systems go to die.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop for keeping us honest about shipping costs and the existence of Mike.
If you want to share your own fastener inventory spreadsheet or argue with Hilbert about the best local supplier, leave us a voicemail at my weird prompts dot com. We'll read the best ones on a future episode.
We'll be back soon.
