Daniel sent us this one — he's building out a hardware collection for electronics repair and general DIY, everything from M4 screws down to the stuff you need to open a phone, plus masonry anchors and washers and drill bits. The problem is that even when his inventory system tells him which box something is in, that box might have hundreds of undifferentiated fasteners inside it, and finding the right one still takes forever. He's seen the plastic organizers at hardware stores, but he wants the system that serious DIY people actually use. Type-first, category-first, something else? And separately, what's the right size for a daily toolbox that holds just the essentials, assuming the deep inventory is the general store?
This is the kind of problem that seems trivial until you've spent forty-five minutes looking for one M2 point five by eight millimeter screw and your project is dead on the bench and your tea is cold and you're questioning every life choice that brought you here.
The paradox of the organized hoarder. You know exactly which box it's in. You just can't find it inside the box.
That's the bottleneck. Most people think the problem is storage — I need more bins, more drawers, more boxes. But the problem is the second tier of organization. Tier one is knowing which container. Tier two is finding the part inside that container in under thirty seconds. Hardware store organizers fail at tier two because they give you twenty-four identical bins and a label maker and call it a day. But when you've got fifty variants of M3 machine screws, twenty-four bins doesn't cut it, and even if it did, the bins are the wrong shape for half the things you're storing.
I want to sit with that thirty-second number for a moment, because I think it's more important than it sounds. Why thirty seconds specifically? Why not a minute? Why not ten seconds?
Because thirty seconds is the threshold where your working memory starts to decay. You're holding the mental model of your project — the screw goes here, the standoff goes there, the washer goes on top — and every second you spend searching is a second that model erodes. By forty-five seconds, you've forgotten whether the washer goes above or below the bracket, and now you're re-reading the instructions or re-examining the part you just removed. By two minutes, you've pulled out your phone to check a reference photo, and the project has lost all momentum. Thirty seconds keeps you inside the cognitive flow state. That's not a number I made up — it comes from interruption studies in software development, where the cost of a context switch is measured in the time it takes to reacquire the mental model. Hardware assembly has the same problem. The screw isn't just a screw. It's a specific location in a three-dimensional assembly that you're holding in your head.
The retrieval time isn't about convenience. It's about not losing the thread of the project.
And that's why the hardware store organizer fails. It solves tier one — you know the screw is in bin fourteen — but tier two is a rummage through a bin of fifty near-identical fasteners, and that rummage can easily take ninety seconds. You've lost the thread. The project is now fighting you on two fronts.
Let's unpack this. We've got three competing ways to slice the taxonomy. Option one — type-first. All M4 screws together, all number eight wood screws together, all drywall anchors together. Option two — function-first, which is basically organizing by project or by application. Everything for drywall in one box, everything for laptop repair in another. Option three — geometry-first. All pan-head screws together, all flat-heads, all hex-heads, regardless of thread size.
I want to address option three immediately, because it sounds clever and it's actually terrible for retrieval. I've seen people do this — they have a drawer of pan-head screws and a drawer of flat-heads and a drawer of socket-heads. The problem is, when you need an M3 by ten millimeter pan-head, you go to the pan-head drawer, and now you're sorting through M2 and M4 and M5 and number six and number eight, all mixed together, all pan-head, all looking nearly identical from the top. You've just moved the sorting problem from one drawer to another. Geometry-first is the musical equivalent of organizing your record collection by album spine color. It looks great, it's completely useless for finding anything.
The album spine analogy is perfect, but I want to push on it a little, because someone listening is going to say, "Wait, I actually do organize my books by color and it works fine for me." What's the difference?
The difference is that books have spines with titles printed on them. You can scan a shelf of color-organized books and read the titles. Fasteners don't have titles. An M3 by ten and an M4 by ten look identical from six inches away. You have to pick them up, maybe measure them, maybe test-thread them into something. A book organized by color still has its metadata visible. A screw organized by head geometry has hidden its metadata. That's the distinction. Organize by the attribute you need to retrieve by. When you're building something, you need a specific thread size and length. You almost never need "a pan-head screw of any size whatsoever." So organizing by head geometry is organizing by the answer to a question nobody asks.
We can kill geometry-first right there. The real fight is type versus function.
For the deep inventory, type-first is the correct answer, and I'll tell you why. Function-first — organizing by project — creates duplicate inventory. If you have a drywall box with number eight drywall screws and a shed repair box with number eight drywall screws and a bathroom remodel box with number eight drywall screws, you now own three separate stashes of the same screw, you don't know how many you actually have, and when you run out in one box you'll raid another box and now your system is broken. Function-first also wastes space because each project box needs its own container infrastructure, and projects change. The bathroom remodel is done, but the box lives on, half-full, taking up a drawer.
There's a retrieval cost either way. With type-first, when you're doing a specific project, you visit multiple drawers — one for the M4 screws, one for the washers, one for the anchors. With function-first, you visit one box. But function-first falls apart the moment a part gets used across projects, which is almost all of them. The M4 by twelve millimeter socket-head screw doesn't care whether it's going into a 3D printer or a drone frame.
So the deep inventory should be type-first. But we need to talk about what type-first actually means, because there's a hierarchy within type, and getting that hierarchy right is where most people mess up. The top level should be the thread standard — metric or imperial. These should never share a drawer. The cost of cross-threading an M4 nut onto a number eight screw because you grabbed the wrong bin is a stripped thread and a ruined part. After that, it's thread diameter — all M3 together, all M4 together, all number six together. Then within that, it's length — six millimeter, eight millimeter, ten millimeter. Then within length, it's head type — pan, flat, button, socket. Then within head type, it's material — steel, stainless, brass, nylon.
That's five levels of nesting. Most people give up after two.
That's why most people have a box of hundreds. But here's the thing — you don't need a separate physical bin for every leaf node in that tree. What you need is nested sub-containers. Let me walk through the M3 drawer problem, because it's the perfect example.
This is my favorite kind of problem. The one that sounds like a joke but is actually the entire discipline in miniature.
A serious electronics repair person might have fifty different M3 screws. Different lengths — six, eight, ten, twelve, sixteen, twenty millimeters. Different head types — pan, flat, button, hex socket, countersunk. Different materials — steel, stainless, brass, nylon. If you put all fifty variants loose in one drawer, congratulations, you've built a haystack and you're looking for a needle. If you give each variant its own bin in a plastic organizer, you need fifty bins, and a standard twenty-four bin organizer from Home Depot gives you bins that are two inches by two inches by one and a half inches. That's fine for M3 by six millimeter screws — you can fit maybe three hundred in one bin. But it's also way more volume than you need for the nylon M3 by sixteen millimeter standoffs you own twelve of. You're using a whole bin for twelve parts.
The bin density problem.
And it gets worse in the other direction. That same two-by-two-by-one-point-five bin holds maybe forty number eight drywall anchors because they're physically bulky. So you've got bins that are half-empty because the parts are tiny, and bins that are overflowing because the parts are large, and they're all the same size bin. The hardware store organizer assumes all fasteners are roughly the same volume. They are absolutely not.
This reminds me of the problem with fixed-size apartment mailboxes. Every unit gets the same slot, whether you get two letters a week or you're running a small business out of your apartment and getting padded envelopes every day. The system breaks at the edges.
And the apartment mailbox has the same false assumption — that all mail is a number-ten envelope. Fastener storage has the same false assumption — that all fasteners are a one-inch drywall screw. They're not. You've got M2 washers that are smaller than a grain of rice, and you've got quarter-inch by four-inch lag bolts. They cannot share the same bin geometry.
What's the fix?
One drawer for M3. That drawer contains small sub-containers — and I mean small. Two-by-two inch ziplock bags, thirty-five millimeter film canisters if you can still find them, or my personal favorite, the little plastic vials that COVID test kits came in. Each sub-container holds one variant. M3 by eight millimeter pan-head steel in one bag, M3 by eight millimeter pan-head stainless in another, M3 by eight millimeter flat-head steel in a third. The bags are labeled, and they all live in the same drawer. When you need an M3 screw, you open the M3 drawer, you scan maybe fifteen bags, and you find the right one in five seconds.
The bags conform to the volume of the parts, so you're not wasting space on empty bin volume.
The bag is only as big as the parts inside it. This is the Sortimo principle, by the way. Sortimo is the company that makes professional in-vehicle storage systems for field service technicians. Their core insight is that the drawer is the unit of organization, not the bin. Each drawer has a fixed depth — sixty millimeters, ninety millimeters, or a hundred and twenty millimeters — and inside the drawer, you have a grid of one-inch by one-inch dividers that can be reconfigured without tools. So you can have a three-by-three inch section for your M4 nuts and a one-by-six inch section for your long standoffs, all in the same drawer. The bins are variable within the fixed footprint.
A Sortimo drawer costs more than my first bicycle.
A single Sortimo drawer costs about what a whole stack of Plano boxes costs, yes. But the system is expandable and the retrieval time is dramatically lower. For a serious DIYer, I recommend the sixty-drawer cabinet. Sortimo calls their system L-Boxx, and a sixty-drawer configuration runs about two hundred fifty dollars retail. Husky makes a comparable system for less. The drawers are modular — you can pull one out and take it to your workbench, do the job, and slot it back in. That alone is worth the price. You're not walking back and forth to the storage wall six times because you grabbed an M4 by ten when you needed an M4 by twelve.
I want to pause on that walk-back-and-forth cost, because I think it's invisible to most people setting up a workshop for the first time. You walk six times, that's maybe two minutes of transit. Doesn't sound like much. But multiply that across a year of projects.
Let's do the math. Say you do three projects a week — a mix of repairs, builds, and maintenance. Each project involves maybe eight trips to the deep inventory. If your storage is poorly organized and you make one wrong trip per project, that's three wrong trips a week. At two minutes per wrong trip, that's six minutes a week. Over a year, that's five hours of walking back and forth to grab the wrong screw. Five hours of your life, gone, because the M4 by ten and the M4 by twelve were in the same bin and you grabbed without looking closely enough. And that's just the transit time — it doesn't count the frustration, the re-threading, the stripped heads.
Five hours buys you a very nice dinner. Or in this case, it buys you the drawer system that eliminates the problem.
The drawer system lasts ten years. So you're trading five hours a year for a one-time purchase. That math is not complicated.
The layout for sixty drawers?
Drawers one through ten — metric machine screws, M2 through M5. Eleven through twenty — imperial machine screws, number two through quarter-inch. Twenty-one through thirty — wood screws and construction fasteners. Thirty-one through forty — anchors, washers, nuts, and threaded inserts. Forty-one through fifty — drill bits, taps, dies, and cutting tools. Fifty-one through sixty — adhesives, heat-shrink, solder, consumables. That's a layout that covers electronics repair, 3D printing, light construction, and general DIY without overlap. And you still have room to expand because the cabinets stack.
I want to pause on one thing before we move to the daily toolbox. We've got sixty drawers, each with sub-containers, each with a specific variant. How are we labeling this so we don't spend more time reading labels than finding parts?
Brother P-Touch labeler with twelve millimeter laminated TZe tape. That tape is rated for five-plus years of adhesion on ABS plastic and it's resistant to oils and solvents. You can spill acetone on it and it won't smudge. The label format should be consistent across the entire system. Thread size, length, head type, material. So: M3 space ten millimeter space pan head space stainless. That's it. Five seconds to read, unambiguous, and it sorts correctly when you scan visually because all the M3s are together and within M3 all the six millimeters are before the eight millimeters.
What about color coding?
It fails at scale. If you've got eight colors and fifty variants, you're doing math in your head every time you open a drawer. Is M3 blue or green? Is M4 the darker blue? Wait, no, blue is metric and green is imperial? Now you're playing a memory game while you're trying to fix a laptop. Laminated text labels are faster, more scalable, and they don't require you to remember a lookup table. The same goes for QR codes. I've seen people put QR codes on their bins that link to a parts database. It's clever engineering and terrible user experience. You have to pull out your phone, unlock it, open the camera, scan the code, wait for the page to load — that's a fifteen-second retrieval tax every single time. The cost of a touch in inventory management is estimated at fifty cents to two dollars per retrieval in industrial settings, based on time-motion studies from the automotive repair industry. At home, the cost is your attention and your patience. Don't spend it on QR codes.
I love the QR code thing because it's such a perfect example of solving the wrong problem. The engineer thinks, "The problem is that I don't have perfect metadata at my fingertips." The actual problem is, "I need the screw in my hand in under thirty seconds." A QR code gives you perfect metadata and a fifteen-second penalty. A text label gives you sufficient metadata and a two-second scan. The engineer optimized the wrong variable.
That's the trap with inventory systems in general. People build a beautiful database with photos and datasheet links and supplier part numbers, and then they spend more time maintaining the database than they spend using the parts. The database becomes the hobby. Which is fine if your hobby is databases. But Daniel's hobby is electronics repair. The inventory system should be transparent. You shouldn't notice it.
The inventory system itself — Daniel mentioned he has one that points to a box. With this setup, it should point to the drawer, not the bin.
The drawer label is the index. The bin label is the detail. If you're using a spreadsheet or an app like PartKeepr, the search should return drawer fourteen, bin C three — not box four. The physical system and the digital system need to use the same hierarchy. Otherwise you're translating between two different mental maps every time you look for a part, and that's where errors creep in.
We've got the deep inventory. Type-first, sixty drawers, nested sub-containers, consistent text labels, digital index at the drawer level. Now the daily toolbox. Because nobody wants to walk to the deep inventory wall every time they need a number eight drywall screw.
The daily toolbox is a completely different system with completely different logic. The deep inventory is organized by type because retrieval is the primary operation. The daily box is organized by task because speed and portability are the primary operations. You grab one pouch and you have everything for that task. You don't want to grab the M2 point five pouch and the Torx bit pouch and the spudger pouch and the magnetic mat — you want the laptop repair pouch that has all of those things in it.
Task-specific pouches inside a small case. Like a chef's knife roll but for fasteners and bits.
Let me give you two concrete examples. The laptop repair pouch contains M2 point five by six millimeter screws, a spudger, a Torx T5 bit, a Phillips zero bit, and a small magnetic mat. That's maybe twelve items total. You grab that pouch and you can open and service ninety percent of laptops. The drywall patch pouch contains number eight drywall screws, a few drywall anchors, a utility knife blade, a six-inch putty knife, and a small tube of spackle. That's it. Each pouch is a self-contained task kit.
The pouches live in what?
A twelve-by-eight-by-four inch tool roll or a small Pelican case — the ten-fifty or eleven-twenty. The constraint is that it has to fit in a backpack side pocket, because if it doesn't fit in your backpack, you won't take it with you, and a daily box you don't carry isn't a daily box. Inside, use zippered mesh pouches — Maxpedition makes good ones, but honestly the no-name ones on Amazon are fine for this. Each pouch is labeled with the task name. Laptop repair, drywall patch, electrical outlet replacement, picture hanging, 3D printer maintenance. Whatever your common tasks are.
How many fastener types should the daily box contain before it's just a second deep inventory?
That's the ceiling. If you have more than twenty unique fastener types in your daily box, you're not curating, you're hoarding. The daily box is for the things you use every week. Everything else lives in the deep inventory. And here's the discipline — when you use the last of something in the daily box, you restock it from the deep inventory. The deep inventory is the general store, the daily box is the pantry. You don't keep a fifty-pound bag of flour in the pantry.
This is where the psychology gets interesting. The bigger the daily box, the more you put in it, and the more you put in it, the slower retrieval gets, and pretty soon you're back to the box of hundreds problem but in a smaller box. The constraint is the feature.
A larger toolbox encourages hoarding. I've seen people with rolling tool chests that weigh three hundred pounds and they use maybe twenty percent of what's in there. The daily box should be small enough that you have to make decisions about what goes in it. If everything fits, you'll put everything in, and you'll be right back where you started.
I want to dig into that twenty-percent number, because I think it's one of those statistics that sounds made up but is actually observable everywhere once you start looking. What's the mechanism there? Why do people fill toolboxes with things they never use?
The first is aspirational stocking. You buy the tap and die set because you might need to rethread something someday. That day hasn't come in four years, but the set is in the drawer, taking up space, adding weight, and adding visual noise that makes it harder to find the things you actually need. The second is the sunk-cost retention problem. You bought a specialty bit for a one-time project — say, a triangular driver for opening a specific appliance. The project is done. The bit cost twelve dollars. You can't bring yourself to throw it away because you paid money for it, so it lives in the toolbox forever, even though the probability of needing it again is near zero. Between those two mechanisms, the average toolbox accumulates about five percent dead weight per year. After four years, you're at twenty percent bloat. After ten years, you're at fifty percent. That's not a rigorous study, but it matches what I've seen in every workshop I've ever visited.
The small daily box isn't just about portability. It's about forcing a regular curation cycle. If the box is full and you want to add something new, something old has to leave.
The constraint enforces the discipline. It's the same reason a closet organizer works — you can only keep what fits on the hangers you have. The physical limit does the decision-making for you.
To summarize the two-tier system. Deep inventory — sixty-drawer cabinet organized type-first with nested sub-containers. Daily box — twelve-inch tool roll or small Pelican case organized task-first with zippered pouches, maximum twenty fastener types.
The total cost to set this up from scratch. About two hundred fifty dollars for the sixty-drawer cabinet, forty dollars for a good tool roll, thirty dollars for a set of zippered pouches, and maybe thirty dollars for a Brother P-Touch labeler and tape. So around three hundred fifty dollars total. That's cheaper than buying ten plastic organizers from Home Depot that don't fit together, don't stack, and leave you with the same bin density problems we've been talking about.
One thing we haven't addressed — the orphan parts. The one-off screws that come with flat-pack furniture, the spare parts from appliances, the random standoffs that ship with a Raspberry Pi case. What do you do with those?
Call it the orphan drawer. No organization, no sub-containers, no labels beyond the word orphan on the front. If you need a screw and you think it might be in the orphan drawer, you search manually. If you haven't used something from the orphan drawer in six months, throw it out. The cost of keeping it is higher than the cost of buying a replacement if you ever need it, which you probably won't.
The retrieval tax on an orphan drawer is high, but the alternative — trying to catalog every random screw that enters your house — is infinite overhead for near-zero benefit.
That's the real principle here. Organization has a cost. Every bin, every label, every sub-container — you pay a tax every time you touch it, and you pay a tax every time you maintain it. The goal isn't to organize everything. The goal is to organize the things you actually use so that the retrieval cost is lower than the maintenance cost. For the things you might use someday, the retrieval cost can be high because you're almost never paying it.
This is the part where we tell people what to do this weekend.
Audit your current small-parts situation. Dump everything onto a table. Count how many unique fastener types you have. If you have more than twenty in a single container, you need sub-containers. If you have more than two hundred total, you need a drawer system. If you can't find a specific part in under thirty seconds, your taxonomy is wrong. And if you have parts you haven't touched in a year and they're not in an orphan drawer, either use them or lose them.
I want to add one step to that weekend audit, because I've seen people do the dump-everything-on-a-table step and then get overwhelmed and just scoop it all back into the box. What's the intermediate step that prevents the scoop-of-shame?
Sort into three piles before you even think about containers. Pile one: I used this in the last month. Pile two: I used this in the last year. Pile three: I don't remember using this, ever. Pile one goes into the daily box or the most accessible drawers. Pile two goes into the deep inventory, properly labeled. Pile three goes into the orphan drawer, and you set a calendar reminder for six months from now. When that reminder fires, whatever's still in the orphan drawer gets thrown out, no exceptions, no "but what if I need it." You won't. And if you do, you'll spend three dollars at the hardware store and learn nothing, because the alternative was keeping a thousand random fasteners for a decade on the off chance one of them saved you a three-dollar trip.
The three-pile sort is the emotional detachment mechanism. You're not deciding what to throw away. You're deciding what you actually use. The discards are just the remainder.
That's the psychology that makes it work. People resist decluttering because it feels like loss. But you're not losing anything — you're identifying what's actually in rotation and giving it a proper home. The stuff that isn't in rotation isn't inventory. It's clutter that happens to be shaped like fasteners.
One future implication before we wrap. 3D printing is changing the economics of this whole problem. If you can print a custom standoff for a Raspberry Pi project in fifteen minutes, you might not need to stock fifty variants of M3. You stock the common ones and print the weird ones. That's going to shrink the deep inventory for a lot of people. But that's a whole other episode.
I suspect we'll get to that one. For now, if you're building out a hardware collection, start with the sixty-drawer cabinet and the tool roll. Buy once, cry once. The frustration you save is worth more than the money you spend.
Now: Hilbert's daily fun fact.
Hilbert: In the seventeen-eighties, crofters in the Outer Hebrides cultivated a landrace barley called bere, which produced roughly three hundred kilograms of grain per hectare — about one-third the yield of modern barley varieties grown in the same soil today.
I genuinely don't know what to do with that information.
If you found this useful, please rate and review the show. It helps other serious DIYers find us. And if you have a weird prompt of your own, send it to prompts at myweirdprompts dot com. This has been My Weird Prompts. I'm Herman Poppleberry.
I'm Corn. Go organize something.
