Cost tracking is like weighing the dog. Once a month, useful. Three times a day, you're not learning anything new — you're just disturbing the dog.
The same is true of inventory detail. Somewhere between "I have no idea what anything costs" and "I have a six-digit allocation key tied to my refrigerator amperage," there's a level of detail that's right for your business. Most of the cost-tracking advice that gets handed down in maker forums quietly pushes you past that middle and into the territory where the tracking itself becomes the cost.
A boundary matters here, because tracking inventory is genuinely one of the highest-leverage habits a small maker business can build. This is not an argument against tracking. It's an argument against three specific habits — the kind that look rigorous, feel responsible, and quietly cost more hours per quarter than they ever recover in margin clarity.
The short version. Granularity is a tool, not a virtue. For most cottage-scale sellers, four numbers per SKU — materials, packaging, an estimate of make-time, and a monthly overhead line — capture eighty percent of the margin signal. Per-batch electricity allocation, sub-gram input attribution, and continuous fulfillment timing usually don't earn back the time they take. Track them only during specific decision windows or in regulated categories.
This post is a rebuttal. Not against tracking, which is sometimes exactly the move, but against treating granularity as a moral position when it's, almost always, an engineering question.
The 80/20 of cost tracking
Almost everything you need to price your products correctly is captured by four numbers per SKU:
- What your materials cost per unit.
- What your packaging costs per unit.
- A reasonable estimate of how long the unit takes to make.
- A monthly overhead line item divided across the SKUs you actually produced that month.
That's the floor, and for most cottage-scale sellers it's also close to the ceiling. With those four numbers you can price intelligently, spot a thinning margin, and decide which products to keep and which to retire. None of those decisions require per-batch electricity allocation, sub-gram input tracking, or stopwatch-grade fulfillment timing.
Rule of thumb. If a tracking habit takes thirty minutes a month to maintain and changes a decision you make less than once a year, it's decoration. There's no shame in decoration — but call it what it is, instead of mistaking it for accounting.
The trouble is that maker-business advice — especially the advice handed down in Facebook groups and YouTube comment sections — treats granularity as a virtue in itself. More detail equals more rigor equals more legitimacy. That's the same logic that produces beautiful, useless spreadsheets with seventeen tabs nobody opens twice. The detail isn't doing the work. It's just sitting there, looking responsible.
Here are the three habits where this matters most.
Habit 1: Per-batch electricity allocation
Somewhere in the second year of running a candle business, a particular thought arrives: I should really know what each pour costs me in electricity. The melter ran for forty minutes. The thermostat was on. The studio lights were on. There's a kilowatt-hour cost in there somewhere, and a serious business — a real business — would know what it is.
So you read the wattage off the back of the melter, divide your monthly electric bill by total hours of equipment use, and try to allocate the result across batches. Three months in, you have a spreadsheet that calculates electricity to the third decimal of a cent per candle.
You also have an answer: it's about four cents. Maybe seven on a long day. The candle sells for nineteen dollars.
The reason this exercise quietly disappoints almost everyone who attempts it is that for most cottage-scale crafts, electricity is one to three percent of unit cost. That's a real number, and it should land somewhere in your overhead bucket. But chasing it batch-by-batch doesn't give you cleaner data. It gives you the same data, computed thirty different ways.
Where it does pay off. Electricity allocation per batch starts to matter when energy crosses roughly five percent of unit cost. That generally means kilns running twenty-plus hours a week, walk-in coolers, freezer-dependent ice cream production, commercial coffee roasters, or 3D printers running unattended overnight. If you're a candle maker doing weekend pours, the bill belongs in overhead.
The right move for most cottage-scale sellers is to treat electricity as a flat monthly line item, divide it by SKUs produced that month, and move on. You'll be off by single-digit pennies per product. Those pennies are not what's setting your margin.
Habit 2: Sub-gram raw-material attribution
The second habit shows up in soapers, jewelers, herbalists, and ceramic artists. It looks like this: every drop of fragrance oil is logged. Clay scrap is weighed before and after each session. Every chip of silver from a wire cut is collected, weighed, and added to a "metal reclaim" inventory record. The wax that didn't make it into the mold gets scraped out and re-attributed to the next pour.
This is the habit that feels most virtuous, because it sounds like waste reduction. And in moderation, in a specific subset of crafts, that's exactly what it is. But moderation isn't usually what's happening. What's happening is a maker spending forty-five minutes per production session managing a tracking system designed for an industry where the inputs cost a hundred times what theirs do.
Here's the question that decides whether this habit is paying for itself: if I'm off by ten percent on this input, does anything change?
For a candle maker tracking soy wax to the half-gram, the answer is no. Soy wax at three dollars a pound means a ten-percent measurement variance costs roughly two cents per candle. The label won't change. The price won't change. The customer won't notice. The IRS won't audit you over it.
For a jeweler tracking sterling silver to the half-gram, the answer is yes. Silver at thirty dollars an ounce means a ten-percent variance is roughly three dollars per pendant. That's the difference between a sixty-percent margin and a fifty-percent margin. That's a real number that should be tracked.
The threshold. Track sub-unit material use when the input is either expensive enough that a ten-percent measurement error meaningfully changes your pricing, or regulated enough that a ten-percent measurement error could fail an inspection. Precious metals, essential oils above roughly two hundred dollars per ounce, IFRA-restricted fragrance load, scheduled-process acidified foods, and gemstones cross that bar. Soy wax, beeswax, common clay, generic fragrance oil, and stock cotton yarn do not.
The other place this falls apart is that the maker who tracks every gram usually doesn't also have a clean record of labor. They know the milligram of fragrance in the candle. They have no idea whether the candle took forty minutes or seventy. Labor is the variable that's actually moving the margin around, and the granular-input habit is, in practice, displacing it.
Habit 3: Hand-entered fulfillment minutes per order
The third habit is the most recent arrival, and it shows up most in sellers who've been told to "treat your time like a cost." Which is correct advice. The execution is where it goes sideways.
The execution looks like a spreadsheet, or a notebook, or a Notion database with a stopwatch column. Each order gets timed. Pick, pack, label, customer email, drop at the post office — broken down to the minute. After a few weeks, the maker has a beautiful dataset that tells them the average order takes 11.4 minutes.
What they do with that number, in practice, is nothing.
The reason is that for most cottage-scale sellers, fulfillment time isn't a variable cost in the way labor accounting treats it. It's a fixed evening. You set aside Tuesday and Thursday after dinner, you pack what came in, and the labor cost of fulfillment is your time — which is mostly already spent. What's varying is order volume, not the time-per-order.
That doesn't make the data useless. It makes continuous tracking of the data useless. The number itself is genuinely worth knowing — once. Maybe twice a year. Possibly during a decision window.
The threshold. Track fulfillment minutes only during decision windows — when you're evaluating a hire, testing a new packaging system, considering a price change, or building a wholesale price sheet that has to account for labor. The rest of the year, batch-sample for one week per quarter and then stop. You won't get better data by tracking every order in perpetuity. You'll only get less time to pack.
The version of this advice that does work is the spot check: once a quarter, time a representative week. Compare it to last quarter. If the average is creeping up, look at what changed in packaging, product mix, or how returns are getting handled. Then put the stopwatch away.
Three places "track everything" is actually right
Some conditions genuinely warrant the full-detail tracking advice. They're narrower than the advice itself suggests.
New product introductions. When a SKU is brand new and you're still stabilizing the recipe, granularity is how you find out where the actual cost is landing. Track everything for the first six batches, write it down, and then pick a level of detail that's sustainable. Granular tracking is a tool for learning a process, not for running one.
Compliance-driven categories. If you make acidified foods, cosmetics, herbal supplements, or pet treats, the regulatory regime decides your tracking detail — not your time budget. The FDA doesn't care that you found ingredient tracking tedious. AAFCO doesn't negotiate. State agriculture departments will ask for batch records, and the records they ask for are not optional. Track to the level the regulator requires, full stop. The exact detail varies by category and state — the Cottage Food Laws by State reference is a good starting point for figuring out what your state actually demands, and the batch tracking guide for food sellers walks through the lot-tracking spine that makes inspector requests a five-minute problem instead of a five-hour one.
The first ninety days of a new system. When you migrate from a spreadsheet to a real inventory system — or from no system to any system — granular tracking for the first three months is how you build trust in the numbers. Once the numbers feel right, prune. The granularity was a calibration tool, not a permanent operating mode.
Outside those three windows, the right level of detail is "enough to price correctly, spot trouble, and answer questions during a tax audit." A lot of the advice you're being given is calibrated for businesses an order of magnitude larger than yours.
What to actually do with the time you reclaim
The hours you stop spending on per-batch electricity, sub-gram fragrance attribution, and stopwatch-grade order timing don't disappear. They get redeployed. The redeployment usually moves the margin around more than the tracking did.
Spend the reclaimed hours on the things that actually shift unit economics: renegotiating with a supplier who quietly raised prices, photographing the SKU that's underperforming because the listing photo is bad, retiring the product line that's been sitting at a twelve-percent margin for two years, or building the wholesale relationship that turns one eight-hundred-dollar order into a recurring monthly account.
Those decisions need four numbers per SKU, plus a clean trail of which SKUs are eating which percentage of your time. They don't need three decimal places of electricity allocation.
The point of cost tracking is not to know everything. It's to know enough to make better decisions. The maker who knows their top three SKUs by margin, their slowest-moving stock, their average monthly overhead, and their supplier-price trend over the last six months is in better shape than the maker who has perfect granular records and no idea which product line to retire.
Ardent Seller is built around that gap. It tracks materials and packaging at the level that actually changes pricing, treats overhead as a single monthly line, and keeps fulfillment-time tracking optional rather than mandatory. There's a free tier and a feature tour if you want to see what "enough detail" looks like in software instead of a spreadsheet.
The dog is fine. Stop weighing her.
Related reading
- Against "Just Charge More" — The pricing-side companion to this post, on the popular advice that confidently sidesteps the actual cost stack underneath the price.
- Recipe Costing 101 — The foundational primer on which costs actually belong in a maker's unit cost calculation, and which can sit in overhead without losing accuracy.
- The Spreadsheet Breakup — Seven signals that your current tracking system has crossed from "useful" into "tax on your time," with a four-gate decision framework for what to replace it with.
Free resources
A few free downloads from the Ardent Workshop library that pair well with this post:
- Spreadsheet vs. Inventory Software Decision Guide — The decision aid for the question this post implies: at what scale and which workflows does a real inventory system out-earn the time it takes to set up?
- Monthly Inventory Count Sheet — A printable count sheet for the once-a-month rhythm this post argues for, instead of perpetual real-time tracking.
This article is provided for educational purposes only and does not constitute financial, tax, or business advice. Cost structures, pricing examples, and margin figures are illustrative and will vary by your specific circumstances. Consult a qualified accountant or small-business advisor before making financial decisions based on this content.