01
Why revenue is the least useful number you have
Revenue is easy to find, easy to compare, and easy to feel good about, which is why it dominates conversations between sellers. It answers no question that matters.
Two sellers turning over the same amount can be in entirely different businesses: one working ten hours a week on stock that converts in a month, the other working forty on stock that converts in a year and carrying most of their profit as unsold inventory. Revenue cannot tell them apart. Ratios can, because a ratio has a denominator, and the denominators — time, capital, stock — are your actual constraints.
02
The six worth tracking
Deliberately few. A dashboard nobody reads is worse than three numbers checked every month, because it produces the feeling of measurement without the fact of it.
| Metric | How to calculate it | What it tells you |
|---|---|---|
| Sell-through rate | Units sold in a period, divided by units listed at the start of it. | Whether what you are buying is what people want. The earliest warning available. |
| Inventory turns | Cost of goods sold over a year, divided by average inventory at cost. | How many times your capital cycles. Multiply by margin to see what the business actually returns. |
| Average days to sell | 365 divided by turns; or measure it per item from listing date to sale. | The same fact in units people can feel. Also the input to the ageing thresholds in the aged stock guide. |
| Net margin per unit | Sale price less fees, postage, packaging, and cost. | Whether individual transactions work. The margin calculator does this one. |
| Profit per handling hour | Net profit over a period, divided by hours worked in it. | What your time earns. The number that decides between categories and between activities. |
| Return on inventory investment | Annual gross profit divided by average inventory at cost. | What a dollar of capital produces in a year. The best single answer to "what should I buy more of". |
Three service numbers sit alongside these and belong to the account health guide rather than here: dispute rate, late dispatch rate, and cancellations. They are not profit metrics, but they gate your visibility, and visibility determines whether any of the six above can be earned at all.
03
Leading and lagging numbers
Metrics split into ones that tell you what is about to happen and ones that confirm what already did. Both are needed and they are used differently.
- Leading: sell-through, days to sell, impressions and click-through on new listings, the age profile of your stock. These move first, and they move while you can still do something.
- Lagging: profit, revenue, turns over a full year. These confirm, and by the time they move the causes are months old and already locked in.
The mistake is running the business on lagging numbers, because they feel more real. A quarter of declining sell-through is a quarter of buying decisions you can still change; the profit drop that follows it is a quarter you cannot.
The same logic applies to the metrics that control your standing, and it is worth saying twice: direction matters more than level. A dispute rate that has doubled while remaining under the threshold is the signal. The threshold breach is merely the consequence.
04
Segment or the averages will lie to you
A business-wide average hides the thing you needed to know. Two categories, one excellent and one losing money, average out to adequate — and adequate is a conclusion that recommends no action.
The four cuts worth making, in order of how often they change a decision:
- By category. Almost always the biggest spread, and the cut that tells you what to buy more of.
- By source. Which sourcing channels earn their Saturdays — the audit in the sourcing guide is this cut applied to supply.
- By channel. Fee structures and audiences differ enough that the same item earns materially different amounts in different venues.
- By price band. Cheap items are hurt disproportionately by per-order fees and postage, and many sellers find an entire band below which nothing is worth listing.
Segmentation is only possible if the records support it, which is the practical reason the COGS guide insists on cost, source, date, and channel per unit. Those four fields are what turn a pile of sales into an analysable business, and they cannot be added retroactively.
05
Check that you were paid what you expected
A metric nobody names and everyone should watch: the gap between the payout you predicted and the payout that arrived.
Marketplace deductions are numerous and some are easy to miss — commission on the postage portion, promoted-listing fees settled separately, currency conversion, per-order charges, adjustments from earlier refunds. A persistent gap between expectation and reality means your pricing model is wrong, and every price you set is wrong by the same amount.
- Predict before the payout For one week, predict the net on each sale before the payout lands.
- Compare line by line Compare against the actual payout line by line, not in total.
- Fold surprises into pricing Any deduction you did not predict goes into your pricing model permanently.
- Repeat after every fee change Repeat whenever a platform changes its fee structure, which they do.
The fees guide covers the anatomy of a deduction in full. The reason it appears again here is that reconciliation is a metric, not an accounting chore: it measures whether your model of your own business is accurate.
06
How often to look
Checking numbers too often is its own failure. Daily data on a small business is mostly noise, and reacting to noise produces the thrash of repricing everything on a quiet Tuesday.
- Weekly, briefly: anything with a deadline attached — orders to dispatch, open cases, listings expiring.
- Monthly, properly: the six metrics, segmented, plus the account standing check. Half an hour with the monthly review worksheet.
- Quarterly: the structural questions. Which categories and sources deserve more capital, which deserve none, and what the ageing profile of the shelf says about buying decisions two quarters ago.
- Annually: turns, return on inventory investment, and the honest profit-per-hour number for the whole year.
One rule that keeps this useful: every monthly review ends with a single decision, written down. Not a list — one. A review that produces observations and no change is a pleasant way to spend half an hour and nothing more.
07
Practice
Exercise
Calculate your turns
- Total the cost — not the sale price — of everything you sold in the last twelve months.
- Estimate the cost value of your average inventory over that period. Beginning plus ending, divided by two, is close enough.
- Divide the first by the second. That is your inventory turns.
- Then divide 365 by that number to get average days to sell, and compare it to what you would have guessed.
Check yourself
Category A returns 60 percent margin and sells in 200 days. Category B returns 25 percent and sells in 30 days. Which deserves your next dollar?
Category B, by a wide margin, and it is not close. The dollar in category A produces 60 cents once a year. The same dollar in category B produces 25 cents roughly twelve times, because it comes back and gets redeployed — around three dollars of gross profit against sixty cents. Margin percentage describes one transaction; return on capital describes what a dollar does over a year, and capital is the constraint that actually limits a small reselling business.
Your sell-through rate has fallen for three consecutive months while profit has stayed flat. Is anything wrong?
Yes, and this is precisely why sell-through is worth watching. Profit is flat because you are still selling the good stock bought earlier; falling sell-through means a rising share of what you buy is not moving. The unsold portion is accumulating as capital on the shelf, and the profit figure will follow the ratio down once the good stock is gone. Leading indicators move first, which is the entire reason to track them.
Progress is saved in this browser only. No account, nothing sent anywhere.
08
Common questions
What is a good sell-through rate?
It varies enormously by category, so the useful comparison is against your own trend rather than a published benchmark. A rate that is falling month on month is informative regardless of its level; a rate compared to somebody else's category tells you almost nothing.
Should I optimise for margin or for turns?
For return on capital, which is the product of the two. A modest margin turning twelve times a year beats a large margin turning once, because the dollar comes back and gets redeployed. Margin percentage describes one transaction; capital return describes a year.
How do I calculate profit per hour honestly?
Include the hours nobody counts: sourcing and travel, research, photography, messages, packing, post office trips, and admin. Most sellers who do this the first time find the number lower than expected and the distribution across activities more uneven, which is itself the useful result.
What is inventory turns and why does it matter?
Cost of goods sold over a year divided by average inventory at cost — how many times your capital cycles through stock. It matters because capital, not shelf space, is usually the binding constraint on a small reselling business, and turns is what determines how much work each dollar does.
How many metrics should I track?
Few enough that you actually look at them. Six on a monthly schedule, segmented by category, changes more decisions than a dashboard of twenty that gets opened twice a year.