The KPI cockpit: what you check daily, weekly, monthly
- You have a fixed KPI routine in three rhythms: daily, weekly, monthly.
- You know the definitions and target values of the central KPIs.
- You calculate the real monthly profit per product — not just revenue.
- You catch deviations early instead of chasing them.
Revenue is noise, profit is music — and without a cockpit you only hear the noise. This lesson builds your KPI routine: short enough to actually do, complete enough to miss nothing important. You know everything in it from previous lessons — here it gets rhythm.
1The daily check: 5 minutes, coffee-length
- Sales & revenue yesterday vs. the 7-day average — question downward outliers immediately (Buy Box? stock? suppressed listing?).
- Days of stock of top products against the reorder point.
- Ad spend within range? (Control only — optimization happens weekly, L2.)
- New customer questions/messages answered (24-hour rule).
- Account health scanned for red flags.
Tuesday, 7:40 a.m.: the AURELO spice grinder set shows 9 sales yesterday against a 7-day average of 19. Stock: there. Ads: ran. But the listing says “currently unavailable” — Monday's new main image ran into an automated check, and Amazon suppressed the listing in the meantime. By 8:00 the case is open with support, by noon everything is live again. Without the daily check this would have surfaced days later — at around €475 of revenue per lost day (19 units × €24.99).
Most important rule: the daily check is a CONTROL look, not an intervention trigger. Whoever twiddles prices and bids daily optimizes noise.
2The weekly review: 60–90 minutes
- Full PPC routine (harvest, negative, bids — L2).
- Rankings of core keywords updated (L3).
- Business reports: sessions and CVR per ASIN vs. last week (L4).
- Return reasons of the week read — every pattern onto the product list (L5).
- Inventory planning: reorder points checked, cash forecast rolled one week forward (L7).
3The monthly P&L: the only number that counts
| Line | Source |
|---|---|
| Revenue (net, after returns) | Amazon reports / settlement summary |
| − Amazon fees (referral, FBA, storage) | fee reports — don't forget storage! |
| − Cost of goods sold | your landed-cost calculation (Getting-started L7) |
| − Ad spend (total!) | ads console — this is where TACOS comes from |
| − Side costs (software, insurance, EPR, prorated) | bookkeeping |
| = Product profit before taxes | the number everything is measured against |
Revenue is what RUNS through the till — profit is what STAYS in the till at the end of the month. A market stall selling every apple for €1 that buys them for €1.05 makes a proud €10,000 revenue on 10,000 apples sold — and still a €500 loss. The monthly P&L is nothing but reaching into the till at month's end: what is actually still in there?
Plus three vitals per product: margin % (target ≥ 20 after everything), TACOS (trending down?), capital locked (stock value at purchase prices). This monthly statement drives portfolio decisions (L10) — and at exit, a clean history of these numbers is hard cash.
The monthly P&L of the AURELO spice grinder set: 580 net units sold × €24.99 = €14,494 revenue. Minus the 15 % referral fee (€2,174), minus FBA fulfilment and storage fees (here €2,550 combined), minus goods (580 × €4.50 = €2,610), minus ads (€1,750), minus prorated side costs (€320) = €5,090 product profit. The three vitals to go with it: margin 35 %, TACOS 12 %, capital locked €8,100 (1,800 units of stock × €4.50). Only this statement shows that a good third of the pretty revenue curve actually sticks.
4The KPI reference
| KPI | Meaning | Orientation |
|---|---|---|
| Sessions | product-page visitors | trend matters; a drop = visibility problem |
| CVR / unit session percentage | purchases ÷ visitors | <5 % weak · 10–15 % good · 20 %+ strong |
| ACOS | ad spend ÷ advertised revenue | steer against break-even ACOS |
| TACOS | ad spend ÷ total revenue | mature: often 5–15 %; trend beats value |
| Margin after everything | product profit ÷ net revenue | ≥ 20 % healthy |
| Days of stock | stock ÷ daily sales | above replenishment time + buffer |
| Order defect rate | defects ÷ orders | < 1 % (account health) |
| Return rate | returns ÷ sales | category-dependent; a rise = alarm |
5Tool: spreadsheet or software?
For 1–5 products a maintained spreadsheet is entirely enough — and it forces you to touch every number once (the real learning effect). Profit-analytics tools pay off with a growing portfolio: they pull fees and ad spend together automatically. The PPC-automation rule applies here too: understand manually first, then automate — otherwise you believe dashboards you can't verify.
Revenue euphoria: €30,000 monthly revenue feels like success — until the monthly P&L shows €400 left after fees, goods, ads and storage. Watch only the revenue curve and you may scale a loss-making business for months with growing conviction. The monthly statement is non-negotiable.
- Daily check established as a 5-minute routine (control, not intervention).
- Weekly review with a fixed slot: PPC, rankings, CVR, returns, stock.
- Monthly P&L per product calculated — including ALL cost blocks.
- Three vitals per product: margin, TACOS, capital locked.
- Targets noted; deviations produce tasks, not just worries.
- Tool fits the size (spreadsheet → software), but stays understood.
6Expert insight: Signal or noise — control limits for your cockpit
The most expensive cockpit mistake is not missing signals — it is reacting to noise. Sales figures fluctuate even with no cause at all. How much, you can calculate — and that turns gut feelings into fixed alarm thresholds.
- Daily sales: sales behave like random, independent arrivals. Rule of thumb: the typical fluctuation equals the square root of the average. At 19 sales a day: root of 19 ≈ 4.4 — the normal band (average ± 2 fluctuations) runs from about 10 to 28. A day with 14 is not news. A day with 9 sits below the limit — exactly the case from the daily-check example, and there a suppressed listing really was behind it.
- CVR: the spread of a proportion is roughly the square root of CVR × (1 − CVR) ÷ sessions. At 900 sessions a week and 15 % CVR: ± 1.2 points, normal band 12.6–17.4 %. The same math for ONE day (130 sessions) gives ± 3.1 points — daily CVRs between 9 and 21 % are pure noise. That is why CVR is read weekly, never daily.
- Small samples lie loudly: 3 returns on 30 sales is 10 % — sounding twice the usual 5 %. But at 30 sales the rate scatters by ± 4 points; the gap can be pure chance. Only from roughly 200 sales in the period does such a difference become reliable. (You know the same logic from reviews: 4.2 stars from 12 reviews says almost nothing.)
| Value | Basis | Normal band (± 2 fluctuations) |
|---|---|---|
| Sales/day: 19 | square-root rule | approx. 10–28 |
| Weekly CVR: 15 % | 900 sessions | 12.6–17.4 % |
| Return rate: 5 % | 30 sales | 0–13 % |
Two additions make the limits practical: trends beat single values — three declines in a row are rare under pure chance (roughly a 4 % probability), so the three-point rule from the FAQ is statistics, not gut feeling. And: limits live. After real level shifts (price change, season start, new main image), recompute the average — otherwise the cockpit alarms against a state that no longer exists. Weekday patterns are absorbed by comparing against the 7-day average, not against yesterday.
Cut the price after two weak days and you will see sales rise on day three — sales that would have come anyway, because noise returns to the average. From then on the price cut wrongly counts as the savior and stays: margin permanently gone, and the wrong lesson learned. Control limits prevent exactly this most expensive kind of decision.
The first lessons of every track are open to everyone. From here on you just need a free account — no subscription, no costs.
- 2The weekly review: 60–90 minutes
- 3The monthly P&L: the only number that counts
- 4The KPI reference
- 5Tool: spreadsheet or software?
- 6Expert insight: Signal or noise — control limits for your cockpit
- ✓Quiz: 6 questions with instant feedback
- All 41 lessons in 3 tracks — completely free
- Your progress is saved across devices
- Certificate + 100 free Listimo credits per completed track
Check yourself
6 quick questions — one at a time, instant feedback. With a free account your progress is saved.
Frequently asked
Isn't Amazon's own overview enough as a cockpit?
As a raw-data source yes, as a cockpit no: Amazon shows revenue prominently, but no Amazon view calculates your real product profit (after goods, ads, side costs). Yet that one number decides every one of your products — hence the own monthly P&L.
Which single KPI would I automate monitoring for first?
Days of stock against the reorder point — the most expensive avoidable mistake is out-of-stock. Then Buy Box loss and the ranking of your top 3 keywords: all values where hours matter, not weeks.
How do I catch creeping deterioration?
Through trends instead of absolutes: keep a small time series per core value (weekly values suffice) and watch for three consecutive points in the same wrong direction. Single values fluctuate — three weeks of falling CVR is a pattern with a cause.
The free Listing Check scores an ASIN from 0 to 100 in one minute: title, keywords, bullet points, images, A+ content and compliance — biggest weak spots first.
Everything in this academy comes from day-to-day selling practice — the same playbook behind Listimo, the tool that turns product photos into complete Amazon listings.