Build the Scorecard Before You Build the Board Pack

A client of ours lost roughly $300,000 of Amazon revenue (call it ~$80,000 of contribution plus a rank-recovery tail) in about two weeks because a unit-ship problem at their third-party fulfillment center never surfaced. Inbound stalled, units shipping per day dropped from ~1,800 to a few hundred, the listing rank fell, and the bleed compounded for fourteen days before anyone flagged it in a leadership meeting.

The fix was a flight, a week onsite, and a quarter of damage at their best channel. Any competent ops manager checking Seller Central daily catches this in 48 hours, and that is exactly the point. There was no system that would have forced the catch. A weekly one-page scorecard with "units shipped vs. plan" as a single red/green line owned by one named human is exactly that system. It would have caught it in week one.

The most expensive pattern I see at $5M to $80M ecommerce and DTC brands is the same one every time: finance builds the dashboard before it builds the scorecard. So the company has elegant visibility that only the five smartest people in the building can act on, and zero of the early-warning structure the rest of the team needs.

A dashboard is not a scorecard, and most finance teams treat them as the same thing

A dashboard is a wall of indicators. CAC, CPM, contribution margin, inventory turns, AOV, cash on hand. All true, all useful, all needing a smart human to interpret them. Boards love dashboards. Investors love dashboards. There is nothing on a dashboard, though, that tells a non-expert reader that something is wrong and names the person who has to do something about it.

A scorecard does both. It is 10 to 20 measurables per team. Each line has a minimum acceptable threshold. Each line has one named owner. Each week, each line is red or green. If it is red, it gets discussed. If it is green, it doesn't.

That structure is deliberately simple so that a customer support lead and a head of operations are both looking at the same kind of object and both know instantly what they own and whether it is on fire. A CX lead can read a scorecard. A CX lead cannot reliably act on a dashboard. Both matter; only one of them catches the fulfillment problem before financial close.

Five rules for a scorecard that survives the quarter

1. Weekly. Not monthly. Not "when things feel off." Two weeks is already too long to react to most operational problems. Financial close is monthly because of accrual mechanics; that is not a reason to run the business at monthly cadence. Operations and cash move faster than that.

2. Minimum acceptable thresholds, not stretch targets. This is the single most common reason scorecards die. Finance puts the budget number or the stretch goal in the red/green column, the team starts the week red on half their lines, defensiveness sets in, the meeting becomes a tribunal, and inside a quarter the scorecard is quietly abandoned. Set the threshold at what "acceptable, no fire" looks like. A practical method is trailing-12-week average minus one standard deviation, not the plan number. Track stretch elsewhere: monthly performance reviews, quarterly planning. A fire alarm that goes off every week is wallpaper.

3. Measurables, not KPIs. Same data, better word. Yes, this is the EOS / Traction vocabulary and it works for the reason Wickman flagged: teams get less defensive about "measurables," and defensiveness determines whether problems get raised in the room or hidden until close. Hidden problems compound.

4. One named owner per line, and a rule for when ownership is contested. Shared-surface metrics will fight you. CAC is owned by marketing but moved by fulfillment speed. Gross margin is owned by finance but moved by procurement. The rule: pick the role with the most direct lever, not the most direct accountability. CAC owner is marketing because marketing decides what to spend tomorrow morning; ops owns fulfillment-speed as a separate line on its own scorecard. If two leaders genuinely both move it, the CEO assigns one and the other accepts it. Ambiguous ownership costs more than a debatable call.

5. The scorecard is a business-management tool, not a performance-management tool. It exists to surface problems and route them to a person. It is not a firing instrument. If someone owns important numbers and the numbers stay red across quarters, you'll know who needs to go without ever having to point at the scorecard to make the case.

How the meeting actually runs

A working scorecard meeting is 60 minutes, same day every week, hard stop. Tuesday 9 to 10am is the version I run; pick a day and don't move it. Last week's to-dos are reviewed first, before anything else, so accountability compounds.

Then the scorecard read itself is five minutes: the owner of each line says "on track" or "off track," and that is all. No commentary, no explanation in the read. Off-track lines roll to an issues list.

The next ~40 minutes is the issues block: Identify, Discuss, Solve (IDS, the EOS verb, because the audience will recognize it). For each of the top three issues: identify the real problem, discuss, decide a single next action with one owner and a 7-day due date. The last 10 to 15 minutes are headlines and to-dos.

One non-negotiable rule for chronic red: if a line has been red three weeks in a row, it does not get another week of conversation. It becomes a 90-day project with one accountable owner and a stated outcome. Chronic red is the single failure mode that quietly turns a scorecard back into a dashboard, and this rule is the only thing I've found that prevents it.

Data discipline is what determines whether the scorecard exists at all in month three

The scorecard collapses on data plumbing more often than on philosophy. Four rules:

– One person owns updating the scorecard file. Finance ops or a senior analyst, not a committee. A shared Google Sheet is fine; Ninety.io and Bloom Growth are the EOS-native tools.

– Owners submit their numbers by Monday 5pm. If the number isn't in by Tuesday morning, the cell is red by default. That single rule fixes most data discipline problems within two months.

– Every measurable shows 13 trailing weeks side by side. One quarter of context, enough to see a trend break before it becomes the trend.

– The source-data link lives in the cell comment. Seller Central report URL, ad platform view, bank balance. Any reader can verify the number in under thirty seconds. Numbers you cannot verify get gamed.

A sample executive scorecard for a $20M ecommerce brand

Five lines to make this concrete. Plan is ~$385K/wk net new revenue; thresholds are set roughly 12% below plan (trailing-12 average minus one standard deviation), not at it.

Executive scorecard (illustrative subset of 12 to 15 lines):

– Weekly net new revenue. Owner: CEO. Threshold: ≥ $340K

– Blended ROAS, trailing 7 days. Owner: Head of growth. Threshold: ≥ 2.4

– Units shipped vs. plan. Owner: Head of ops. Threshold: ≥ 95%

– Cash on hand (weeks of opex). Owner: CFO. Threshold: ≥ 14

– Inventory in stock on top-20 SKUs. Owner: Head of supply. Threshold: 100%

Each team runs its own scorecard with the same structure. A customer experience team's scorecard (shown abbreviated; a full CX scorecard runs 10 to 12 lines) might look like:

– CSAT (rolling 7-day). Owner: CX lead. Threshold: ≥ 4.6 / 5

– First-response time. Owner: CX lead. Threshold: ≤ 4 hrs

– Refund rate (orders). Owner: CX lead. Threshold: ≤ 2.5%

Lines roll up where they need to; nothing else is shared. The team builds its own measurables with finance in the room. A scorecard the team picked is one they will defend. A scorecard finance imposed is one they will work around.

Why the CFO should build it first, before anything else

For a company that does not have one, the scorecard goes in before the financial statement package. That is the order, and the reason is data integrity. A measurable is tied to source data: bank, ad platform, Seller Central, GA, warehouse log. You can verify it against the producing system in seconds. The number is true.

A financial statement is true-ish. It depends on accruals, deferrals, inventory cutoffs, bookkeeper judgment, and a chart of accounts someone built in a hurry in year one. Financials are essential; they're how you audit a quarter. But they are slow, late, and aggregated. You steer a business off the scorecard. You audit it off the financials.

The single best diagnostic for whether a finance organization actually runs well is what the scorecard meeting sounds like in the second month. The version you want sounds something like: "Units shipped is off. I'm tracking it to the 3PL receiving log, I think it's an inbound issue but I'm not sure yet. I need help." That sentence is hedged, partial, honest, and live, and it routes the problem to the room three months before close would have. You only get it when the scorecard has been used as a flashlight, not a stick.

What to do on Monday

If your company doesn't run a scorecard, the version to build this week:

– Pick 12 to 15 measurables that, if those were the only numbers you saw for a quarter, would tell you whether the business was healthy.

– Tie each one to source data, not a derived report.

– Set minimum acceptable thresholds. Trailing-12 minus one standard deviation is a fine starting heuristic.

– Put one human name beside each line.

– Recurring 60-minute meeting, same day every week. Re-baseline targets quarterly, not weekly.

The dashboard can come after. The board pack can come after. The forecasting model can come after. None of them catches the fulfillment problem in week one. The scorecard does. Build that one first.