Do I really need 10 numbers? Can't I just look at revenue?

Revenue alone tells you what happened, not why. With 10 well-chosen numbers, you can spot a refund spike before it eats the quarter, notice that ad spend is quietly underperforming, and catch a churn problem 30 days before it shows up in revenue. Ten is small enough to fit on one screen and big enough to catch the problems revenue alone hides.

I don't run ads. Do the marketing numbers still matter?

If you don't run paid ads, replace the two ad metrics with two organic metrics — like organic sessions and email revenue. The shape of the scorecard is stable; the specific numbers track whichever channels you actually use.

Where do I put the final report?

Wherever you look during your weekly review. For most solo operators that's an Excel file, a Google Sheet, or a Notion page. The pipeline writes the numbers; the display is your call. The key is to see all ten in one place — not to open eight dashboards and add up in your head.

What if a number looks wrong?

That's the most valuable outcome of the first few weeks. Nine times out of ten, 'looks wrong' means you've discovered a data problem you didn't know you had — duplicate orders, a refund that didn't reconcile, an ad campaign that stopped tagging. The scorecard is a detector, not just a display.

How long until I get used to reading it?

The first week it's unfamiliar. By the third week you're scanning it in 30 seconds. By the second month it's the first thing you look at each week, and you'll wonder what you used to do instead.

All articles

Small Business Metrics KPIs Beginner Weekly Report

The 10 Numbers Every Small Business Should Track Each Week

A flagship weekly scorecard for small business owners: what to track, where to find each number, and how to stitch them into one report in under an hour.

By Edward van Eechoud April 16, 2026 7 min read

TL;DR. Most small businesses check three or four dashboards each week and reconcile the difference in their heads. A real weekly scorecard consolidates ten numbers — four about revenue, three about customers, and three about marketing — into one sheet you pull from your actual data. This post walks through which ten to track, where each one lives, and how to stitch them together in about an hour using a visual pipeline tool.

Why ten and not three

Revenue is the one number every owner checks. It’s also the latest-lagging indicator in your business. By the time revenue drops, the cause happened weeks ago: a churn spike in March, an ad campaign that died in February, a refund rate that climbed quietly for a month. Ten numbers, chosen well, let you see the causes before they move revenue.

The goal isn’t more metrics. It’s the smallest set that catches the biggest surprises.

The 10 numbers

Revenue health (4 numbers)

1. Gross revenue this week. Total of all orders placed, before fees and refunds. The headline number everyone asks about.

Where it lives: Shopify Orders export, WooCommerce, Stripe Payments, or your accounting tool.
How it can mislead: Gross revenue includes orders that later get refunded, which is why the next number matters.

2. Net revenue this week. Gross revenue minus refunds, minus processing fees.

Where it lives: Stripe Balance report (payouts are already net), or compute it yourself: gross - refunds - fees.
Why it matters: This is the number that actually lands in your bank account. It is the only honest baseline for marketing spend.

3. Orders (or transactions) this week. Count of completed orders.

Where it lives: Same place as revenue. For SaaS, count successful Stripe charges.
Why it matters: Flat revenue with declining orders usually means you’re relying on a few big customers — an exposure you want to see.

4. Average order value (AOV). Gross revenue ÷ orders.

Where it lives: Computed in the pipeline.
Why it matters: Small changes in AOV often explain revenue swings that look mysterious on the surface.

Customer health (3 numbers)

5. New customers this week. First-time buyers.

Where it lives: Orders where customer_created_at is within the week, or email not seen in previous orders.
Why it matters: If new customers dry up, everything downstream dries up 60 days later.

6. Repeat purchase rate (last 90 days). Share of customers who bought more than once in the last 90 days.

Where it lives: Computed from the orders export — group by customer, count orders.
Why it matters: Single best leading indicator of retention. For most small businesses, the target floor is 20–30% depending on category.

7. Refund / chargeback rate. Refunds and chargebacks divided by orders, for the week.

Where it lives: Stripe disputes + Shopify refunds.
Why it matters: A climbing refund rate eats margin silently. Catch it early.

Marketing & traffic (3 numbers)

8. Website sessions this week. Total sessions across all sources.

Where it lives: Google Analytics 4, Plausible, or whichever analytics tool sits on your site.
Why it matters: Traffic leads revenue by days to weeks for most businesses. Stagnant sessions mean the top of the funnel is narrowing.

9. Conversion rate. Orders ÷ sessions.

Where it lives: Computed from (8) and (3).
Why it matters: Reveals whether changes in revenue are traffic-driven or conversion-driven — two very different problems with two very different fixes.

10. Blended ROAS. Net revenue ÷ total ad spend across all channels.

Where it lives: (2) divided by the sum of Meta Ads spend + Google Ads spend + any other paid channel.
Why it matters: Blended ROAS is the one marketing number that can’t be double-counted. If per-platform dashboards say 6.0 but blended says 2.5, your bank balance is agreeing with the 2.5. More on this in Meta vs. Google Ads: How to Actually Tell Which One Is Selling.

The scorecard layout

The output is deliberately tiny. One row per week, one column per number, one comparison column showing week-over-week movement:

Week of	Gross	Net	Orders	AOV	New cust.	Repeat %	Refund %	Sessions	Conv. %	Blended ROAS
Mar 30	€18,420	€16,940	412	€45	284	22%	3.1%	14,800	2.78%	4.8
Apr 6	€17,630	€16,210	398	€44	261	24%	3.3%	13,900	2.86%	4.6
Apr 13	€19,100	€17,640	429	€45	302	23%	3.0%	15,400	2.79%	5.1

Ten columns. Three rows. Done. This is all that a small business owner needs to see during a weekly review to know whether the ship is still pointing in the right direction.

Building it once

In a visual pipeline tool like Flowfile, the scorecard is about six inputs and about twelve transformation nodes. Here’s the high-level map:

Inputs. Shopify Orders export, Stripe Balance report, GA4 export (or Plausible CSV), Meta Ads spend export, Google Ads spend export. Point each at a folder that holds the latest file; drop new files in each week.
Normalize. A Select node on each input renames columns into a consistent shape (date, order_id, email, amount, channel, etc.) and lower-cases emails.
Filter to the current week. A Filter node on each keeps only rows inside the reporting window.
Compute revenue metrics. A Group By produces gross revenue, orders, AOV, and refund totals from the orders data.
Compute net revenue. A Formula node subtracts refunds and fees from gross.
Compute customer metrics. Read the full historical orders file (not just this week) to calculate new-vs-returning and the 90-day repeat rate. A Join against this week’s orders tags each buyer as new or returning.
Sum ad spend. A Concat node unions Meta and Google ad spend into one table, followed by a Group By for the week total.
Assemble the scorecard. A series of Join nodes combines the summary tables into a single row of the ten numbers.
Append to history. A Read → Concat → Write pattern appends this week’s row to the running history file.
Write the output. A Write Data node saves the updated scorecard as weekly-scorecard.xlsx in whichever folder you check each week.

Budget about 90 minutes the first time. After that, running it is: drop this week’s exports into the input folders, open the pipeline, click Run. Ten seconds. Your scorecard is up to date.

What changes once you have it

Three things, within the first month:

You stop flying on revenue alone. The first time refund rate climbs a point while revenue holds, you’ll catch a problem two weeks before it would have shown up otherwise.
You make marketing decisions against honest numbers. Blended ROAS is uncomfortable the first time you see it. It is also the number your bank account is using. Better to disagree with your ad dashboard than with your bank.
You reclaim the hour you used to spend on dashboards. Ten seconds beats an hour of dashboard-hopping every single week.

Getting started

Flowfile is free, open-source, and runs locally on your laptop — your sales, customer, and ad data never leave your machine. The browser demo lets you try the canvas before installing.

A good first version of this scorecard: start with just the four revenue numbers. Build that much this week. Add the customer metrics next week. Add the marketing metrics the week after. By the end of the month you’ll have the full ten, and you’ll be running the cleanest weekly review of any small business in your category.

Related reading. Meta vs. Google Ads: How to Actually Tell Which One Is Selling goes deep on the blended ROAS number, and Stop Copy-Pasting Between Spreadsheets explains the broader pipeline pattern.

Frequently asked questions

Do I really need 10 numbers? Can't I just look at revenue?: Revenue alone tells you what happened, not why. With 10 well-chosen numbers, you can spot a refund spike before it eats the quarter, notice that ad spend is quietly underperforming, and catch a churn problem 30 days before it shows up in revenue. Ten is small enough to fit on one screen and big enough to catch the problems revenue alone hides.
I don't run ads. Do the marketing numbers still matter?: If you don't run paid ads, replace the two ad metrics with two organic metrics — like organic sessions and email revenue. The shape of the scorecard is stable; the specific numbers track whichever channels you actually use.
Where do I put the final report?: Wherever you look during your weekly review. For most solo operators that's an Excel file, a Google Sheet, or a Notion page. The pipeline writes the numbers; the display is your call. The key is to see all ten in one place — not to open eight dashboards and add up in your head.
What if a number looks wrong?: That's the most valuable outcome of the first few weeks. Nine times out of ten, 'looks wrong' means you've discovered a data problem you didn't know you had — duplicate orders, a refund that didn't reconcile, an ad campaign that stopped tagging. The scorecard is a detector, not just a display.
How long until I get used to reading it?: The first week it's unfamiliar. By the third week you're scanning it in 30 seconds. By the second month it's the first thing you look at each week, and you'll wonder what you used to do instead.