It's not a dashboard. It's the full cycle on top of your GA4: we find what's stalling and how much money is at stake, turn it into tasks with their own baseline, and measure each one at +7, +14 and +30 days.
Two sources report the same sale differently, and you don't know which one to use.
The one that declares its numerator.03
Nobody can explain whether what they did last month worked.
That's what the fourth step is for.
The cycle
Four steps. The fourth is the one that matters.
A dashboard stops at step one. The cycle closes when the action is measured against its own baseline.
01 · Sources
Connect
You connect your GA4 and the diagnosis runs on its own. Nothing to install.
OAuth to Google Analytics 4 and property selection, in minutes and without touching code.
Read-only. Credentials encrypted with AES-256-GCM, never returned by the API.
02 · Diagnosis
Discover
Findings prioritized by money at stake, confidence and urgency, not by what's easy to chart.
Rules engine with severity and action, status by pillar and score history.
The right metrics for each vertical: in ecommerce, CR uses purchases; elsewhere, key events.
03 · Tasks and alerts
Decide
Every task starts with its metric and its baseline already captured. That's what makes it measurable later.
A target metric per task and automatic capture of the before.
Alerts when a number drops sharply from one day to the next.
04 · Results
Measure
Every closed task is measured at +7, +14 and +30 days against its own baseline.
For volume metrics, net of a control metric.
If there isn't enough signal, the number degrades to N/A.
Three windows, never mixed
We know which period every number refers to.
They coexist on purpose, and the interface labels them.
Impact ledger
Results. The part almost nobody shows you afterward.
One row per closed task. What happened, what it was compared against, and what else was going on.
One-step checkout: from 412 to 468 purchases at +30 days. The control rose 4.1%, so what gets measured is +9.5% net. And if a promotion ran in the same window, we flag it in the row.
Purchases, before → +30 d412 → 468
Gross change+13.6%
Control metric+4.1%
Net of control+9.5%
Co-movement, not causation.
One-step checkoutpurchases per day
purchasescontrol: sessionsbaseline · 30 d
Impact ledger · one row per closed task
Closed task
Net of control
Reading
One-step checkoutpurchases · +30 d
+9.5%
412 → 468 · control +4.1%
Reorder the payment steprevenue · +14 d
+4.2% promotion in the window
not fully credited
Fix the source mappingsessions · +7 d
N/A
not enough signal to say
Sample data. No figure comes from a real account.
01
Before and after
At +7, +14 and +30 days against a baseline of the same length.
02
Control metric
For volume metrics, the change is measured net of a metric the task doesn't touch. It's co-movement, not an experiment.
03
Confounders flagged
Other tasks and logged campaigns in the same window are flagged in the row.
04
Money measured
The indicator only adds up metrics that are money. Calibration tells you how much to trust the estimate.
We don't claim the task caused the result. We say what happened, what it was compared against, and what else was going on.
Mercado
Your price, against the market. Product by product.
We import your catalog from GA4 and look up each product across the stores and marketplaces in your country. You see where you're expensive, where you're cheap and what's no longer selling.
PricesComparing 0 of 120 · best sellers first
MercadoLibreVTEX storesListings with Schema.orgYour listing
ProductYour priceMedianCompetitorsDifference
Market alert
50" 4K Smart TV: the competitors' median dropped 6.0% since yesterday.
Sample brands, products and prices.
Catalog
From your GA4
Your catalog comes from what you sell. We read your list price from your listing, or you enter it.
Sources
The ones in your country
MercadoLibre in seven LATAM countries, VTEX stores and any listing with structured data.
No false matches
Pending confirmation
A doubtful match stays out of the median until you confirm it's the same product.
Alerts
When the market moves
We let you know when competitors change the price of what you sell.
Points and redemptions
The real price
If you sell through a points program, you enter what a point is worth and we compare what's actually charged.
Studio
Dashboards for decision-makers. On the same data.
Dashboards built on the diagnosis numbers, not on a separate export. Each brand in the group with its own identity, and presentations you send as a single file.
New dashboardFerraroFerraro OnlineClub FerraroGenerate HTML
Presentation · OctoberFerraro · month to dateData as of each day's close · prepared for leadership
Sales this month
ferraro-october.htmlopens in any browser
Sample brands and data
01 · Templates
Start from a built layout
Month to date with best sellers, executive, ecommerce, acquisition and product. Every row of the grid, filled in.
02 · Builder
Widgets on your data
KPIs, time series, best sellers with each product's photo, and images, on the same metrics as the diagnosis.
03 · Brands
One for each brand in the group
You import the kit as JSON, CSS or text. We show you what we understood before applying it, and if there's no primary color, we don't make one up.
04 · Presentation
A file, not a login
Cover, sections, free text and dashboards frozen at publish time, in an HTML file that opens in any browser.
Built for groups
Several brands, several teams, the same number.
The main brandThe online storeThe loyalty programThe reporting agency
Standards
What we claim. And what backs it up.
Every claim on this list can be shown in a demo, with your account.
We don't alert you over noise.
A trend is flagged only if the model fits it well (R²) and the change is big enough to act on.
You compare apples to apples.
In ecommerce, conversion uses purchases; elsewhere, key events.
We'd rather say we don't know.
Without signal, the measurement stays at N/A. Without measured pillars, there's no score.
What we claimWhat backs it up
We tell you what worked, not what you'd like to hear.Before-and-after measurement, net of a control metric for volume metrics and with confounders flagged, that refuses to claim causation.
We know which period every number refers to.Three explicit, labeled windows, never mixed.
Our judgment is codified, not in someone's head.Rules engine with severity, measurement and a concrete action: an actionable sentence, not a chart.
Unusual changes show up separately.Regime changes and values outside normal variation (IQR) are flagged as findings of their own.
The cycle closes, and gets measured.The closed task is measured against its own baseline, with confounder detection.
How it's used
Your team uses it. If you need us, we run it with you.
Revenue runs the diagnosis, prioritizes and measures on its own. Your team decides and executes. And if you'd rather have someone run it, you can add that as a service.
Revenue
Runs the diagnosis on your GA4 and ranks findings by money at stake.
Revenue
Sets up each task with its metric and captures its baseline.
Your team
Decides what to do first and makes the changes.
Revenue
Measures every closed task at +7, +14 and +30 days and tells you whether it worked, without dressing it up.
NORQ · optional
If you'd rather we run it with you, or build the changes, you add it as a service, with a fixed-scope proposal.
Your access stays yours
Revenue looks. It doesn't touch.
Read-only access to your GA4, through Google OAuth.
Encrypted credentials with AES-256-GCM.
Never returned by the API, not even to us on screen.
Pricing
Everything is custom. The price is set before we start.
No plans, no packages: every operation is different. The path is always the same.
01
Diagnosis
You connect your GA4 and the diagnosis runs on its own: what's stalling, by how much and since when.
02
Fixed-scope proposal
What gets built, how long it takes and what it costs. Fixed before we start: we don't sell hours.
03
Build and measurement
The same people who ran the diagnosis build it, and the delivered system measures whether it worked.
It's for you if
Your revenue is between US$2M and US$50M.
You're a retail ecommerce business.
You have GA4 installed and suspect it isn't working right.
Someone on your team can make changes.
You've already spent on outside help and can't explain what you got back.
It's not for you if
You don't have GA4, or you installed it last month: without history, there's no trend.
I closed the month and a week later the number changed
Source data gets reprocessed, and nobody froze what was presented. NORQ Revenue labels which period every number refers to and alerts you when one stops matching.
Two sources report the same sale differently. Which one do I use?
The one that declares its numerator. In ecommerce, CR is calculated on purchases, not key events, and the product declares which one it uses before comparing.
How do I know whether what we did last month worked?
With the baseline captured when the task is created, measurement at +7, +14 and +30 days (net of a control metric for volume metrics), and flags for whatever else was happening in the same window. If there isn't enough signal, the number degrades to N/A.
Do I have to install anything or touch the site?
No. You connect your GA4 through Google OAuth, read-only, and choose the property. The diagnosis runs on its own.
Is it for me?
If your revenue is between US$2M and US$50M, you have GA4 installed with history, and someone on your team can make changes. It's not for you if what you want to buy is a dashboard.
Start with a diagnosis.
You connect your GA4 and the diagnosis runs on its own. If you then want us to run it with you or build the changes, we'll send you a fixed-scope proposal.