Buyer’s guide Sourced and dated
If clients send you CSV or Excel exports and Tableau workbooks and expect a finding, not a platform, use xmetric: it's free, needs no seat in the client's tenant, opens .twbx files, and keeps client data on your laptop. Pick Data Studio when the deliverable is a live Google marketing dashboard, Power BI when the client runs on Microsoft 365, and Tableau when they already pay for it.
Our pick
xmetric
No per-seat licence, no guest access to the client's tenant, works on whatever laptop you carry, opens the .twbx the client's last analyst left behind, and keeps the client's data on your machine.
Pick something else if
The client wants a live dashboard on Google Analytics, Ads or Sheets (Data Studio), lives in Microsoft 365 and will maintain the report in Power BI, or already pays for Tableau and wants the work inside it.
xmetric makes this page. Every vendor fact below links to the vendor’s own page and the date we checked it.
A retailer's order export, nine months of it. Marketplace is 43% of booked sales and the client wants to grow it. Take out returns (30% of marketplace orders come back) and it's 36% of net sales. Take out marketplace fees, shipping and cost of goods and it's 12% of the margin, at 8.9% against 37.8% for the webshop. Wholesale, with 30% of booked sales, earns the most. That's a finding a client pays for, and it took three formulas.
Computed by the xmetric engine
The data
An outdoor-gear retailer selling tents, jackets and boots through its own webshop, a third-party marketplace and wholesale to shops. 1,800 orders with discounts, fees, shipping, cost of goods and returns.
order_id | order_date | channel | category | units | gross_sales | discount | fees | shipping | cogs | returned |
|---|---|---|---|---|---|---|---|---|---|---|
| R00259 | 2025-01-01 | Webshop | Tents | 1 | 249.1 | 13.22 | 7.08 | 7 | 137 | 0 |
| R00536 | 2025-01-01 | Webshop | Tents | 1 | 235.58 | 18.03 | 6.53 | 7 | 129.57 | 0 |
| R00975 | 2025-01-01 | Webshop | Jackets | 1 | 173.38 | 5.32 | 5.04 | 7 | 78.02 | 0 |
| R01259 | 2025-01-01 | Webshop | Accessories | 1 | 24.17 | 1.76 | 0.67 | 7 | 9.67 | 0 |
Step 1
Marketplace is the biggest channel by a distance. The client wants to move budget into it.
Tableau
SUM([gross_sales] - [discount])SQL
SELECT
channel,
ROUND(SUM(gross_sales - discount)) AS booked_sales,
ROUND(SUM(gross_sales - discount) / SUM(SUM(gross_sales - discount)) OVER (), 3) AS share
FROM input
GROUP BY channel
ORDER BY booked_sales DESCRows
| Channel | Booked sales | Share |
|---|---|---|
| Marketplace | 152,195 | 43.3% |
| Wholesale | 103,729 | 29.5% |
| Webshop | 95,519 | 27.2% |
Step 2
Almost a third of marketplace orders come back. After refunds the three channels are close.
Tableau
SUM(IF [returned] = 1 THEN 0 ELSE [gross_sales] - [discount] END)SQL
SELECT
channel,
ROUND(SUM(CASE WHEN returned = 1 THEN 0 ELSE gross_sales - discount END)) AS net_sales,
ROUND(AVG(returned), 3) AS return_rate
FROM input
GROUP BY channel
ORDER BY net_sales DESCRows
| Channel | Net sales | Return rate |
|---|---|---|
| Marketplace | 105,210 | 30.2% |
| Wholesale | 100,977 | 2.4% |
| Webshop | 86,538 | 9.4% |
Step 3
Marketplace pays a fee on every sale and double shipping on every return; the returned goods earn nothing. It ends up last by a wide margin. Wholesale, a smaller channel by sales, makes the most money.
Tableau
[Margin] = [Net sales] - [fees] - [shipping] - [cogs]
Margin % = SUM([Margin]) / SUM([Net sales])SQL
WITH t AS (
SELECT
channel,
CASE
WHEN returned = 1 THEN 0
ELSE gross_sales - discount
END AS net_sales,
fees,
shipping,
cogs
FROM input
)
SELECT
channel,
ROUND(SUM(net_sales)) AS net_sales,
ROUND(SUM(net_sales - fees - shipping - cogs)) AS margin,
ROUND(SUM(net_sales - fees - shipping - cogs) / SUM(net_sales), 3) AS margin_pct,
ROUND(SUM(net_sales - fees - shipping - cogs) / SUM(SUM(net_sales - fees - shipping - cogs)) OVER (), 3) AS share_of_margin
FROM t
GROUP BY channel
ORDER BY margin DESCRows
| Channel | Net sales | Margin | Margin pct | Share of margin |
|---|---|---|---|---|
| Wholesale | 100,977 | 33,729 | 33.4% | 44.5% |
| Webshop | 86,538 | 32,742 | 37.8% | 43.2% |
| Marketplace | 105,210 | 9,317 | 8.9% | 12.3% |
Common mistake
A returned order has zero net sales, so its margin % is a division by zero and Tableau drops it from AVG. The average then ignores every return, which is exactly what sinks the marketplace. Divide the sums instead.
Tableau
AVG([Margin] / [Net sales])SQL
WITH t AS (
SELECT
channel,
CASE
WHEN returned = 1 THEN 0
ELSE gross_sales - discount
END AS net_sales,
fees,
shipping,
cogs
FROM input
)
SELECT
channel,
ROUND(AVG((net_sales - fees - shipping - cogs) / NULLIF(net_sales, 0)), 3) AS margin_pct
FROM t
GROUP BY channel
ORDER BY channelTableau
SUM([Margin]) / SUM([Net sales])SQL
WITH t AS (
SELECT
channel,
CASE
WHEN returned = 1 THEN 0
ELSE gross_sales - discount
END AS net_sales,
fees,
shipping,
cogs
FROM input
)
SELECT
channel,
ROUND(SUM(net_sales - fees - shipping - cogs) / SUM(net_sales), 3) AS margin_pct
FROM t
GROUP BY channel
ORDER BY channelDrop the client's CSV here: one row per order with a channel, the sales and the costs you have. The steps rerun in your browser and nothing is uploaded: the client's file never leaves your laptop.
Drop a CSV anywhere on this page
Up to 10,000 rows. The sample, orders.csv, shows the shape.
If this is you
A diagnostic or due-diligence project on exports and workbooks the client sends
xmetric
Free, private by default, and the finding goes back as a view-only link.
If this is you
A marketing client who checks GA4 and Ads numbers every morning
Data Studio
Free, live Google connectors, shared by link. Team workspaces are $9 per user per project per month.
If this is you
The client's staff will maintain the report inside Microsoft 365
Power BI
Build in their tenant; sharing needs Pro at $14.00 user/month, paid yearly.
If this is you
The client already runs Tableau Cloud and gives you a seat
Tableau
Work where their reports live; a Creator seat costs $75 per user/month, billed annually.
Yes. Open it in the xmetric app in your browser to keep building on it, or drop it into the workbook inspector to list its data sources, sheets and every calculated field.
No. CSV, Excel and Parquet files and .twbx workbooks are read inside your browser unless you choose to save the workbook online. That's easier to put in an NDA than a cloud upload.
Sign in, save the workbook online and send the client a view-only link. Until you save it, everything stays in your browser.
Today xmetric works with files: CSV, Excel, Parquet and Tableau workbooks. Live Postgres and BigQuery connections are coming.