Appearance
Cross-Dimensions: segments and revenue by location, channel, category and brand
Cross-Dimensions breaks your segments and predicted revenue down by another dimension, such as where customers live or which brand they buy. Open it from Insights > Customers > Cross-Dimensions.
Problem and intended user
A media buyer (the person who runs your paid ads) wants to know which channel and which place bring customers who are worth more. An ecommerce manager (the person who runs the store) wants to know which brands and categories attract the best customers, and which brands see the most returns. A segment list alone cannot answer these questions. Without the split, ad budgets and range decisions rest on averages that hide where value comes from.
Current workflow
A merchant often starts from a customer export and an order export from Shopify admin. The merchant adds a column for city, region, brand or category, then builds a pivot table in a spreadsheet. Acquisition channels come from Meta Ads Manager, Google Ads or GA4, and the merchant joins them to customers by hand. Return rates by brand come from a separate report. Each question needs a new pivot.
Prerequisites
- Your order and customer data is available in Nexus.
- Your RFM segments are built.
- For the Predicted revenue x Channel view, your orders carry a recorded traffic source.
How Nexus works
Nexus reads your customers, their segments and their orders. Fixed rules group and count them by region, city, brand, category or channel. Predicted revenue is not computed by these rules. It is the predicted revenue model's estimate, and Nexus shows it as given.
You approve nothing. The tab only reads data. It depends on your store data being available and on your RFM segments being built. This tab has no date range. It shows the current state of your customers.
Choose a view
A row of nine buttons sits at the top of the tab. The first one, RFM x Region, opens by default.
| Button | What it shows |
|---|---|
| RFM x Region | Customers per segment in each region. |
| RFM x City | Customers per segment in each city. |
| RFM x Brand | Customers per segment for each brand. |
| RFM x Category | Customers per segment for each category. |
| Predicted revenue x Location | Predicted revenue per city. |
| Predicted revenue x Channel | Predicted revenue per acquisition channel. |
| Predicted revenue x Category | Predicted revenue per category. |
| Predicted revenue by RFM | Predicted revenue against RFM score, for each customer. |
| Return Rate by Brand | Orders, returns and lost revenue per brand. |
Sort a table
Select a column header to sort a table. RFM x Brand, RFM x Category, Predicted revenue x Category and Return Rate by Brand always sort. The other views do not sort. See the limits below.
Exact output
Each view shows one table or one chart under its title.
RFM views
The four RFM views are tables, titled RFM Distribution by Region, by City, by Brand and by Category. Each row is a place, a brand or a category. Each column is a segment, followed by Total.
A cell shows how many customers of that segment belong to the row. A darker cell means more customers. A dash means none. Point at the (i) icon in a segment header to read what the segment means.
In the region and city views, a small grey line under each row name gives the country or the region.
Predicted revenue views
Predicted revenue is the model's estimate of lifetime revenue per customer. It is revenue, not profit.
In the channel view, a channel is the traffic source recorded on a customer's first order. Nexus shows it as recorded. Customers whose first order has no recorded source are not in this view.
The three tables share these columns:
| Column | What it shows |
|---|---|
| First column | The city, channel or category. |
| Avg predicted revenue | The average estimate. For a category, it covers the paying customers in the group. For a city or channel, it covers the customers in the group that have an estimate. |
| Median predicted revenue | The middle estimate across the same customers. Category rows show a dash on a live store. |
| Customers | The number of customers counted in the group. |
| Total predicted revenue | The sum of the estimates across the group. |
In the category view of a live store, the last column is Revenue to date instead. It shows actual revenue from paid lines in the category. It is not a prediction.
Predicted revenue by RFM score
This view is a scatter chart. The horizontal axis is RFM Composite Score. The vertical axis is Predicted revenue. Each dot is a customer from a sample of your customers. Dots take the colour of their segment. The legend lists the segments in the chart. Dots from a custom group appear in grey under Other.
Return Rate by Brand
Return Rate by Brand has these columns:
| Column | What it shows |
|---|---|
| Brand | The brand name. |
| Total Orders | Paid order lines for the brand, whatever their order status. It counts lines, not orders. Free-gift lines are left out and shown as a +N gifted note under the number. |
| Returns | Returned order lines. |
| Return Rate | Returned lines as a share of the lines in Total Orders. The colour changes with the rate. |
| Revenue Lost | Revenue lost to returns. |
What to do with it
| What you see | Decision or action | Owner |
|---|---|---|
| A region or city with many VIPs or Loyal Customers | Decide where to aim local offers and ad location targeting. Set the targeting in Meta Ads Manager or Google Ads, outside Nexus. | Media buyer |
| A channel with a high Avg predicted revenue and enough customers | Decide whether to move budget toward it. Change budgets in the ad platform, outside Nexus. Read the ads' own results on Performance > Ad performance before you move money. | Media buyer |
| A brand or category with many At-Risk customers | Decide whether to feature it in a win-back campaign or to review the range. The campaign runs in your email tool, outside Nexus. | Retention marketer (the person who runs email and repeat-purchase campaigns) |
| A brand or category that your best segments buy | Decide which products to show in ad creatives. | Creative strategist (the person who plans and makes ad creatives) |
| A brand with a high Return Rate or a large Revenue Lost | Decide whether to review product pages, sizing or the supplier. | Ecommerce manager |
Expected value and measurement
Cross-Dimensions can help you aim budget and effort at the places, channels and brands that bring the most value, and find brands that lose revenue to returns.
To check, open Insights > Customers > Cross-Dimensions later and read the same view again. Compare the Total of a row, Avg predicted revenue for a channel, or Return Rate for a brand with your earlier reading. Nexus keeps no history on this tab, so note the earlier figures yourself.
Nexus cannot isolate the effect of one decision. Predicted revenue is an estimate. Promotions, seasonality and ad changes move the same figures at the same time.
Limitations and exceptions
- Predicted revenue is an estimate. It comes from the predicted revenue model, it is revenue and not profit, and it can be wrong.
- Some views use a sample. The region, city, location and channel views are built from a sample of your customers, not from your full customer list. The Predicted revenue by RFM scatter uses the same sample, but shows no note about it. The sampled tables do not sort. They list rows alphabetically, and a line above the table says so.
- Channel needs a recorded traffic source. The channel view shows nothing when your orders carry none.
Rows with too few buyers
In the Predicted revenue x Category view, a category with too few paying customers shows a dash in Avg predicted revenue and the words too few buyers. An average over so few customers would mislead you. Point at the words to read the minimum number of paying customers. The Predicted revenue x Location and Predicted revenue x Channel views have no such rule. A city or channel with one buyer still shows an average.
Free gifts
RFM x Brand, RFM x Category, Predicted revenue x Category and Return Rate by Brand leave out free gifts. A note under the table says how many entries, or order lines, Nexus left out. A brand can show 0 orders with a +N gifted line when your store only ever gave it away.
What the tab shows when data is missing
Until your store data is available and your RFM segments are built, the page shows a panel that says what it is waiting for. If the page opens but no view holds data, the tab shows No cross-dimension data yet.
An empty view shows No data for this view and a line that says where the data comes from. The view fills once your data import is complete. The channel view says Channel attribution not available for this store when your orders carry no recorded traffic source.
Demo Mode
INFO
Demo Mode data is complete, so every view sorts and the sample note does not appear. The category view shows Total predicted revenue, not Revenue to date.