Customer Analytics

Cohorts, retention, and lifetime value — computed from your real orders, not a sampled estimate.

Customer Analytics turns your order history into the cohort, retention, and LTV picture most stores never get to see. It reads every order, groups customers by the month they first bought, and tracks how each cohort comes back and what it's worth over time. Blend in your analytics and you can see which acquisition channels actually produce loyal customers.

What it does

Customer Analytics runs a daily job against your store's full order history and builds five views:

Weekly revenue — net revenue trend over the last 12 weeks.
Revenue by channel — a stacked breakdown across organic, paid, email, direct, referral, and social for the last four weeks. This view is driven by Google Analytics 4 attribution; without GA4 it shows an "attribution unavailable" state and the rest of the dashboard stays intact.
Cohort retention — a heatmap with one row per acquisition month and one column per month-since-first-purchase, so you can read repeat behavior at a glance.
Retention curves by cohort — the same retention data drawn as curves, one line per cohort, over 12 months.
Cumulative LTV by cohort — cumulative revenue per customer for each cohort over time, so you can see which months produced your most valuable buyers.

It's analysis only. It reads your data and draws the picture — it never writes anything back to your store.

Who it's for

Ecommerce teams that want real cohort and LTV analysis without a data analyst or a separate BI tool. If you can answer "what's my AOV" but not "are the customers I acquired in March worth more than the ones I acquired in June," this is for you.

Prerequisites

Customer Analytics runs through your connected integrations. You need one commerce source, and GA4 is recommended:

One ecommerce platform — Shopify or Magento 2 / Adobe Commerce. (Magento is in early access and gated per org — ask us if you need it.)
Google Analytics 4 — optional but recommended. It's what powers the Revenue by channel view by attributing your orders to acquisition channels. Every other view works without it.
Slack — optional, for the weekly digest.

Heads up: Customer Analytics is not available for Shopify App Store accounts. If your org signed up through the App Store, the fieldset shows a notice instead. Email support@onbelay.ai to discuss.

Setup

Customer Analytics is included in your On Belay plan at no extra charge. Enable it yourself like any other fieldset — open Dashboard → Fieldsets, find Customer Analytics, and turn it on (admin only). Once it's enabled, the setup wizard appears:

Go to the fieldset. Open Dashboard → Fieldsets → Customer Analytics. If the fieldset is enabled for your org, you land on the setup wizard.
Connect your sources. Pick your commerce platform (Shopify or Magento) — it must be connected and active in Integrations first. Connect GA4 too if you want channel attribution. The wizard won't continue until one commerce source is selected and connected.
Choose your look-back window. Pick 24 months (recommended) or 12 months of order history to analyze.
Finish setup. The fieldset is now configured and queued for its first run on the next daily cycle.

How does the setup wizard work?

The wizard is a two-step flow that runs the first time you open the fieldset, before any dashboard exists. You stay in the wizard until you finish it — entitlement and setup are separate, so an org that's enabled but hasn't completed the wizard still sees the wizard, because On Belay doesn't yet know where your order data lives.

Step 1 of 2 — Connect your data sources. This is where you tell Customer Analytics where your orders live and which optional sources to join in.

Commerce source (required). Pick exactly one: Shopify or Magento. The source has to be connected and active in Integrations before you can continue — if it isn't, the row shows a Connect link (or a red needs reconnection prompt if it's connected but in a broken state), and Continue stays disabled. If neither commerce source is connected at all, the wizard shows a single "Connect Shopify or Magento to continue" prompt with a link to Integrations. Magento is early-access only and gated per org behind a feature flag; until your org is flagged it carries a Beta badge and a "Contact your On Belay rep to enable it" note, and selecting it can't satisfy the gate.
GA4 (recommended). Connecting Google Analytics 4 is what populates the channel-attribution chart. It's recommended, not required — if you continue without GA4 connected, the wizard pops a "Continue without channel attribution?" confirmation reminding you that you can connect GA4 later in Integrations and your next daily run will pick it up. Cohort, retention, and LTV all work without it.
Slack (recommended). Optional, for the weekly brief. Without Slack you still get the dashboard — you just check it manually instead of getting the Monday brief.

If you change the commerce source after one was pre-selected, the wizard warns that your first analytics run will rebuild from scratch.

Step 2 of 2 — Confirm look-back window. Pick how far back to anchor cohorts: 24 months (recommended) or 12 months. As you toggle, the wizard shows the live Cohorts will start from date and the First refresh time (your next daily 09:00 UTC run). Cohorts start from the first order inside the window; earlier orders still count toward retention totals but don't anchor a cohort. The first run pulls the full window of orders and can take 5–20 minutes depending on store size — subsequent runs are incremental and finish in under a minute. Click Finish setup to save your configuration and queue the first run. If the save fails, the wizard keeps your selections and shows an inline error so you can try again.

Configuration

Set during the setup wizard, stored on the fieldset config:

commerceSource — shopify or magento. Chosen in the wizard; this is the store the analysis reads from.
lookbackMonths — how far back to analyze. Defaults to 24; the other option is 12.
Slack channel — optional, set on the Settings page after setup, for the weekly digest.

Settings (after setup)

Once the wizard is done, manage everything from Dashboard → Fieldsets → Customer Analytics → Settings. It's a calm control panel rather than a wizard, with one card per setting. Admins can change things; non-admins see everything read-only.

How do I change my commerce source, look-back window, or Slack channel?

Commerce source. The Settings page shows your current source (Shopify or Magento) with a connection status chip, and — for Shopify — your shop domain. Click Change… to switch sources. Because switching rebuilds your entire analytics history from scratch, On Belay asks you to confirm; the rebuild happens on the next daily run (09:00 UTC), and your current cohorts, retention, LTV, and dashboard stay visible until it finishes. You can only switch to a source that's connected and active (and not Magento-gated). If your current commerce source falls into a broken state, the card warns that daily runs will fail until you reconnect it in Integrations.
Channel attribution (GA4). Optional. The card shows GA4's connection status with a Connect or Reconnect link. Connecting GA4 here populates the channel-breakdown chart on the next daily run — no rebuild needed. Without it, cohorts, retention, and LTV still work.
Weekly brief delivery (Slack channel). Pick the channel where the Monday brief should post, using the channel picker. Leave it blank to skip Slack — the dashboard still updates every day. Slack must be connected in Integrations for delivery to work; if it isn't, the card tells you so.
Look-back window. Switch between 24 months (recommended) and 12 months. Changing the window is a heavy recompute, so when you save a changed window On Belay confirms first: your next daily run rebuilds every cohort, retention curve, and LTV figure from the new window (5–20 minutes), and the dashboard shows your current data until the rebuild lands.

The Slack channel and look-back window save together with the Save settings button (the page flags unsaved changes); the commerce-source switch commits through its own confirmation.

How it runs

The dashboard refreshes daily at 09:00 UTC — every chart is recomputed from your latest orders each morning. If you've connected Slack, a weekly digest posts on Mondays highlighting cohorts that crossed meaningful LTV thresholds (or noting when nothing moved). The daily refresh keeps the numbers current; the brief stays weekly so it isn't noise.

Data freshness, stale data, and the first-run state

How do I know how fresh the data is?

A small freshness strip sits just above the dashboard tabs and shows two times, in UTC: Last refreshed (when the most recent run completed) and Next refresh (when the next daily run is due). It's the quick at-a-glance answer to "am I looking at this morning's numbers?"

What happens if a refresh fails?

If the most recent run failed, the freshness strip turns amber and reads "Last run failed [time]. The next attempt runs [time]." When a run fails but you already have an earlier successful snapshot, the dashboard stays on your last good data and shows a stale-data banner across the top of the tab content — "Showing data last refreshed [date]. The most recent refresh failed — we'll retry on the next cycle." Your charts don't disappear; the banner just makes sure you never mistake stale numbers for fresh ones. The retry happens automatically on the next daily cycle.

What do I see before the first run completes?

Right after setup, before any run has finished, the dashboard shows a "Your first snapshot is being generated" state. The agent runs daily at 09:00 UTC, and your first snapshot appears within about 30 minutes — so check back around 09:30 UTC, or watch your Slack channel for the weekly brief. There's a shortcut to Settings in case you want to adjust anything while you wait.

Read-only — no approval gate

Customer Analytics is read-only. It computes cohorts, retention, and LTV from your order history and renders them on the dashboard. There's no writeback to your store, no publish step, and no approval queue — nothing to review before anything happens, because nothing leaves On Belay.

Outputs & where they land

Everything lands on the Customer Analytics dashboard at Dashboard → Fieldsets → Customer Analytics, organized into tabs: Overview (weekly revenue + revenue by channel), Cohorts (the retention heatmap), Retention (retention curves), and LTV (cumulative LTV curves). The optional weekly digest lands in your Slack channel.

Billing

Customer Analytics is included in your On Belay plan at no extra charge — no per-fieldset fee and no usage metering. Enable it yourself from Dashboard → Fieldsets like any other fieldset. The only thing you pay for is your base plan tier. See Plans & billing for tier details.

Tips

Connect GA4. Without it you still get cohorts, retention, and LTV — but you lose the channel attribution that tells you where your best customers came from.
Start with the 24-month window. More history means more complete cohort and LTV curves; drop to 12 months only if you have a reason to.
Give it a day. The dashboard refreshes on the daily cycle — your first full picture appears after the first run completes.
App Store account? Customer Analytics isn't available on App Store accounts — reach out to support@onbelay.ai if you need it.