Skip to main content
The Explorer lets you build custom analytics queries using a flexible query builder (AftersellQL). You can select metrics, group results by dimensions, apply filters, and visualize data in charts or tables. Saved queries can be added to reports as widgets for ongoing monitoring.
You build queries visually with menus — no syntax required. If you prefer to type queries directly, the Explorer also exposes the underlying AftersellQL text. See Writing AQL queries below for the syntax reference.

​
Available metrics

These are the metrics you can pick in the Explorer, grouped the same way the metric picker groups them.

​
Revenue & profit

MetricDescription
RevenueUpsell revenue in your store’s native currency.
Revenue (USD)Upsell revenue normalized to USD for cross-currency comparisons.
Revenue Per VisitUpsell revenue per impression session. Cannot be broken down by product or placement.
Avg. Conversion ValueRevenue per accepted offer. Also referred to as Average Upsell Value.
Upsell Revenue Per OrderUpsell revenue (USD) divided by total orders. Store-level only.

​
Conversions

MetricDescription
ConversionsUnique sessions that accepted an offer.
Accept RateConversions divided by impressions.
Units SoldTotal units sold through upsell offers.
Decline RatePercent of post-purchase offers explicitly declined. Post-purchase only.

​
Engagement

MetricDescription
ImpressionsUnique sessions that saw an offer.
Show RatePercent of decisions that resulted in an impression.

​
Store performance

MetricDescription
Total Store RevenueTotal Shopify-paid order revenue. Store-level only — cannot be broken down by surface, funnel, placement, or device.
OrdersTotal Shopify-paid orders. Store-level only.
Average Total Order ValueStore revenue divided by orders. Store-level average order value.

​
Rokt network

MetricDescription
Rokt RevenueRokt-network revenue attributed to your store.
Rokt TransactionsRokt-network transaction count for your store.
Rokt Revenue / TransactionRokt revenue divided by transactions per time bucket.
Rokt ImpressionsTotal Rokt-network impressions across your store’s placements. Distinct from upsell Impressions.
Rokt ReferralsRokt-network referrals — positive engagements that sent the shopper to a Rokt partner.

​
Available dimensions

Dimensions let you break down metrics by a specific attribute. Not all dimensions are compatible with every metric.
DimensionDescription
DateGroups results by day, week, or month.
SurfaceThe upsell surface (for example, post-purchase, checkout, product page).
FunnelThe specific funnel the offer belongs to.
ProductThe upselled product.
PlacementThe placement within a funnel.
DeviceThe device type (desktop or mobile).
Flow typeThe type of upsell flow.
ExperimentThe A/B test or experiment variant.
OutcomeThe decision outcome (for example, eligible, out of stock).
Reason codeThe reason for a decision outcome.
ScopeThe decision scope (Flow, Experience, Placement, or ItemSlot).
Response typeThe offer response (Accepted, Declined, or Timeout).
CurrencyThe ISO currency code (for example, USD, EUR, GBP). Useful for multi-currency stores. Compatible with store-level metrics and most upsell metrics.
Some dimension and metric combinations are incompatible. For example, Decline rate and Show rate cannot be broken down by Currency. The Explorer automatically prevents incompatible combinations.

​
Writing AQL queries

Every query you build in the Explorer is an AftersellQL (AQL) statement. Most of the time you build queries visually — picking metrics, dimensions, filters, and a date range from menus — and never need to write AQL by hand. For power users, the Explorer also exposes the underlying query as editable text. This section is the reference for that text form: what the clauses mean, what values they accept, and a few ready-to-use examples.

​
How an AQL statement reads

An AQL statement is a single question made of clauses. Only SELECT and a time range (SINCE) are required; everything else is optional. When you include optional clauses, they must appear in the order shown below. 
SELECT    <metrics>                    -- what to measure (required)
WHERE     <filters>                    -- narrow the data
GROUP BY  <dimensions>                 -- break the numbers down
SINCE     <time range>                 -- the period to cover (required)
GRAIN     <time grain>                 -- bucket size for time series
COMPARE   <comparison>                 -- compare against another period
CHART     <visualization>              -- how to display the result
TIMEZONE  "<timezone>"                 -- timezone for date buckets
ORDER BY  <field> <direction>          -- sort the results
LIMIT     <number>                     -- cap the number of rows
A minimal example — daily upsell revenue and accept rate for the last 30 days: 
SELECT revenue, accept_rate
GROUP BY date
SINCE last_30d
GRAIN day
Keywords are not case-sensitive (SELECT and select both work) and statements do not end with a semicolon. String values are wrapped in double quotes; numbers and lists are not.

​
Picking what to measure and how to slice it

  • SELECT lists the metrics to measure, separated by commas — for example SELECT revenue, impressions, accept_rate.
  • GROUP BY breaks those metrics down by one or more dimensions, such as date, device, surface, or funnel. Without GROUP BY, you get a single total for the whole period.
For the full list of available metrics and dimensions — and which combinations are allowed — see Available metrics and Available dimensions above. The Explorer prevents incompatible metric and dimension combinations automatically.

​
Filtering with WHERE

WHERE narrows the data before it is measured. Combine conditions with AND. 
SELECT revenue, impressions, accept_rate, rpv
WHERE device = "mobile"
GROUP BY date
SINCE last_month
GRAIN day
Supported comparisons are =, !=, IN, NOT IN, >, <, >=, and <=. Use a list with IN to match several values: 
WHERE experiment IN ["variant_a", "variant_b"]

​
Time ranges and comparisons

Every query needs a time range, set with SINCE. Use a preset or a custom window.
FormExampleMeaning
PresetSINCE last_30dA rolling window ending today.
Custom windowSINCE 2025-11-28 UNTIL 2025-12-01A fixed range, using ISO dates (YYYY-MM-DD).
Available presets: last_1d, last_7d, last_30d, last_90d, this_month, last_month, and this_year.
  • GRAIN sets the bucket size for time series — hour, day, week, or month.
  • COMPARE overlays a second period so you can see change at a glance. Use previous_period (the equal-length window immediately before) or previous_year (the same window shifted back one year).
-- BFCM 2025 vs BFCM 2024
SELECT revenue, impressions, accept_rate
GROUP BY date
SINCE 2025-11-28 UNTIL 2025-12-01
GRAIN day
COMPARE previous_year

​
Choosing a chart and timezone

These optional clauses are usually set for you by the Explorer’s visual controls, but you can also write them directly:
  • CHART sets how the result is displayed: scorecard, line_chart, bar_chart, area_chart, or table.
  • TIMEZONE sets the timezone used to bucket dates, as a quoted IANA name — for example TIMEZONE "America/New_York". Defaults to UTC.
SELECT revenue
GROUP BY date
SINCE last_30d
GRAIN day
CHART line_chart
TIMEZONE "America/New_York"

​
Sorting and limiting

  • ORDER BY sorts results by a metric or dimension, with ASC or DESC.
  • LIMIT caps the number of rows returned — useful for “top N” style questions.
-- Top 20 products by upsell revenue this month
SELECT revenue, conversions, avg_conversion_value
GROUP BY product
SINCE this_month
ORDER BY revenue DESC
LIMIT 20

​
More examples

-- Daily performance vs the previous period
SELECT revenue, impressions, conversions, accept_rate
GROUP BY date
SINCE last_30d
GRAIN day
COMPARE previous_period

-- Mobile vs desktop accept rate over the last 90 days
SELECT impressions, accept_rate, rpv
GROUP BY device
SINCE last_90d
ORDER BY rpv DESC

-- Which surface is driving the most revenue?
SELECT revenue, impressions, accept_rate
GROUP BY surface
SINCE last_30d
ORDER BY revenue DESC
Once you have a query you like, save it and add it to a report as a widget so it keeps updating — see Manage widgets.

​
Timezone support

By default, queries run in UTC. You can override the timezone for any query directly in the Explorer toolbar, so that date-bucketed results (daily, weekly, monthly breakdowns) reflect your local time rather than UTC.

​
Setting a timezone for a query

  1. Open the Explorer in your Aftersell admin.
  2. In the toolbar, click the Timezone selector (next to Compare).
  3. Choose one of the available timezones from the list, or select Account default to use the timezone configured in your analytics settings.
  4. Run your query. Results are bucketed using the selected timezone.
The selected timezone is saved with the query. When you save and reload a query, the timezone is restored automatically.

​
Account default timezone

If you have Lock reporting timezone enabled in your analytics settings, selecting Account default in the toolbar uses that locked timezone for your query. The toolbar label shows the resolved zone, for example Timezone: Account default (Paris (CET)). If Lock reporting timezone is not enabled, Account default falls back to UTC.

​
Specifying a timezone in AQL

You can also specify a timezone directly in your AQL query using the TIMEZONE clause, which appears between CHART and ORDER BY: 
SELECT ...
CHART ...
TIMEZONE "Asia/Tokyo"
ORDER BY ...
When present, the clause overrides the toolbar selection for that query. The timezone is preserved when you save and reload the query.

​
Available timezones

The timezone picker includes the following options:
TimezoneExample location
UTCUniversal Coordinated Time
America/New_YorkNew York (ET)
America/ChicagoChicago (CT)
America/DenverDenver (MT)
America/Los_AngelesLos Angeles (PT)
America/Sao_PauloSão Paulo (BRT)
Europe/LondonLondon (GMT/BST)
Europe/ParisParis (CET/CEST)
Asia/DubaiDubai (GST)
Asia/TokyoTokyo (JST)
Australia/SydneySydney (AEST/AEDT)

​
How timezone affects query results

When a timezone is set, date-bucketing in your query uses local time instead of UTC. For example, an event that occurred at 2026-03-29T01:30:00Z (UTC) falls on March 28 in New York time (ET) but on March 29 in Paris time (CET). Setting the correct timezone ensures your daily, weekly, and monthly breakdowns match your business reporting expectations. Queries that do not include a timezone — including previously saved queries — continue to run in UTC, so existing results are not affected.

​
Need help?

If you have questions about the Explorer or want to enable access, contact the Aftersell support team through the in-app chat.