# product

Product analytics for the storefront website and the iOS and Android apps, shaped like a Segment warehouse export: sessions, the page, screen and track calls, identify calls, anonymous ids tied to customers, campaign touches and order attribution, the experimentation tool's tests and assignments, the feature flag list, and the analytics pipeline's daily rollup. It sees only what the tracking sent: visitors who block tracking or decline consent are missing, so it undercounts commerce. Every time is in UTC. Customers are commerce's ids (cus_...) and orders are commerce orders.

| Table | Rows | What it holds |
|---|---:|---|
| [anonymous_alias](#anonymous_alias) | 2,580 | One row per anonymous id the shop has tied to a customer: when it was first seen and when it was linked (a sign-in, an account made at checkout, or a session that opened signed in). A customer usually has several ids: other devices, the app and the web, cleared cookies. |
| [attribution](#attribution) | 4,224 | The analytics tool's credit for each order over the customer's campaign touches in the 30 days before it, under two models. last_click gives the last touch a credit of 1; linear gives each of n touches 1/n, rounded to four places with the remainder on the last. Each order's credits sum to exactly 1 per model; an order with no touch is direct with credit 1. A nightly job credits the previous day's orders, so orders of the last day in the data have none. |
| [daily_metrics](#daily_metrics) | 12,648 | The analytics pipeline's daily rollup: one row per market, local day and metric (sessions, users, orders, revenue). A day is the market's local day (India's, Amsterdam's, New York's). The job recomputes the last three days each morning, so events arriving later than that are not counted. Europe starts on 1 February 2024, when its store opened. |
| [event](#event) | 333,783 | One row per page call (a web page view), screen call (an app screen) or track call (a named action such as Product Viewed or Order Completed), with Segment's fields, as the warehouse loader delivers them. |
| [experiment](#experiment) | 14 | One row per A/B test, as the experimentation tool lists it. A visitor is assigned at their first page view while a test runs, or at Checkout Started for a checkout test. |
| [experiment_assignment](#experiment_assignment) | 5,889 | One row per visitor assigned to a test, as the flag service logged it: the visitor's first exposure while the test ran. The client then sends an Experiment Viewed call to event. |
| [feature_flag](#feature_flag) | 48 | One row per feature flag in the flag service: release flags, test flags (exp- and the test's key), operational kill switches (ops-) and market switches, with their rollout at the end of the data. |
| [identify](#identify) | 1,183 | One row per Segment identify call, which ties a visitor's anonymous id to a customer with their traits: when the account is made at checkout, and when the customer signs in on a device that does not yet know them. |
| [session](#session) | 58,337 | One row per session on the storefront website or the app, as the warehouse's sessions model builds it from events: a web session ends after 30 minutes without activity, an app session after 5 minutes in the background. Only sessions the tracking saw are here. |
| [utm_touch](#utm_touch) | 3,360 | One row per session of a known customer that arrived with campaign parameters, or from a social site without them, as the warehouse reads it. These are the touches attribution credits. |

## anonymous_alias

One row per anonymous id the shop has tied to a customer: when it was first seen and when it was linked (a sign-in, an account made at checkout, or a session that opened signed in). A customer usually has several ids: other devices, the app and the web, cleared cookies.

2,580 rows; key `alias_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `alias_id` | integer, key | 0.0% | 1 to 2580 | The alias row's key. |
| `anonymous_id` | text | 0.0% | `2eab287b-9038-4b39-9767-b9ed0446cac3`, `a28c2cef-8cf0-43c6-9185-76b2b99ab4ec`, `71514270-2966-41d1-ae63-643bda188970` | The device or browser's anonymous id. |
| `user_id` | text | 0.0% | `cus_Nv3DeFWiyRTLlA`, `cus_ENIcR9NSxcGpCo`, `cus_0AUFe33DuAzLdY` | The commerce customer the id is tied to. |
| `first_seen_at` | timestamp, UTC instant | 0.0% | 2023-09-16 02:10:21 to 2026-09-30 20:16:55 | When the anonymous id was first seen. |
| `aliased_at` | timestamp, UTC instant | 0.0% | 2023-10-01 08:21:18 to 2026-09-30 20:16:55 | When the id was tied to the customer. |

## attribution

The analytics tool's credit for each order over the customer's campaign touches in the 30 days before it, under two models. last_click gives the last touch a credit of 1; linear gives each of n touches 1/n, rounded to four places with the remainder on the last. Each order's credits sum to exactly 1 per model; an order with no touch is direct with credit 1. A nightly job credits the previous day's orders, so orders of the last day in the data have none.

4,224 rows; key `attribution_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `attribution_id` | integer, key | 0.0% | 1 to 4224 | The credit row's key. |
| `order_id` | integer, → `commerce.order.order_id` | 0.0% | 1 to 11247 | The commerce order credited. |
| `model` | text | 0.0% | `linear`, `last_click` | The attribution model: last_click or linear. |
| `channel` | text | 0.0% | `paid_social`, `direct`, `paid_shopping` | The touch's marketing channel, as the UTM map reads its source and medium (for example paid_search, paid_social, email, messaging, organic_social); direct when the order had no touch. |
| `campaign_code` | text | 43.0% | `CMP-2023-011`, `CMP-2022-013`, `CMP-2022-012` | The marketing campaign the touch's campaign parameter names, in its canonical form (CMP-<year>-<number>). Empty when the touch carried no recognisable campaign code. |
| `credit` | number, 4 places | 0.0% | 0.0588 to 1.0 | The share of the order credited to this touch, 0 to 1. |
| `computed_at` | timestamp, UTC instant | 0.0% | 2023-10-02 02:30:21 to 2026-09-30 02:32:59 | When the nightly attribution job credited the order, about 02:30 the day after it was placed. |

## daily_metrics

The analytics pipeline's daily rollup: one row per market, local day and metric (sessions, users, orders, revenue). A day is the market's local day (India's, Amsterdam's, New York's). The job recomputes the last three days each morning, so events arriving later than that are not counted. Europe starts on 1 February 2024, when its store opened.

12,648 rows; key `metric_row_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `metric_row_id` | integer, key | 0.0% | 1 to 12648 | The row's key. |
| `date` | timestamp, UTC instant | 0.0% | 2023-10-01 00:00:00 to 2026-09-29 00:00:00 | The market's local day the row counts. |
| `market_code` | text | 0.0% | `IN`, `US`, `EU` | The market: IN, EU or US. |
| `metric` | text | 0.0% | `sessions`, `users`, `orders` | sessions (sessions with an event that day), users (distinct visitors with an event that day), orders (Order Completed events) or revenue (the revenue of those events). |
| `value` | number, 2 places | 0.0% | -0.0 to 15520.0 | The metric's value. Revenue is in the market's currency (INR, EUR, USD) in major units and excludes shipping and tax. |
| `computed_at` | timestamp, UTC instant | 0.0% | 2023-10-04 06:10:16 to 2026-09-30 06:24:46 | When the job last computed the row: between 06:00 and 06:25 three days after the day, or on the last day of the data for the final days. |

## event

One row per page call (a web page view), screen call (an app screen) or track call (a named action such as Product Viewed or Order Completed), with Segment's fields, as the warehouse loader delivers them.

333,783 rows; key `event_row_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `event_row_id` | integer, key | 0.0% | 1 to 333783 | The row's key, in timestamp order. |
| `message_id` | text | 0.0% | `ajs-next-ad9ac20cd7023970d79a4d93404e2314`, `ajs-next-af40bd5791e8ee6ec276c3b333f4ae53`, `ajs-next-629921d96694baae9339b689dbdb22cf` | Segment's message id for the call: ajs-next- and a hash on the web (ajs-next-, epoch milliseconds and a uuid from June 2024), an uppercase UUID on iOS, a lowercase uuid on Android. |
| `session_id` | integer, → `product.session.session_id` | 0.0% | 1 to 58337 | The session the call belongs to. |
| `anonymous_id` | text | 0.0% | `9e70e38d-16f9-4340-ad74-9b1d569024bc`, `5fdae6c0-a5dd-44fc-bc57-7d41d306e0db`, `29FA9927-C965-4388-8BF6-03C76D6813EF` | The browser's or app install's anonymous id. |
| `user_id` | text | 61.8% | `cus_BbYHTEPSr1B6cP`, `cus_GCAJzH0Ybl1Y5M`, `cus_1XudikaNGJg7p0` | The commerce customer id once the visitor is signed in; empty before. |
| `type` | text | 0.0% | `page`, `track`, `screen` | page (web), screen (app) or track (a named action). |
| `event` | text | 0.0% | `Product`, `Product Viewed`, `Club` | The track call's name (for example Product Viewed, Product Added, Checkout Started, Order Completed, Experiment Viewed, Application Opened), or the name of the page or screen viewed (Home, Collection, Product, Cart, Checkout). |
| `timestamp` | timestamp, UTC instant | 0.0% | 2023-10-01 02:55:35 to 2026-09-30 21:14:59 | When the call happened, corrected for the device clock's error. |
| `properties` | text | 0.0% | `{"path":"/products/munduchilli","referrer":"https://www.querent.example/","search":"","title":"Mundu Chilli \| Pepperwind`, `{"url":"https://us.pepperwind.example/products/munduchilli","name":"Mundu Chilli","product_id":224,"sku":"SKU-SCHL-10019`, `{"path":"/pages/club","referrer":"https://us.pepperwind.example/products/munduchilli","search":"","title":"Pepperwind","` | The call's properties as a JSON object: a page's path, url, title and referrer; a product's id, sku, category, price and currency; an order's id, subtotal, total, revenue, shipping, tax, discount and currency; a test's id and variation; a sign-in's username and whether it succeeded. A customer whose deletion the privacy desk completed has no username here. |
| `sent_at` | timestamp, UTC instant | 0.0% | 2023-10-01 02:55:36 to 2026-09-30 21:33:39 | When the device sent the call, by the device's own clock, which can be off. |
| `received_at` | timestamp, UTC instant | 0.0% | 2023-10-01 02:55:35 to 2026-09-30 21:33:42 | When Segment's servers received the call. App calls are sent in batches, so they can arrive well after they happened. |
| `context_ip` | text | 0.0% | `2001:db8:56af:11b2:01cb:f548:330c:bf5d`, `198.18.105.240`, `2001:db8:064c:f0a9:dcd2:a075:79b5:b426` | The IP address the call came from. Kept after a customer's deletion, as the security logs it feeds are. |
| `context_user_agent` | text | 37.1% | `Mozilla/5.0 (iPhone; CPU iPhone OS 17_-1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.-1 Mobile/15`, `Mozilla/5.0 (Linux; Android 10; K) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/118.0.0.0 Mobile Safari/537.36`, `Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/118.0.0.0 Safari/537.36` | The browser's user agent string; empty on app calls. |
| `context_locale` | text | 2.0% | `en-US`, `en-IN`, `hi-IN` | The device's language and region setting, for example en-IN; empty where it was not sent. |
| `context_timezone` | text | 1.6% | `America/Los_Angeles`, `Asia/Kolkata`, `America/Chicago` | The device's time zone; empty where the browser withheld it. |
| `app_version` | text | 62.9% | `3.19.0`, `3.20.0`, `3.20.1` | The app's version on app calls; empty on web calls. |
| `plan_tier` | text | 97.0% | `reserve`, `spice`, `leaf` | The signed-in visitor's Pepperwind Club tier (leaf, spice or reserve), as the call sends it under the property name plan_tier. Empty when the call carries no tier. |
| `tier` | text | 99.0% | `leaf`, `reserve`, `spice` | The signed-in visitor's Pepperwind Club tier (leaf, spice or reserve), as the call sends it under the property name tier. Empty when the call carries no tier. |

## experiment

One row per A/B test, as the experimentation tool lists it. A visitor is assigned at their first page view while a test runs, or at Checkout Started for a checkout test.

14 rows; key `experiment_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `experiment_id` | integer, key | 0.0% | 1 to 14 | The test's key. |
| `key` | text | 0.0% | `pdp-sticky-add-to-cart`, `home-hero-everyday-chai`, `eu-free-shipping-banner` | The test's key in the tool, kebab-case. |
| `name` | text | 0.0% | `Sticky add-to-cart bar on product pages`, `Everyday chai hero on the home page`, `Free-shipping threshold banner` | The test's name. |
| `hypothesis` | text | 7.1% | `If the add-to-cart button stays on screen while shoppers scroll a product page, more of them will add to cart, because o`, `Leading the home page with everyday chai blends instead of the seasonal gift range will raise revenue per visitor, since`, `Showing the free-shipping threshold in a banner on every page will lift average order value in Europe, as shoppers add a` | The hypothesis the test's owner wrote before it started. Empty where none was written. |
| `primary_metric` | text | 0.0% | `add_to_cart_rate`, `revenue_per_visitor`, `aov` | The metric the test is judged on: checkout_conversion, add_to_cart_rate, revenue_per_visitor, aov (average order value), signup_rate or app_d7_retention. |
| `allocation` | text | 0.0% | `{"control":50,"treatment":50}`, `{"control":33.34,"variant_a":33.33,"variant_b":33.33}`, `{"control":25,"variant_a":25,"variant_b":25,"variant_c":25}` | The planned split of visitors between the test's arms, in percent, as JSON. |
| `started_at` | timestamp, UTC instant | 0.0% | 2023-10-09 06:09:02 to 2026-08-03 04:33:34 | When the test started. |
| `ended_at` | timestamp, UTC instant | 0.0% | 2023-11-06 05:19:00 to 2026-08-31 08:50:33 | When the test ended. |
| `owner_employee_no` | text | 0.0% | `EMP-10025`, `EMP-10034`, `EMP-10277` | The employee number of the product manager who owns the test. |

## experiment_assignment

One row per visitor assigned to a test, as the flag service logged it: the visitor's first exposure while the test ran. The client then sends an Experiment Viewed call to event.

5,889 rows; key `assignment_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `assignment_id` | integer, key | 0.0% | 1 to 5889 | The assignment's key. |
| `experiment_id` | integer, → `product.experiment.experiment_id` | 0.0% | 1 to 14 | The test. |
| `anonymous_id` | text | 0.0% | `0697f486-c2c7-401b-9c5c-ead9201a0e80`, `95f6a425-61b8-4c31-8ce1-a02917bc4663`, `777de0a9-c077-421f-a4d9-843e267aa7bd` | The visitor's anonymous id; assignment sticks to it across sessions. |
| `user_id` | text | 89.2% | `cus_HzaRNWoFUYFvkg`, `cus_KQ2qXl4TxNU4Wm`, `cus_84ZsMECvQpa8o2` | The commerce customer id when the visitor was signed in at assignment; empty otherwise. |
| `variant` | text | 0.0% | `treatment`, `control`, `variant_b` | The arm assigned: control, treatment, or variant_a to variant_c. |
| `assigned_at` | timestamp, UTC instant | 0.0% | 2023-10-09 06:21:46 to 2026-08-31 06:07:28 | When the visitor was assigned. |
| `browser` | text | 7.1% | `Chrome`, `Mobile Safari`, `Edge` | The session's browser, or the social app whose in-app browser opened it; empty on the app. |

## feature_flag

One row per feature flag in the flag service: release flags, test flags (exp- and the test's key), operational kill switches (ops-) and market switches, with their rollout at the end of the data.

48 rows; key `flag_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `flag_id` | integer, key | 0.0% | 1 to 48 | The flag's key. |
| `flag_key` | text | 0.0% | `checkout-guest-checkout`, `exp-pdp-sticky-add-to-cart`, `ops-disable-cod` | The flag's key in the service, kebab-case with a team prefix. |
| `description` | text | 0.0% | `Guest checkout without an account`, `Sticky add-to-cart bar on product pages`, `Turn off cash on delivery` | What the flag controls. |
| `rollout_percent` | integer | 0.0% | 0 to 100 | The share of traffic the flag is on for at the end of the data, in percent. |
| `created_at` | timestamp, UTC instant | 0.0% | 2023-10-04 12:03:49 to 2026-09-15 09:48:05 | When the flag was created. |
| `owner_employee_no` | text | 0.0% | `EMP-10047`, `EMP-10025`, `EMP-10051` | The employee number of the engineer or product manager who owns the flag. |
| `archived_at` | timestamp, UTC instant | 45.8% | 2024-01-26 15:56:40 to 2026-08-29 14:11:26 | When the flag was archived. Empty for a flag still live; kill switches are never archived. |

## identify

One row per Segment identify call, which ties a visitor's anonymous id to a customer with their traits: when the account is made at checkout, and when the customer signs in on a device that does not yet know them.

1,183 rows; key `identify_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `identify_id` | integer, key | 0.0% | 1 to 1183 | The identify call's key. |
| `message_id` | text | 0.0% | `ajs-next-5b54f20d3971af3928b7feb31191f9ed`, `ajs-next-87d85854e814317339d3f5fabac951df`, `ajs-next-971e5b4c238dc64aea3234a0536f0856` | Segment's message id for the call, in the same shapes as event.message_id. |
| `user_id` | text | 0.0% | `cus_1XudikaNGJg7p0`, `cus_8qHESP9k0XUUub`, `cus_KQ2qXl4TxNU4Wm` | The commerce customer identified. |
| `anonymous_id` | text | 0.0% | `30c9d093-ff86-48bc-bffa-47993467f299`, `884ceb0b-2969-4c18-88c4-8d5a335c7abf`, `e18b82af-c0ce-43c7-b7e1-51e9252f3bbc` | The anonymous id the call ties to the customer. |
| `timestamp` | timestamp, UTC instant | 0.0% | 2023-10-03 05:07:36 to 2026-09-30 04:59:03 | When the call was made. |
| `traits` | text | 0.0% | `{"email":"hema.murugesan@mail.example","first_name":"Hema","last_name":"Murugesan","created_at":"2022-09-17T03:17:46","m`, `{"email":"patrick.carpenter@usmail.example","first_name":"Patrick","last_name":"Carpenter","created_at":"2023-04-05T03:4`, `{"email":"malavika.mohan@inmail.example","first_name":"Malavika","last_name":"Mohan","created_at":"2023-10-05T09:06:44",` | The customer's traits as JSON: email, first_name, last_name, created_at (when the account was made) and market; only created_at and market for a customer whose deletion the privacy desk completed. |

## session

One row per session on the storefront website or the app, as the warehouse's sessions model builds it from events: a web session ends after 30 minutes without activity, an app session after 5 minutes in the background. Only sessions the tracking saw are here.

58,337 rows; key `session_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `session_id` | integer, key | 0.0% | 1 to 58337 | The session's key, in order of the sessions' start. |
| `anonymous_id` | text | 0.0% | `9e70e38d-16f9-4340-ad74-9b1d569024bc`, `5fdae6c0-a5dd-44fc-bc57-7d41d306e0db`, `29FA9927-C965-4388-8BF6-03C76D6813EF` | The browser's or app install's anonymous id. |
| `user_id` | text | 70.6% | `cus_BbYHTEPSr1B6cP`, `cus_GCAJzH0Ybl1Y5M`, `cus_1XudikaNGJg7p0` | The commerce customer id when any of the session's calls carried one (the visitor was signed in or made an account); empty for an anonymous session. |
| `platform` | text | 0.0% | `web`, `ios`, `android` | web, ios or android. |
| `app_version` | text | 82.0% | `3.19.0`, `3.20.0`, `3.20.1` | The app version on an app session; empty on the web. |
| `browser` | text | 18.0% | `Mobile Safari`, `Chrome`, `Samsung Internet` | The browser on a web session (for example Chrome, Mobile Safari, Samsung Internet), or the social app whose in-app browser opened it; empty on the app. |
| `os` | text | 0.0% | `iOS`, `Android`, `Windows` | The device's operating system: Android, iOS, Windows, macOS (an iPad reports macOS) or Linux. |
| `device_id` | text | 82.0% | `5C91FE0A-3ED1-40B9-B664-69E4FD79E722`, `955e23c400299680`, `3eacf737a3ed58eb` | The app's device identifier on an app session; empty on the web. |
| `started_at` | timestamp, UTC instant | 0.0% | 2023-10-01 02:55:35 to 2026-09-30 21:14:09 | When the session's first event happened. |
| `ended_at` | timestamp, UTC instant | 0.0% | 2023-10-01 02:59:21 to 2026-09-30 21:15:02 | When the session ended. |
| `landing_page` | text | 18.0% | `https://us.pepperwind.example/products/munduchilli`, `https://in.pepperwind.example/products/malabarearlgrey?utm_source=sb&utm_medium=paid&utm_campaign=cmp_2023_011`, `https://us.pepperwind.example/pages/our-estates` | The full URL of the session's first page, with its query string; empty on the app. |
| `referrer` | text | 54.5% | `https://www.querent.example/`, `https://www.querent.example/?hl=en-IN`, `https://m.palaver.example/` | The referring site, as the browser sends it (the origin only); empty when there was none and on the app. |
| `utm_source` | text | 51.2% | `sb`, `push`, `querent` | The utm_source parameter as it arrived, spelled as people and tools typed it; empty for an untagged visit. |
| `utm_medium` | text | 51.2% | `paid`, `push`, `shopping` | The utm_medium parameter as it arrived; empty for an untagged visit. |
| `utm_campaign` | text | 54.2% | `cmp_2023_011`, `newsletter-23`, `shopping_2023` | The utm_campaign parameter as it arrived: a campaign code in one of several spellings or a hand-written name; empty when absent. |
| `country` | text | 0.0% | `US`, `IN`, `FR` | The shopper's country, as a two-letter code. |
| `is_bounce` | boolean | 0.0% | False to True | True for a bounce: one page or screen, under 10 seconds, no conversion. |

## utm_touch

One row per session of a known customer that arrived with campaign parameters, or from a social site without them, as the warehouse reads it. These are the touches attribution credits.

3,360 rows; key `touch_id`.

| Column | Type | Empty | Range or values | Meaning |
|---|---|---:|---|---|
| `touch_id` | integer, key | 0.0% | 1 to 3360 | The touch's key. |
| `session_id` | integer, → `product.session.session_id` | 0.0% | 8 to 58328 | The session that is the touch. |
| `customer_id` | text, → `commerce.customer.customer_id` | 0.0% | `cus_BbYHTEPSr1B6cP`, `cus_84ZsMECvQpa8o2`, `cus_1XudikaNGJg7p0` | The commerce customer. |
| `occurred_at` | timestamp, UTC instant | 0.0% | 2023-10-01 08:21:18 to 2026-09-30 19:51:14 | When the session started. |
| `utm_source` | text | 0.0% | `m.palaver.example`, `l.shutterbloom.example`, `querent` | The utm_source as it arrived, or the social site's host for an untagged social visit. |
| `utm_medium` | text | 0.0% | `cpc`, `referral`, `shopping` | The utm_medium as it arrived, or referral for an untagged social visit. |
| `utm_campaign` | text | 8.8% | `CMP-2023-011`, `CMP-2022-013`, `cmp_2022_012` | The utm_campaign as it arrived; empty when absent. |
| `landing_url` | text | 0.0% | `https://in.pepperwind.example/collections/groundcoffee?utm_source=m.palaver.example&utm_medium=cpc&utm_campaign=CMP-2023`, `https://in.pepperwind.example/`, `https://us.pepperwind.example/?utm_source=querent&utm_medium=shopping&utm_campaign=CMP-2022-013&qclid=Cj0K6c10e621fbbf5f` | The full URL the session landed on, with its parameters. |