> For the complete documentation index, see [llms.txt](https://docs.drivepoint.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.drivepoint.io/using-the-web-app/product-mapping.md).

# Product mapping

How product mapping works in Drivepoint and how to run a mapping session end to end: discover live sources, choose datasets, review canonical products, kits, and coverage, then Approve & publish.

Product mapping decides when two product records are the same physical product and records that decision. One canonical product per physical unit; every channel row ties back to it. The display name and SKU are picked from strings you already use in your own data. Drivepoint does not write a new catalog for you.

You run product mapping in a **mapping session** with your Drivepoint team. The session shows what is actually connected, waits for you to choose which datasets to map, builds a review, and publishes only after you reply **Approve & publish**.

{% hint style="info" %}
The screenshots on this page are product UI rendered from the Products page components and the mapping session's output format, using sample data for a fictional brand. Your own session shows your connections and your products.
{% endhint %}

### What the session can see

| Source type                  | Examples                                                             | What it contributes                                                                                      |
| ---------------------------- | -------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------- |
| Storefronts and marketplaces | Shopify, Amazon Seller                                               | Catalog or taxonomy, sales line items, kits                                                              |
| Retail feeds                 | Target, other retailer data                                          | Sales line items                                                                                         |
| Finance                      | QuickBooks invoices                                                  | Invoice line items (the sales-item line with its item reference and quantity). Never the general ledger. |
| 3PL and inventory            | Trackstar (Finale and other 3PLs), native Cin7, Cin7 Omni, Cin7 Core | Shipments or sale order lines, kit flags, units per case where the source reports one                    |

Ads, analytics, and general-ledger tables (profit and loss, balance sheet, accounts, journals, payments, invoice headers) are seen and skipped. A dataset that exists but is empty or stale is not treated as live.

### Run a mapping session

{% stepper %}
{% step %}

#### Start the session

Ask your Drivepoint team to map your products. "Map products", "reconcile products", "crosswalk SKUs", or "match Amazon to Shopify" all start the same flow. If you name a channel up front, the session still shows every live source first and does not group anything until you confirm the list.
{% endstep %}

{% step %}

#### Discovery

The session reads the product roster once (the cross-channel list the pipeline can already save and publish: one row per channel, store, SKU, and title) and lists the raw datasets behind your live connections. For each connection it keeps the tables that carry a product master, an item catalog, or SKU-bearing lines, and notes whether a 3PL or inventory connection carries a units-per-case or `unit_quantity` field, since warehouse rows are often recorded in cases.

If dataset listing is not available for your company, discovery is limited to the channels already on the roster, and the session says so.
{% endstep %}

{% step %}

#### Choose the datasets

You get one table, one row per product-bearing connection.

![Mapping session discovery table listing each live connection, what it carries, roster rows, and whether it is mappable today](https://80593965-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FePRvSjErXFousUDSWAz7%2Fuploads%2F2BAeYrGTcqQC2ldVwB09%2Fsession-discovery-table.png?alt=media)

| Column          | Meaning                                                                   |
| --------------- | ------------------------------------------------------------------------- |
| Source          | System and store or account                                               |
| Status          | Live, or empty or stale                                                   |
| Carries         | Catalog, sales line items, invoice line items, shipments, kits, case pack |
| Roster rows     | How many rows the pipeline can already save                               |
| Mappable today? | Yes, or review only                                                       |

Under the table, one line names the live connections that were seen and skipped, and any connection that looks empty or stale.

Reply with the sources you want, or `all`, and say whether review-only sources should be included. The session stops until you answer; "ok" or "thanks" is not a choice. Anything you leave off the list stays as it was.

{% hint style="warning" %}
A source already on the roster can be saved and published in this session. A source that is connected but not on the roster yet can only be reviewed. QuickBooks invoice lines, Trackstar shipments, and native Cin7 rows appear on the roster after the analytics model runs in your project. Until then, a save drops any source keys the roster does not know.
{% endhint %}
{% endstep %}

{% step %}

#### Review canonical products, flagged rows, kits, and coverage

The session groups the chosen rows into physical products. The canonical id is derived from product family, flavor or variant, and size, so the same product gets the same id every time; different flavors, sizes, and pack counts stay separate. The display name and SKU are selected from your own rows, preferring a storefront title and SKU. If there is no consistent SKU, the SKU stays blank.

![Mapping review with canonical products, flagged rows, kits and BOM, coverage by channel, and connected sources that are review only](https://80593965-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FePRvSjErXFousUDSWAz7%2Fuploads%2FDLoSDU0LiMRjUOmvL6Q9%2Fsession-review.png?alt=media)

1. **Canonical products.** One row per product: mapped id, name, SKU, family, category, format, size, how many source rows roll up to it, and which channels they came from.
2. **Flagged rows.** Rows that should not be confirmed silently, and why: name-only matches, a UPC that does not agree with the title and size, samples and donations, and a product seen in only one channel.
3. **Kits / BOM.** See the next step.
4. **Coverage.** For the channels you chose: total rows, confirmed, rejected, and still unmapped. Channels you did not choose are named under the table.
5. **Connected, not in roster (review only).** Only when you included a review-only source. It proposes which canonical each warehouse SKU looks like and the units per case the source itself reports (blank when the source has none). Nothing in this section is saved.

If you already have mappings, prior decisions stay. The review calls out what is new, what drifted, and what is being corrected. Unchanged rows are not sent again.
{% endstep %}

{% step %}

#### Kits and bill of materials

A kit sold as one SKU is still one catalog product: the parent. The review lists the parent, the kit type, and each component with its quantity. Every component must already be a canonical product in the same review. The kit type is `fixed` unless you describe a virtual or variable kit. If there are no kits, the section says "No kits". Demand units can be exploded from the bill of materials while the catalog stays one row per product sold.

Units per case is separate from the kit BOM. On save and publish, `units_per_case` is copied from the roster onto the catalog only when a Trackstar product has one inventory item and one positive `unit_quantity`. Cin7 lines and QuickBooks lines do not get a factor. Demand facts do not multiply by that factor yet.
{% endstep %}

{% step %}

#### Approve & publish

Read the review with your Drivepoint team. Reply **Approve & publish** to apply it ("publish these mappings" or "looks good, publish" also count; "ok" does not). If something is wrong, say what to change and you get a new review. Nothing is saved before that reply.

![Mapping session after Approve & publish, reporting what was saved, that the catalog is live, and that the review is complete](https://80593965-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FePRvSjErXFousUDSWAz7%2Fuploads%2FcOh1oYb1j2p6HVwRIvwy%2Fsession-approve-publish.png?alt=media)

After you approve, the decisions are saved and the catalog is published together. You hear one outcome:

| Outcome                                         | What it means                                                                                         | What to do                                                                                      |
| ----------------------------------------------- | ----------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------- |
| **Published**                                   | The catalog is live. Product Mix is live when the publish says reporting refreshed.                   | Nothing.                                                                                        |
| **Catalog published, reporting refresh failed** | The catalog is live; the reporting rebuild did not finish.                                            | Retry the publish. Do not save a second time. The review stays open until the refresh succeeds. |
| **Product Mix skipped**                         | The catalog is live. Product Mix was not rebuilt (the table or the mapped columns are not there yet). | Nothing; retail and ecommerce still refresh overnight.                                          |

If the review covered every product that was new since the last review, that review is marked complete on the successful publish. If you mapped only part of the new products, the rest stay New.
{% endstep %}
{% endstepper %}

### The Products page after publish

The **Products** page in the web app shows the published catalog with its mapping details: total products, mapped, rejected, unmapped, kits, last modified, and the reporting state from the last publish.

A published kit shows a **Kit** chip and a component count. Expand the row to see the bill of materials.

![Products page with a kit row showing the Kit chip and component count](https://80593965-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FePRvSjErXFousUDSWAz7%2Fuploads%2FUQhiJofYHOEMcznH4fnU%2Fproducts-page-kit-chip.png?alt=media)

![Products page with the kit row expanded to show its components and quantities](https://80593965-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FePRvSjErXFousUDSWAz7%2Fuploads%2FEgbYcpRy4E36Q3shmlVv%2Fproducts-page-bom-expand.png?alt=media)

Products that arrived since the last full review stay marked **New** until a publish covers that whole batch. The banner tells you how many are new, sorts them first, and **Show new products** filters to just those rows.

![Products page banner showing products that are new since the last human review, with New chips on the top rows](https://80593965-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FePRvSjErXFousUDSWAz7%2Fuploads%2FemO96ZoQGSvudOvggYkX%2Fproducts-page-new-since-review.png?alt=media)

The **Reporting** field reflects the last publish outcome. When Product Mix was skipped, the page says so and notes that retail and ecommerce still refresh overnight.

![Products page Mapping Details card showing Catalog published with Product Mix not live yet](https://80593965-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FePRvSjErXFousUDSWAz7%2Fuploads%2FhKDixZ9GNrgeWUVy1AXp%2Fproducts-page-publish-outcomes.png?alt=media)

### Source-specific rules

<details>

<summary>QuickBooks</summary>

QuickBooks is offered when its invoices table is live. The product grain is the sales-item line only: `SalesItemLineDetail` with an `ItemRef` (item id and name) and quantity. If an items table is also live, it is the item master for those lines, not a second dataset to map. Profit and loss, the balance sheet, accounts, journals, payments, and invoice headers are not mapped. QuickBooks lines do not receive a units-per-case factor.

</details>

<details>

<summary>Cin7 and Trackstar</summary>

The session reads the connection, not the display name alone. A Trackstar connection named Cin7 is the Trackstar path, not native Cin7, and its shipments are Finale shipments when the connection is Trackstar. Native Cin7 and Cin7 Omni rows come from sales order line items. Cin7 Core rows come from sale order lines. A Cin7 Core connection with data refresh disabled is offered only when its Sale table already exists; if it does not, there is no Cin7 Core source to offer. Cin7 lines do not receive a units-per-case factor.

</details>

<details>

<summary>Where the mapped taxonomy shows up downstream</summary>

Right after a successful publish, the mapped-products catalog is live, and Product Mix is live when the publish reports that it refreshed. Retail and ecommerce figures that are not rebuilt at publish time refresh overnight.

In the warehouse, roster rows for QuickBooks invoice lines, Trackstar shipments, and native Cin7 appear after the analytics model runs in your project. After the warehouse models run, ecommerce and retail product-level reports gain the mapped taxonomy as added fields. The existing `product_category` column is not overwritten. Older partitions stay empty until those dates are backfilled.

</details>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.drivepoint.io/using-the-web-app/product-mapping.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
