Data Warehouse & Modelling

Your marketing data warehouse, in your own BigQuery.

Verity builds and runs a marketing data warehouse for ecommerce in a BigQuery project that is yours: the data models, the semantic layer and the business glossary. You own the data, you decide who has access, and you keep everything if you ever leave.

How data gets in: managed ETL Next step: dashboards

Lineage

your BigQuery project

Sources

Shopify

GA4

Meta Ads

Google Ads

Data models

Orders

Customers

Ad spend

Sessions

Semantic layer

Revenue

New customers

ROAS

Margin

Used by

Data Chat

Dashboards

Reports

Segments

The problem

Your data sits in someone else's system

Most marketing tools store your data on their side. You see it through their screens, at their level of detail, for as long as you pay. Switch tools and the history is gone.

A warehouse of your own solves that, but building one takes a loading tool, a modelling project and someone to maintain both. Verity is the managed route: it builds the warehouse in your BigQuery and keeps it running, so you get the ownership without the engineering team.

Who it is for

✓ Ecommerce teams that want one place for store, ad, analytics and email data

✓ Marketing leads who want the same revenue number in every report

✓ Data teams that want to keep control of the warehouse and its access rights

✓ Companies that need their data stored in the EU, in a project they own

What is in it

From raw data to numbers everyone agrees on

Six parts, one warehouse. All of them live in your own BigQuery project.

Your own BigQuery

Bring the Google Cloud project you already have, or Verity sets one up in your name during onboarding. The data is stored in the EU, in the Netherlands region.

Verity and BigQuery →

Ownership and access

The project is yours and Google bills you directly. You decide who has access, you see what is queried and what it costs, and you keep everything if you ever leave.

Why you should own your marketing data warehouse →

Data models with lineage

Raw source data becomes clean tables for orders, customers, spend and sessions. For every model you see what it reads, what reads it, and whether the last run succeeded.

A semantic layer you approve

Verity proposes the metrics and dimensions from your data. A person reviews and approves them, and from then on every answer, dashboard and report uses them.

AI analytics on the semantic layer →

A business glossary

Each term has a definition in plain language: what counts as revenue, when a customer is new, which costs are in the margin. New colleagues and AI agents read the same one.

dbt, if you already use it

Verity reads the metrics your data team defines in dbt and follows your repository. On a pull request it reports which dashboards and monitors use the models you change.

Verity and dbt →

Modelling

One definition, from source to report

Take revenue. Shopify reports it with VAT, your ad platform counts orders that were later refunded, and finance wants it after discounts. In the warehouse there is one model for orders and one definition of revenue on top of it, and you can follow the line from the source to the report that shows it.

The models run on SQLMesh, with tests and lineage. You can read them, add your own, and export them. Before you change a model you see which dashboards, reports and monitors depend on it.

What you can always see

✓ Which sources and tables a model reads

✓ Which models, dashboards and reports depend on it

✓ When it last ran and whether the run succeeded

✓ The draft you are working on next to the version that runs

✓ The definition of every metric, in plain language

Compare

How this differs from the alternatives

Verity is an AI marketing data platform for ecommerce. The warehouse is one layer of it, managed for you and owned by you.

Compared with building it yourself

A self built stack means a loading tool, a modelling project, a BI tool and someone to keep them running. Verity delivers the same result as one managed service, in a project you own, so you are not locked in either way.

The real cost of the self built stack →

Compared with a platform that hosts your data

Many marketing data tools keep your data in their own systems. With Verity the warehouse is your BigQuery project. Verity works in it, and what was built stays with you.

Compared with a general warehouse service

Verity is built for marketing and ecommerce data on BigQuery. It does not load data into Snowflake, Redshift or Databricks, and it is not a modelling tool for all company data. Snowflake and Postgres can be connected as sources to ask questions on.

Around it

Before, after and further reading

The warehouse sits between loading your data and using it.

Previous and next step

Managed marketing ETL: how data gets in

Marketing dashboards: the first thing built on it

AI analytics and BI

Platform overview

Questions about the warehouse

What is a marketing data warehouse?

One place where the data from your store, your ad platforms, your analytics and your email tool is stored together and modelled the same way, so that an order, a customer and a euro of spend mean the same thing in every report. Verity builds and runs it in your own BigQuery.

Who owns the warehouse and the data?

You do. The BigQuery project is in your own Google Cloud account, or in one Verity creates in your name. You control the access rights and Google bills you directly. If you stop using Verity, the data and the tables that were built stay in your project.

Do I need a Google Cloud project to start?

No. Verity sets one up in your name during onboarding, or connects to your existing project. The guide setting up your Google Cloud organization explains the steps.

What does BigQuery cost?

BigQuery is billed by Google and is not part of the Verity price. For marketing workloads the costs are usually modest, and expected spend is part of the demo. See BigQuery costs for marketing teams.

What is a semantic layer, and why does it matter?

It is the agreed definition of your metrics, kept in one place. Whether revenue includes VAT, whether discounts and refunds are taken off, when a customer counts as new. Without it every tool and every AI assistant calculates its own version. With it, one definition is used everywhere.

How are the data models built?

The models that turn raw source data into reporting tables run on SQLMesh, with tests and lineage. You can read them, add your own models on top as a draft, validate them before they run, and export them.

We already use dbt. Do we have to move?

No. Verity reads the metrics from your dbt project on GitHub and uses them as the source of truth. Per source you choose dbt definitions, Verity definitions, or both. See Verity and dbt.

Does Verity work with Snowflake or other warehouses?

Verity runs on BigQuery. Snowflake and Postgres can be connected as sources that Verity answers questions on. Verity does not load your marketing data into Snowflake, Redshift or Databricks.

Is the warehouse included in Verity?

Data Management is one of the three parts Verity is priced on, next to Analytics and Activation, from €500 per month excl. VAT. It covers the pipelines, the data models and the semantic layer. BigQuery itself is billed by Google. See pricing.

See your data in a warehouse you own

Book a demo and see your own sources modelled in BigQuery, followed by a free trial on your own numbers.

Contact sales