Skip to content
BQSyncby Practical Vision

Managed analytics data
in your warehouse.

BQSync turns the GA4 BigQuery export into maintained, relational tables your teams can query. Collection is reviewed, nested events are structured, processing is monitored, and the workflow is adapted when upstream platforms change.
See pricing

Existing Practical Vision client? You do not need the Data Audit. Request BQSync access.

Each refresh, in your projectAug 20
  1. 1

    The export lands

    analytics_318492371.events_20260820
  2. 2

    The export is read, not changed

    The raw tables stay exactly as Google Analytics 4 wrote them.
  3. 3

    Your transformations run

    The table set is designed around your collection, so it is yours, not a fixed template.sessionsproduct_viewscheckout_stepspurchase_events+3 more
  4. 4

    Anything that fails, we handle

    Runs are retried, escalated to a person if they stay failed, and reported to you in plain words.
Last 14 daysAll tables up to date

The problem

Exporting data is not the same as having data you can use.

The GA4 BigQuery export is a powerful foundation, not finished analytics data teams can use.
Turning it into something BI, product, and marketing can work with takes more than moving events across.

01

Nested data takes repeated work

The event-level detail is valuable. Repeatedly parsing nested fields and rebuilding the same logic across dashboards, notebooks and models gets between your analysts and their questions.

02

A running pipeline is only part of the picture

Infrastructure can work perfectly while the underlying tracking is incomplete or inconsistent. Export success is not the same as collection quality.

03

Your measurement setup keeps changing

Google updates schemas and behavior. Your organization adds events, parameters and tracking requirements. Someone still maintains the work between collection and analysis.

The SQL

One question, asked two ways.

Purchases with page and session, August 1 to 20

Querying the raw export

select
  (select value.string_value from unnest(event_params)
     where key = 'page_location')   as page_location,
  (select value.int_value    from unnest(event_params)
     where key = 'ga_session_id')   as session_id,
  (select value.double_value from unnest(event_params)
     where key = 'value')           as revenue,
  user_pseudo_id,
  timestamp_micros(event_timestamp) as event_time
from `analytics_318492371.events_*`
where _table_suffix between '20260801' and '20260820'
  and event_name = 'purchase'

Twelve lines, half of them unnesting. Every dashboard, notebook and model writes its own copy, and each copy drifts.

Querying your managed tables

select page_location, session_id, revenue,
       user_pseudo_id, event_time
from `metacortex.pv_analytics.purchase_events`
where event_date between '2026-08-01' and '2026-08-20'
purchase_eventsrefreshed every 30 min
checkout_stepsrefreshed daily
product_viewsrefreshed daily

The unnesting moved into maintained relational tables, written once and kept working.

What BQSync does

More than a data pipeline.

Raw where it matters. Relational where it helps.
Human expertise around the process, so the managed pipeline keeps working when the inputs move.

Collection quality

The analytics implementation is reviewed before anyone treats the resulting data as trustworthy.

Auditedbefore the first table is built

Relational data

Nested analytics data becomes structured tables designed for plain SQL, BI tools and downstream models.

sessionspage_viewssearch_eventsform_submissions

Managed operation

The processing is monitored and maintained as an ongoing service, not handed over as a project dump.

every updatewatched, retried, reported

Change support

When upstream structures change, including changes Google introduces, the processing is adapted rather than broken.

new eventsnew parametersnew schema
starting fromanalytics_318492371.events_*

The same event-level information stays available to product, marketing and BI.

How it works

Managed from collection to usable data.

We review how you collect, design the tables around what you actually have and what your teams need to answer, then keep the workflow operating over time.
All of it inside your warehouse.

Collected by you

GA4 property

Your existing measurement setup, exporting to BigQuery daily, streaming, or both.

analytics_318492371events_20260820events_intraday_20260821

Managed by Practical Vision

Audit and design

Your collection is reviewed, and the table design follows from what it actually contains.

Transformations

Nested events flattened into the tables designed for you.

Scheduled updates, last 14 days3 active
SucceededFailedBlockedRunningQueued

When an update fails we retry it, and if it stays failed one of us picks it up. You are told what happened, in words.

Owned by you

Your BigQuery environment

Raw events stay where they landed. The relational tables sit beside them.

sessionsUp to date
form_submissionsUp to date
purchase_eventsUp to date

Read by BI, product and marketing, with your own CRM and revenue data alongside.

01

Audit

Review the implementation, what is actually being collected, and what would undermine confidence in it.

02

Design

Work out which tables and columns you should have, from your collection and the questions your teams need to answer.

03

Build

Create those tables inside your own BigQuery environment.

04

Operate

Refresh them on a schedule set for you, from daily to every 30 minutes.

05

Monitor

Watch the workflow, and show your team the state of it in the dashboard.

06

Maintain

Adapt the design and the processing as relevant upstream structures change.

What you see

Managed doesn't mean invisible.

We operate the infrastructure.
Your team still sees what is happening, in plain words rather than pipeline jargon.

The customer home screen: a notice that one of four tables is behind, tiles for tables up to date, latest updates, scheduled updates and BigQuery access, a chart of scheduled updates over the last fourteen days, and the status of each table.

1 of 4 tables is behind.

The home screen names the tables that did not update on schedule, and says whether anything is needed from you. Usually nothing is.

Every update, with its reason.

Each refresh is listed with its table, attempts, rows written and what went wrong if it failed.

SucceededRunningQueuedFailedCanceled

BigQuery access

Working

The connection to your project is checked on a schedule, so a revoked grant surfaces here before it surfaces in someone's query.

Status is written as words, never as color alone, and a stage and its health are shown separately. Approved and needs reconnect can both be true at once.

The updates screen: a filter row for all, succeeded, running, failed and canceled, above a table of refreshes with table name, status, attempts, rows written, start time and details.
Updates. Every time one of your tables was refreshed.
The BigQuery access screen: access is working, a week of daily access checks, and where the tables are written.
BigQuery access. Where your tables are written and who can read them.

Ownership

Your data. Your warehouse. Your queries.

We build the managed analytics data inside your environment. If you decide to continue without us, the tables and the queries behind them stay exactly where they are. What you would stop paying for is the operation around them: monitoring, maintenance and expertise.

Data stays in your environment

Your analytics data is processed inside your own BigQuery project. It does not move to ours.

Tables remain yours

The relational tables we create stay in your project, including if you later operate them yourself.

Queries remain yours

The queries behind the managed processing live in your environment and are readable by your team.

Organizations and FAQ

Built to work with real organizations.

Data infrastructure rarely lives in isolation. We work with organizations of different sizes, and we can support the processes needed to evaluate and onboard an external data partner.

NDA processes
Security review
Vendor onboarding
Procurement coordination
Stakeholder alignment across analytics, marketing, product and engineering

Evaluating an external data partner?

Tell us what your security, legal and procurement teams need to see, and we will work through it with you rather than around it.

Talk to us about your requirements

A few practical questions

Before we get started.

Want to talk through your setup?

No. The technology includes a customer-facing dashboard, but the service itself is managed by Practical Vision. We configure, operate and maintain the data workflow with you.

New organizations start with a Data Audit.
Existing clients do not.

Talk to us

Already working with Practical Vision? You do not need the Data Audit. Request BQSync access.

Accessibility options

These options stay in this browser. They sit on top of the page's own markup and do not change how the site looks for anyone else.

Display
Motion and reading