BigQuery IntegrationConnect BigQuery to KPI Tree and build the accountability layer your data is missing.
You have the warehouse. You have the data. What you do not have is a system that maps how metrics drive each other, assigns ownership, delivers action plans, and proves what worked. KPI Tree connects to BigQuery with a single project ID, runs one query per metric on a configurable schedule, caches results with a configurable TTL, and performs all comparisons and aggregations off-warehouse. Every query generates a deep link to the BigQuery console so you can inspect exactly what ran and what it cost.
Connected in under an hour
One connection field, three IAM roles, and a guided wizard that validates everything in real time. No agents, no data extraction, no firewall changes.
Choose your authentication method
KPI Tree supports two ways to authenticate. With the managed service account, KPI Tree provides its own GCP service account and you grant it access to your BigQuery datasets. No JSON key to manage. Alternatively, create your own GCP service account, download the JSON key, and provide it to KPI Tree. Either way, the permission surface is three predefined IAM roles.
Grant three IAM roles
Whether you use the managed service account or your own, grant three roles: BigQuery Data Viewer for read-only access to the datasets you specify, BigQuery Job User to run queries, and BigQuery Read Session User for efficient data retrieval via the Storage Read API. No custom roles, no admin access, no data export grants.
Provide your project ID and connect
Enter your GCP project ID. KPI Tree validates the connection in real time, confirming it can authenticate, reach your datasets, and run queries. From there, define metrics with SQL against your tables and views, map how they drive each other, assign ownership, and start closing the loop between data and action.
Your BigQuery investment, with an accountability layer on top
KPI Tree connects to BigQuery with minimal permissions, adds query caching and cost tracking, and then builds the layer no BI tool provides: cause-and-effect trees, ownership, action plans, and proof that actions worked.
Query caching with configurable TTL
Each metric query result is cached with a TTL you control. User interactions such as filtering, comparing periods, and drilling into drivers all hit the cache, not your warehouse. Your BigQuery costs become a function of how many metrics you track and how often they sync, not how many people look at them.
Deep links and cost tracking for every query
Every query KPI Tree runs generates a deep link to the BigQuery console so you can inspect the SQL, execution plan, and bytes processed. KPI Tree tracks bytes-processed metadata from every query, giving you full visibility into what the integration costs without digging through billing exports.
Two auth methods, zero stored passwords
Use KPI Tree's managed service account and skip JSON key management entirely, or provide your own service account key for full control. Both methods use GCP's standard IAM model. Your existing VPC Service Controls, audit logging, and organisation policies stay exactly as configured.
Two authentication methods. Pick the one that fits your team.
Not every team wants to manage service account keys. KPI Tree offers a managed service account option: KPI Tree provides its own GCP service account, and you simply grant it BigQuery access on the datasets you choose. No JSON key to download, rotate, or store. For teams that prefer full control, bring your own GCP service account and provide the JSON key. Both methods require the same three IAM roles and nothing more. Your security team can review the exact permission surface in minutes.
- Managed service account: grant access to KPI Tree's own GCP service account, no JSON key needed
- Custom service account: provide your own GCP service account JSON key for full control
- Both methods require only BigQuery Data Viewer, Job User, and Read Session User roles
- No custom roles, no admin access, no data export grants
Cause-and-effect trees that show how your business actually works.
Dashboards show you what happened. KPI Tree shows you why it happened, who is responsible, and what to do next. Map the causal relationships between the metrics your business runs on. When a number moves, trace the tree to understand which drivers changed. Assign RACI ownership so every metric has a named person accountable for it. Statistical correlation analysis surfaces relationships between metrics that manual inspection would miss. Impact tracking verifies whether actions actually moved the needle.
- Metric trees map cause-and-effect relationships between business measures
- RACI ownership assigns accountability to a named person at every node
- Statistical correlation analysis surfaces what actually drives each metric
- Impact tracking proves whether actions moved the metrics they targeted
Comparisons and aggregations run off-warehouse. Your bill stays flat.
KPI Tree runs one query per metric on a configurable schedule and caches the result. Comparison periods, correlations, regressions, outlier detection, and causal analysis all run in KPI Tree's compute engine without pushing additional queries to BigQuery. Adding users does not add warehouse queries. Every query includes bytes-processed metadata so you can track costs precisely, and a deep link to the BigQuery console so you can inspect exactly what ran.
- One query per metric on a configurable schedule, cached with configurable TTL
- All comparisons, aggregations, and analytics run off-warehouse
- Bytes-processed metadata tracked for every query
- Deep link to BigQuery console for every query job
Three IAM roles. Your security team will approve this quickly.
The entire permission surface is three predefined IAM roles. BigQuery Data Viewer provides read-only access to the datasets you specify. BigQuery Job User allows KPI Tree to run queries. BigQuery Read Session User enables efficient data retrieval via the Storage Read API. No custom roles, no broad project-level permissions, no data export grants. VPC Service Controls, Cloud Audit Logs, and organisation policies stay fully enforced. All credentials are encrypted at rest.
- BigQuery Data Viewer, BigQuery Job User, and BigQuery Read Session User
- No data export grants, metric results are processed in KPI Tree's engine
- VPC Service Controls and organisation policies fully enforced
- All credentials encrypted at rest
What KPI Tree adds on top of BigQuery
BigQuery stores and queries your data. KPI Tree adds the human layer: cause-and-effect trees, ownership, action plans, and proof that actions worked. That is the gap no BI tool fills.
Cause and effect, not just charts
Metric trees map how every KPI drives the ones above it. When revenue drops, trace the tree to find whether it was traffic, conversion, or average order value, and who owns each driver. That causal structure does not exist in any dashboard.
Ownership that creates accountability
Every metric has a named owner with full RACI. Actions are tracked against the metric they were meant to move. Personalised action plans tell each team member what to focus on. Impact is verified statistically, not self-reported.
Off-warehouse analytics that keep costs flat
One query per metric, cached with a configurable TTL. Comparisons, correlations, regressions, and outlier detection all run off-warehouse. Your BigQuery bill stays predictable as your team grows.
Related integrations
Other data sources that work with KPI Tree.
Common questions
- Just your GCP project ID. For authentication, either grant access to KPI Tree's managed service account (no JSON key required) or provide your own service account JSON key. The setup wizard validates the connection in real time.
- Three predefined roles: BigQuery Data Viewer (read-only access to the datasets you specify), BigQuery Job User (permission to run queries), and BigQuery Read Session User (efficient data retrieval via the Storage Read API). No custom roles, no admin access, no data export grants.
- With the managed service account, KPI Tree provides its own GCP service account. You grant it access to your BigQuery datasets and never handle a JSON key. With a custom service account, you create your own GCP service account, download the JSON key, and provide it to KPI Tree. Both methods use the same three IAM roles and the same permission surface.
- KPI Tree caches each metric query result with a TTL you configure. All user interactions, comparisons, aggregations, and analytics, are served from the cache rather than querying BigQuery again. You control how fresh the data needs to be, and your warehouse costs stay predictable.
- Yes. Every query generates a deep link to the BigQuery console where you can inspect the SQL, execution plan, and bytes processed. KPI Tree also tracks bytes-processed metadata from every query so you can monitor costs directly within the platform.
- KPI Tree runs one scheduled query per metric, cached with a configurable TTL. All comparisons, correlations, and analytics run off-warehouse. Your cost per metric is predictable and does not scale with the number of users. Most teams see lower overall BigQuery analytics spend because KPI Tree replaces the per-user query pattern of traditional BI tools.
- Yes. See our dedicated dbt Cloud and dbt Core integration pages for full details. One-click sync imports every metric, dimension, and time grain. Changes in dbt flow through to KPI Tree automatically.
- KPI Tree queries your warehouse and processes aggregated metric results in its own engine. Raw row-level data is not persisted outside your environment. All BigQuery governance, including IAM roles, VPC Service Controls, audit logging, and organisation policies, remains fully enforced.
Related guides
Deep dives into the frameworks and metrics that work with Google BigQuery.
Connect BigQuery in under an hour.
One project ID, three IAM roles, and your choice of auth method. Off-warehouse analytics, query caching, and full cost visibility from the start.
