The pipeline is ours. The warehouse is yours.
We build it, run it and keep it current every day — inside a database your company owns.
Analytics, advertising and app-store data, extracted daily and modeled into documented tables inside a cloud project you own — a data engineering function without hiring one.
- Analytics
- Advertising
- App Stores
7 platforms across 3 categories. Connected read-only.
-
Raw tables
Landed as each platform exports them.
-
Data models
7 documented tables, rebuilt daily.
Refreshed daily, operated by us.
- Anomaly Detection
- Semantic Layer
- Email Digests
- Dashboards
Reading from your tables — never from a copy we keep.
Three things cross that line. Nothing else does.
The boundary is the whole arrangement, so it is worth being exact about what moves in each direction.
-
Going in — read-only, by design
We hold viewer access and nothing more.
Each platform is connected with the narrowest access that will return the data. The pipeline reads; it has no path to write. Nothing in your ad accounts, analytics properties or store listings is altered by connecting them.
-
Inside — we operate it, you hold it
The engine is the service; the tables are the asset.
Extraction, modeling, backfill and the daily refresh are ours to run and ours to fix. What that work produces lands in your own project under your own billing — which is why leaving costs you a supplier, not your history.
-
Coming out — your tools, your tables
Nothing you read is locked behind our login.
Reports, digests and alerts query the same tables your own analysts can query. Anything we can build on top of that warehouse, your team or a future supplier can build too — there is no proprietary store in the middle.
What flows in
The platforms that already hold your numbers, connected once and then read every day — quotas, timeouts and retries handled on our side of the line.
- Analytics
- Traffic, search and performance signals from the platforms that measure your site.
- Advertising
- Ad spend and results, brought next to the revenue they drive.
- App Stores
- Installs, ratings and listing performance for your mobile apps.
Missing the platform you care about? The catalog grows because customers ask for it. Browse all sources
Where it lands
Raw tables give you something auditable to check the modeling against. The data models give your team something queryable without rebuilding the same joins each time.
- BigQuery
- Your data models are built in your own BigQuery project — our access is read-only, and the tables are always yours.
- History without a rolling window
- Each platform is backfilled as far as it allows, and what lands in your warehouse stays there — no retention window quietly deleting last year.
- A different warehouse?
- Tell us which one — the roadmap is driven by requests.
The data models, as they land
- Content Groups
content_groups - Page Types
page_types - Long-tail Queries
long_tail_queries - New Queries
new_queries - Brand vs Non-brand Queries
brand_non_brand_queries - Query Intent
query_intent - Keyword Cannibalization
keyword_cannibalization
These are the shipped starting point, not the ceiling — models are added for what your reporting actually has to answer.
What you do with it
Four ways to read the same warehouse. None of them is a place your data has to move to first.
- Google TimesFM
Anomaly Detection
Notice the drop before it costs you revenue.
- Open Knowledge Format
Semantic Layer
What your data means is written down. Your team and the AI tools you use read it the same way.
- Metrics only
Email Digests
A daily, weekly or monthly summary, in your team’s inbox every morning.
- Data Studio
Dashboards
Live reports, ready on Data Studio (formerly Looker Studio). Nothing new for your team to learn.
How the work runs
Managed means the operational burden sits with us. There is no console for you to learn and no schedule for you to watch.
-
A discovery call
We map which platforms matter, what the reports have to answer, and which of the shipped data models fit before anything is built.
-
We build it
Access is granted, the pipeline and data models are stood up inside your project, and history is backfilled as far as each platform permits.
-
It runs, daily
Extraction and modeling run on schedule. Failures are ours to notice and ours to repair; you find out because the tables are current, not because we asked you to check.
Where it stops
A scope you don’t get belongs in the same conversation as the ones you do.
- We do not run your ad accounts, write your content or make campaign decisions.
- We do not resell you a seat in our own dashboard — the tables are in your project, on your billing.
- We do not claim certifications nobody has audited. What we claim is architectural: read-only access, and data that lands in a project you own.
Frequently asked questions
Anything still unclear is usually settled in one discovery call.
In a cloud project your company owns. Both the raw tables and the data models are written there — so if you stop working with us, the history stays in your warehouse rather than leaving with a vendor.
See the line drawn around your own data.
Give us read-only access and see your own warehouse modeled, current and yours in a live walk-through.
Free discovery call · Your data is always yours