Unlock the Data a Small Sensitive Slice Holds Back
C-level corner · Executive brief
Most data can go to a modern cloud platform with the platform's standard protections. A small slice cannot, or not without a stronger answer, and that slice often holds up everything else. The way out is not "trust us" versus "no cloud": it is deciding which one thing your organisation keeps in its own hands.
The business problem#
Regulated organisations want one modern analytics platform. But a small share of their data carries legal, regulatory or reputational risk that a standard cloud setup can’t fully answer: national ID numbers, bank details, salaries, patient identifiers, or datasets a regulator says may not be stored in a foreign-operated cloud at all.
That slice is small, and it often blocks the whole programme. Projects stall in legal review, data stays in silos, and the value of the other ninety-odd percent goes unrealised. The question that ends most of these conversations is always the same: “If we put this in your platform, what is still in our hands?”
The pressure comes from several directions at once. Privacy law limits where personal data may go. Financial regulators expect some records to stay under domestic jurisdiction. Defence and healthcare have their own rules, and manufacturers often want process data on infrastructure they fully control.
What changes#
For the sensitive slice, the organisation keeps one decisive control, and the platform does the rest:
- The data. The most sensitive records stay on your premises. The platform can reach them only when you allow it, for a limited time, or never.
- The key. The data sits in the platform encrypted. Your organisation holds the key and decides, person by person and query by query, when it is used.
- The switch. You can cut the platform’s access instantly, with a time limit and a full audit trail.
The bulk of the data keeps using the platform’s standard controls. The stronger patterns apply only to the columns and datasets that need them.
What it’s worth#
- Revenue and value from all your data. Sensitive and non-sensitive data can be analysed together, with the tools your teams already use, instead of the sensitive part sitting in a silo.
- Speed through review. Regulators, boards and customers’ legal teams get a concrete, demonstrable answer to “what is still under our control?”, not a contractual promise.
- Lower risk. Depending on the pattern, this protects against the platform operator, over-privileged administrators, stolen credentials and legal compulsion against the provider.
- Accountability by design. Releasing the most sensitive data needs a named person, a second-factor confirmation and a written reason, and every release is logged.
- Proportionate cost. In the technical article’s benchmarks, the extra controls added between a fraction of a second and a few seconds per query, and only where they apply.
What it doesn’t do#
No pattern stops an authorised person from copying what they are allowed to see. That remains a matter of roles, policy and monitoring. And “the key stays with us” is a different requirement from “the data stays with us”: requirements documents often blur the two, and the right answer depends on which one your regulator actually means.
Decisions for leadership#
- Name the slice. Which datasets and columns need a stronger answer than the platform’s standard controls? Usually it is a short list.
- Decide what you keep for each item: the data, the key or the switch. Get legal and compliance to say which requirement applies before engineering picks a design.
- Let everything else move. Don’t let the slice hold the rest of the programme hostage.
- Make the answer demonstrable. Ask for something you can show an auditor or a customer, not a slide.
Questions to ask your team#
- Which data is blocking our platform decisions today, and how much of the total is it?
- Does the requirement say the data must stay with us, or only the key?
- If we had to cut the platform’s access to our most sensitive data tonight, how would we do it, and how would we prove it happened?
- Who can release sensitive values today, and is every release logged with a reason?
Technical background, for your architects
- The Data, the Key, or the Switch: the one-page framework with the decision questions.
- The Data, the Key, or the Switch: Customer-Held Controls for Data Sovereignty in Snowflake: eight runnable patterns with measured latency and the threat model. Snowflake-specific.
- Data Sovereignty First: Integrating an On-Premise Apache Iceberg into Snowflake Through a Direct Outbound Tunnel: keeping data on premises while the platform queries it on demand. Snowflake-specific.
Figures come from the linked articles; check them against your own situation. Snowflake-related articles explore the art of the possible and are not official Snowflake solutions or endorsed by Snowflake.