Test Husky audiences against your CRM in 30 minutes¶
No NDA. No data exchange. No cost.
Husky's APAC + US mobile audience inside a Snowflake Data Clean Room — match against your CRM, conversion log, or seed audience and get an aggregate overlap number same-day. PII never crosses the boundary.
What you get¶
| Output | Example value | Meaning |
|---|---|---|
overlap_count |
234,012 |
Distinct users in both your list and Husky's audience |
overlap_pct |
23.4 % |
What share of your CRM Husky covers |
husky_size (subset) |
33,622,626 |
Husky audience size in your selected geographies |
buyer_size |
1,000,000 |
Your distinct identifier count |
Aggregate only. k-anonymity ≥ 100 enforced — queries that would return fewer than 100 distinct matched users are rejected before output.
What stays private¶
┌────────────────────────────────┬────────────────────────────────┐
│ You keep private │ Husky keeps private │
├────────────────────────────────┼────────────────────────────────┤
│ Raw CRM emails / MAIDs │ Per-user segment assignments │
│ Conversion event log │ Full identifier graph (4B+) │
│ Seed audience identifiers │ Behavioural feature matrix │
│ Customer PII of any kind │ Other buyers' overlap results │
└────────────────────────────────┴────────────────────────────────┘
Only aggregate counts cross the cleanroom boundary. Both sides' raw rows stay encrypted at rest and in motion inside Snowflake's TEE-protected confidential compute.
How it works (5 steps, ~30 min)¶
| # | Step | Who | Time |
|---|---|---|---|
| 1 | Scoping call — confirm question, geographies, ID type | Buyer + Husky | 15 min |
| 2 | Husky issues cleanroom invitation to your Snowflake account | Husky | 2 min |
| 3 | You accept the invitation, attach your hashed-MAID or hashed-email table (read-only) | Buyer | 5 min |
| 4 | You trigger the overlap_analysis template with your geography list |
Buyer | <1 min |
| 5 | Aggregate result lands in your account — review the numbers, decide if Husky fits | Buyer | 5 min |
Total elapsed: ~30 min from "let's test" to a number on the table. Compare to the typical 5-10 day NDA-then-offline cycle.
Prerequisites¶
You need:
- A Snowflake account on Standard Edition or higher (no add-on license required)
- A table containing hashed MAIDs or hashed emails (SHA-256), de-duplicated
- 15 minutes for a scoping call with Husky to confirm your question fits the template
You don't need:
- An NDA — Snowflake's cleanroom contract pre-vets the data isolation
- A DPA — aggregate-only output qualifies as non-personal data under GDPR / DPDP / PDPA
- A new vendor onboarding — Husky is already on Snowflake Marketplace
- Any spend on Husky data — overlap test is free (you pay only Snowflake compute, ~$10–60)
Compliance posture¶
| Regulation | Cleanroom output classification |
|---|---|
| GDPR (EU) | Aggregate-only, k≥100 → not "personal data" under Art. 4(1) |
| DPDP (India) | No PII transfer → "data sharing" exemption applies |
| PDPA (Singapore) | Husky operates as Singapore-licensed data exporter |
| CCPA (California) | Not "sale" or "share" — no personal info crosses |
| K-PIPA (South Korea) | Permitted under domestic data analytics safe harbour |
Husky is incorporated as Husky Technology Pte Limited (Singapore). Cross-border data flows are resolved at the architecture layer — buyer's data never leaves buyer's Snowflake account.
What's the catch?¶
There isn't a hidden one. Honest framing:
- You need a Snowflake account. If you're on AWS-only / Azure-only without Snowflake, our AWS Clean Rooms integration ships in Q3 2026; LiveRamp Safe Haven follows in 2027 H1.
- Templates are pre-published. Custom queries beyond
overlap_analysis,lookalike_build,attribution_liftrequire a Husky engineer to write and review the template — not same-day. - The aggregate output is the contract. If you want row-level matched identifiers, that's a paid commercial share, not a cleanroom output.
Next step — book the scoping call¶
Book: https://www.huskydata.io/contact?interest=cleanroom
Or email sales@huskydata.io with subject "30-min overlap test — [your company]". Include in the body:
- Geographies you care about (e.g. SGP / HKG / JPN / USA)
- Identifier type you'll attach (hashed MAID / hashed email)
- Approximate size of your test list (so we can size warehouse compute)
We confirm template fit and schedule within 24 hours.
About Husky Data¶
The APAC Mobile Audience Foundation. ~3.46B MAIDs across 13 APAC markets, ~2B US MAIDs, schema-identical cross-border. 295 IAB Audience Taxonomy 1.1 segments. Distributed via Eyeota (D&B), Snowflake Marketplace (live), AWS Data Exchange (Q3 2026). Singapore-licensed. https://www.huskydata.io
Document version: 2026-04-27 · For prospective buyers only · Not a formal contract