Microsoft has flipped the switch on the public preview of Azure Storage Discovery, a managed service that pulls tenant-wide Azure Blob Storage telemetry into one portal dashboard and lets teams question their storage estate in plain English through Azure Copilot. The tool is available now in select regions at no charge until September 30, 2025, after which billing is scheduled to begin on October 1, 2025. For enterprises drowning in hundreds of subscriptions and thousands of storage accounts, this marks a pivot from scripting piecemeal inventories to a centralized, analytics-driven view of capacity, activity, errors, and security posture.

What Azure Storage Discovery Is

Storage Discovery is Microsoft’s answer to the fragmented visibility that plagues large blob estates. It automatically discovers storage accounts within a defined workspace, then aggregates metrics and configuration data into interactive reports. A workspace can sweep up to one million storage accounts, making the tool viable for even the largest enterprise tenants.

The service surfaces four families of insights:
- Capacity analytics: trends, distributions, and top accounts by object count and total bytes.
- Activity metrics: transactions, ingress/egress patterns, and hourly or daily aggregates (SKU-dependent).
- Error telemetry: aggregated transaction error rates with anomaly detection.
- Configuration and security posture: encryption, redundancy, access settings pulled directly from resource configurations.

All of this lands in pre-built dashboards inside the Azure portal, filterable by region, redundancy, performance tier, and encryption type. The dashboards are alive within hours of workspace deployment, thanks to an automated backfill that injects up to 30 days of historical data.

Pricing and Retention: Free Now, Metered Later

Preview pricing is zero-cost for both the Free and Standard plans. The Free plan covers capacity and configuration insights with 15 days of backfill and retention. The Standard plan adds transaction, activity, error, and security-configuration insights, extends backfill to 30 days, and retains data for 18 months—long enough to analyze annual cycles and seasonal peaks. Microsoft has not yet published post-preview rates; any per-GB or per-account figures floating around early forums are speculative. The official billing date is October 1, 2025, and the company says detailed pricing will appear on the Azure Storage Discovery pricing page before then.

Copilot: Natural Language Meets Storage Data

Perhaps the most eye-catching piece is the integration with Azure Copilot. Instead of writing KQL queries or clicking through nested drill-downs, operators can type questions like “Which regions are adding the most storage?” or “Where are my costs growing fastest?” Copilot synthesizes answers that span multiple accounts and scopes, linking directly to the affected resources. Finance and product owners who never touch the Azure portal can get data-backed answers without scripting.

But Copilot’s outputs are only as reliable as the underlying telemetry and the clarity of the prompt. Microsoft’s own documentation and early-adopter guidance underscore that conversational insights should be validated with native logs or deterministic exports before anyone makes compliance or configuration changes. “Natural language answers are insight aids, not audit evidence,” one early reviewer noted.

How Workspaces and Scopes Work

Onboarding is deliberately lightweight. An administrator creates an Azure Storage Discovery workspace in a chosen subscription and region, then defines workspaceRoots—up to 100 subscriptions or resource groups that act as discovery boundaries. Inside the workspace, up to five named scopes can be created using ARM resource tags, enabling logical grouping by business unit, workload, or environment. The service requires Reader access on in-scope storage accounts; it checks permissions before ingestion begins.

Workspace metrics usually populate within hours, but the documentation cautions that resource changes can take up to 24 hours to reflect. During preview, only storage accounts that contain blobs are discovered; empty accounts may not appear.

What Early Adopters Are Saying

Microsoft’s announcement highlights Tesco and Willis Towers Watson as early preview users. Tesco’s Cloud Platform Engineering team used Storage Discovery to replace time-consuming PowerShell inventories with a “single-pane-of-glass” dashboard that answered questions like “how much data do we have, and where?” in near-real time. Lead Engineer Rhyan Waine said the tool helped his team “focus on what really matters and addressing our top consumers, as opposed to being snowed in under a mountain of data.”

Willis Towers Watson used the service to spot storage accounts with unusual growth rates, then tied those insights back to cost-control measures and automatic cleanups via Blob Lifecycle Management. Lead DevOps Engineer Darren Gipson described the reaction: “As soon as my team started using Storage Discovery, they were immediately impressed … Very quickly, they identified several storage accounts that were growing at an unusual rate.”

Practical Limitations to Watch

Despite the fanfare, Storage Discovery is still in public preview. Organizations should consider several constraints before wiring it into production governance:

  • Preview SLA gaps: The service has no production uptime guarantee; don’t make it the sole source of truth for compliance audits.
  • Regional availability: The preview is limited to select Azure regions; global deployments must check coverage.
  • Scope granularity: With a maximum of five named scopes per workspace, highly federated enterprises may need multiple workspaces, adding management overhead.
  • Data freshness: A lag of up to 24 hours for resource changes means real-time operational checks should still rely on Azure Monitor or storage analytics.
  • Multi-cloud blindness: Storage Discovery only sees Azure Blobs; organizations with multi-cloud storage will need complementary tools.
  • Pricing uncertainty: Until post-October-1 rates are published, large estates can’t accurately forecast costs. The free preview window is the perfect time to run pilots and gather per-account, per-object metrics for sensitivity analysis.

How to Pilot Azure Storage Discovery

For teams ready to kick the tires, a structured pilot helps surface value while controlling risk:

  1. Inventory your estate: Pick representative subscriptions that cover diverse configurations and regions.
  2. Verify RBAC: The principal deploying the workspace must have Reader access on all workspaceRoots.
  3. Design scopes with tags: Map ARM tags to business units or environments to enable meaningful segmentation.
  4. Start with Standard: Use the free Standard plan during preview to experience the full 18-month retention and activity insights.
  5. Cross-check Copilot: Validate every Copilot-denominated “cost hotspot” against Azure Monitor logs or native storage analytics before acting.
  6. Track object and account counts: Log the number of storage accounts and blobs ingested to build a pricing model once rate cards drop.
  7. Submit feedback: Use the official [email protected] alias to report gaps; preview input can sway roadmap priorities.

Where Storage Discovery Fits in the Market

Storage observability solutions fall into three camps: native cloud tools, third-party multi-cloud platforms, and homegrown scripts. Storage Discovery competes directly with the first category, leveraging deep Azure integration and the Copilot differentiator. Organizations with an Azure-first strategy will likely find the managed service attractive for consolidating FinOps, security, and capacity planning. Multi-cloud estates, however, will still need a normalization layer unless they supplement with a third-party aggregator.

Microsoft’s bet is that ease of onboarding, long retention windows, and conversational AI will entice customers away from script-reliant approaches. The 18-month retention in the Standard plan, for instance, allows year-over-year capacity planning that few ad-hoc pipelines replicate without significant data engineering.

What’s Next

No detailed public roadmap exists, but logical next steps include expanding regional coverage, adding support for Azure Files, Queues, and Tables, and deepening Copilot’s ability to propose—or even execute—remediations within enterprise guardrails. General availability with production SLAs and finalized pricing are the near-term milestones to watch.

Azure Storage Discovery is a genuine leap forward for Azure-centric teams tired of assembling storage intelligence from scraps. The combination of scale, historical analytics, and Copilot-driven natural language access could shrink the time it takes to answer “What’s growing, where, and at what cost?” from days to minutes. But the preview label and the pending price tag mean the next few months are for experimentation, not permanent architecture decisions. Plan a pilot, validate findings, and be ready to adjust when the meter turns on in October.