OBOS BBL unifies analytics with Fabric and cuts costs by 20%

When housing demand turns uncertain, data credibility becomes nonnegotiable

Across Europe’s residential construction sector, a familiar playbook has stopped working. For years, steady demand masked inefficiencies in how builders planned, priced, and paced development. Homes sold before completion. Inventory cycles were forgiving. Reporting delays were tolerated because market momentum absorbed the risk. That context no longer holds.

Rising interest rates, economic pressure, and shifting buyer behavior have tightened margins across the Nordics. Construction timelines stretch years into the future, while demand volatility continues to be felt quarter to quarter. Unsold finished inventory ties up capital. Mispriced developments propagate loss at scale. Regulatory scrutiny around housing models, financing structures, and member equity increases the cost of being wrong.

OBOS BBL (OBOS), one of the Nordic region’s leading residential developers and cooperative housing organizations, sits squarely inside this pressure. Beyond its core business of developing and selling new homes, OBOS manages housing cooperatives, operates real estate brokerage and property management entities, and offers banking and financial services linked to home purchases and savings. Its OBOS Eiendom subsidiary owns and manages commercial properties, which are often located within OBOS residential projects.

Due to its size and scope, the company’s decisions influence not only its own performance but broader market dynamics. Hilde Marie Rustad, Head of IT Group Functions at OBOS, notes, “As one of the Nordic region’s leading homebuilders, the volumes we deliver influence prices, policies, and expectations across our markets, so our data must be precise and reliable.”

Frustrated by fragmentation

OBOS’s data lived across business units with different operational rhythms. Demand planning and pricing, marketing, finance, and member engagement all relied on analytics, but data lived across multiple platforms and business-unit pipelines. OBOS’s systems were no longer delivering the speed or consistency the business required. 

As time-to-insight stretched from days into weeks or months, teams compensated in predictable ways, building local models, exporting spreadsheets, and creating multiple versions of truth. While understandable, this strategy slowed decisions, created operational friction, and increased risk in an increasingly volatile market.

OBOS faced a choice familiar across industries: continue shoring up a platform that had reached its limits or fundamentally rethink how analytics are delivered, governed, and trusted across the organization.

Choosing consolidation over coexistence

OBOS made a deliberate architectural decision: retire parallel systems and consolidate analytics, data engineering, reporting, and governance onto a single unified platform.

OBOS chose Microsoft Azure and Microsoft Fabric to create the foundation required to execute this shift at scale. Azure would provide a highly secure and scalable cloud foundation. Fabric would run as a software as a service (SaaS) data platform on Azure, unifying ingestion, data engineering, governance, and reporting under a consistent operating model.

Rather than lifting individual workloads opportunistically, OBOS sequenced the transition to preserve business continuity while enforcing a clear cutover strategy. Dag Erlend Berger, Product Lead for the Data Platform at OBOS, explains, “We deliberately ran the old and new environments in parallel, comparing row and column counts and validating business KPIs through live reports before we switched anything over. That rigorous validation mattered, because for a data platform like ours, trust is the product.”

This approach enabled OBOS to migrate data processing workloads from Azure Synapse Analytics to Spark in Fabric while maintaining confidence in day-to-day business reporting. Data engineers and platform owners compared results side by side across platforms, ensuring that Spark in Fabric produced equivalent outcomes before retiring Azure Synapse Analytics–based pipelines. 

The scope of the migration was significant, with OBOS transitioning 600-plus notebooks and pipelines while consolidating more than 4,500 data objects into 87 lakehouses in Fabric. Rather than optimizing individual pipelines in isolation, the team focused on migrating complete analytical workflows so that data ingestion, transformation, and consumption could be validated end-to-end on the new platform. 

This workload-level approach reduced the risk of partial parity and ensured that downstream analytics and reports continued to reflect trusted business metrics once Fabric became the system of record. 

Validation extended beyond technical correctness. In addition to schema-level checks, OBOS verified that live business KPIs rendered through reporting workloads remained consistent across both environments. This allowed business users to preview Fabric-based outputs in real operating contexts while platform teams monitored performance and stability.

By maintaining strict governance ownership at the platform level while allowing business domains to continue consuming analytics during the transition, OBOS was able to avoid central bottlenecks without reintroducing fragmentation.

As data processing workloads moved from Spark in Azure Synapse Analytics to Spark in Fabric, OBOS also simplified its capacity model. Spark execution moved to the managed analytics environment of Fabric. Reporting workloads transitioned from Power BI Premium P1 to Fabric reserved capacity, consolidating analytics and reporting consumption under a single capacity framework.

This consolidation allowed OBOS to remove redundant services, simplify platform operations, and align consumption more directly with usage patterns, laying the groundwork for continuing improvements in performance and cost efficiency.

A platform for progress

With the migration complete, OBOS reduced infrastructure complexity; enabled more than 450 OBOS users to use Fabric for faster, deeper insights; and moved to a capacity‑based pricing model that aligned costs more directly with business usage. Early results included a 20% reduction in operational costs by eliminating redundant services and removing idle or overprovisioned compute. OBOS also achieved 30% faster data processing, driven by the optimized execution paths, improved concurrency handling, and managed compute pools in Spark in Fabric.

By unifying ingestion, storage, processing, and consumption under one SaaS platform, governance aligns with the way teams work across business domains. Lineage, classification, and access controls are now applied consistently rather than service by service. Automated deployment pipelines and standardized workflows shortened feedback loops, and optimization replaced reactive firefighting. 

“For us, a big change was that we stopped fighting the platform. We could focus on improving data quality and performance rather than maintaining infrastructure,” explains Atle Vika Røen, Architect at OBOS.

Accelerating decision-making across the organization

More important than infrastructure improvements, however, are the ways faster and more reliable data changed how—and how quickly—business decisions are made. 

With governed, near real-time analytics available across the organization, pricing and sales teams now adjust unit prices and incentives weekly instead of monthly. Development teams can evaluate whether to start, pause, or accelerate construction phases in days rather than weeks. Finance leaders can quickly assess liquidity risk, unsold inventory exposure, and capital allocation, while marketing teams can reallocate campaign spend within one to two days based on real-time conversion data. OBOS’s property management units make quicker renovation and investment decisions using validated cost and member data.

Early indicators are promising, including a 25% increase in decision-making speed and verifiable reductions in long-running data pipelines. Yet OBOS’s leadership remains focused on the long view, recognizing that the solution’s true value will compound over time.

That perspective reflects a fundamental shift in expectations. OBOS now sees its analytics platform as a long-term differentiator—one designed to absorb increasing complexity rather than amplify it, even as regulatory scrutiny intensifies, market volatility persists, and the scale of data and decisions continues to grow.

A measured approach to its AI-enabled future

For OBOS, the story isn’t simply about the transformation completed. It’s about building a foundation strong enough to support decisions whose outcomes may only become visible years from now. In an industry where uncertainty has become the norm, that foundation may be the most strategic asset of all.

While OBOS’s AI initiatives remain deliberately secondary for the time being, Fabric is helping the company ensure those it does pursue will be grounded in high‑quality, governed data rather than ad hoc experimentation. Rustad reflects, “We often say we want to be AI-ready. But that starts with data quality and governance. If we can trust our data today, we can trust the decisions we make tomorrow, whatever the technology looks like then.”

Discover more about OBOS on Facebook, Instagram, LinkedIn, X/Twitter, and YouTube.