Salesforce’s Data Integration and Automation Services Modernizes in Massive Tableau Cloud Migration

Modernizing with Tableau Cloud to support innovation and future AI solutions

Beyond the goal of using platforms before customers do–in order to ensure trust, stability, and maturity in products–the migration supported Salesforce’s goals for facilitating faster analytics innovation and adoption at scale. This strategic modernization was also aligned with the company’s cloud-first direction and data center optimizations. It was the DET team’s opportunity to show what a great migration could look like for any organization.

There were clear benefits to moving onto Tableau Cloud, which include simplified identity management, centralized site administration, and elastic capacity for workloads like background processing, all of which were key to enabling Salesforce to scale analytics with confidence. Throughout the process, the team worked closely with the Tableau Engineering organization to influence platform direction, shape Tableau Migration SDK capabilities, create reusable tools and strategies, and more–ultimately benefitting all migrations to come, whether internally at Salesforce or externally with customers.

This modernization aligned perfectly with the organization's vision of democratizing analytics while keeping pace with the broader Salesforce technology stack and paving the way for an AI future.

Technical considerations

At the start, during their evaluation and scoping of this migration effort, the DET team faced significant architectural and performance considerations due to the immense scale of the Salesforce deployment. They uncovered limitations around postgres access, virtual connections, and custom views, collections, and favorites, and pushed the boundaries of what is currently possible on Tableau Cloud around:

• User limits on a single site
• Group synching
• Backgrounder capacity
• Storage needs
• Timeouts and extract strategy

For example, one technical challenge the teams faced is the fact that Tableau Cloud enforces a strict 2-hour timeout on extract jobs. This timeout required extensive refactoring or replacement of many long-running and large extracts–some 15-20 GB, even up to 100 GB–often solving the challenge by moving them up to the Database, Tableau Bridge, or Prep flows.

To achieve the goal of becoming the best, first users of their products, this migration from Tableau Server to Tableau Cloud was more than just a technical upgrade for Salesforce: it was a strategic imperative.

Business drivers for the migration to Tableau Cloud

When the project began, Tableau usage at Salesforce was growing rapidly but fragmented with multiple on-premise clusters, decentralized governance, and infrastructure nearing capacity. They saw an opportunity to centralize, modernize, and scale in a way that would support long-term innovation and future implementation of AI solutions like Tableau Next and Agentforce.

Several key business drivers made the case for migration:

• Customer Zero and internal adoption goals - As part of Salesforce’s Customer Zero vision, the DET strived to use their own platforms at scale before customers. While the DET team had internal Tableau Server environments, they were limited in reach and didn’t reflect the breadth or depth of enterprise adoption they aspired to.
• Data center strategy and infrastructure rationalization - Already in the process of migrating data centers, the DET team evaluated whether platforms should be modernized or retired. By choosing to transition to Tableau Cloud, they made architecture investments and strategic decisions for the future, reducing overhead and unlocking elastic scale without long-term hardware commitments.
• Faster innovation and feature velocity - Being on Tableau Cloud means gaining faster access to innovation—such as AI-driven insights, improved governance tools, and performance improvements—without the lag and risk of manual upgrades.
• Enterprise scale and simplified operations - With over 80,000 licensed users and tens of thousands of content assets, the Tableau footprint had outgrown what traditional server-based operations could efficiently support. 

These drivers were more than just technical considerations—they were foundational to how Salesforce delivers analytics, supports its internal customers, and models modern cloud transformation. The decision to migrate was about more than Tableau. It was about shaping the future of analytics at enterprise scale.

Developing custom migration solutions and influencing Tableau Cloud growth

Salesforce's DET team partnered deeply with Tableau engineering throughout the migration, influencing platform direction and shaping the capabilities of the Tableau Migration SDK. This collaboration was crucial, as Salesforce's status as a Customer Zero deployment meant their use cases helped uncover edge cases, drive roadmap priorities, and improve SDK capabilities for all future customers.

Initially, the migration adopted a user-led approach where the DET team provisioned projects and permissions, and project leaders migrated content using native Tableau tooling like the Tableau Content Migration Tool (CMT) or Tableau Desktop and Prep. But this proved too slow and inconsistent at scale. So, two months into the project, Salesforce pivoted to a centralized, admin-driven model, powered by the Tableau Migration SDK. The DET team took ownership of content movement using the SDK, dramatically improving consistency, speed, and quality control. They also built custom scripts around the SDK to automate the movement of unsupported artifacts like favorites, custom views, flows, and schedules. This shift was a turning point, enabling the DET team to deliver migrations at scale while minimizing friction for business users.

Originally, the DET team considered separate sites to distribute load across DEP, BTM, and ALPO environments. But following joint analysis with Tableau, they determined that consolidating into a single, top-level Tableau Cloud site was feasible–provided there was sufficient capacity provisioning, especially for backgrounders. The combined environments regularly demanded 50+ concurrent backgrounder jobs, with peaks nearing 75, far above Tableau Cloud’s published default of 25. This had been a major point of negotiation and planning, as backgrounder capacity directly impacts refresh SLAs and overall system responsiveness. By consolidating into a single top-level site, this helped simplify governance, enabled more consistent access policies, and reduced administrative overhead.

Regarding storage needs, Tableau worked with Salesforce to provision more than the standard storage limits, accommodating over 5 TB. Concerns about user scale were also addressed, because Salesforce approached the 100,000 user limit (on a single site), agreeing to roadmap commitments to accommodate future growth. 

Furthermore, Tableau Cloud’s VizQL elasticity and linear scaling model proved sufficient to meet Salesforce’s performance demands, once they were properly configured with the right resource blocks and infrastructure. For example, teams required sub-5-second load times for 80% of requests, and under 10 seconds for 90% of all requests—even during peak periods when concurrent usage spikes to 40,000+.