TSA Exit Acceleration
Move the dashboards before the board notices.
Fixed-fee proposal in 48 hours. Senior team on day one. The first conversation is always free.
TSA Tableau and Power BI separation is the work of moving the Newco's reports, datasources, and refresh schedules out of the seller's BI estate without losing a single executive view. The Newco's TSA exit strategy has to handle BI with the same discipline as the underlying ERP and warehouse, because every report that breaks on cutover is a report a CFO or a sponsor will notice within a week. The buyer that maps every workbook and every dataset to a destination owner before cutover keeps the operating cadence intact. The buyer that does not rebuilds dashboards under deadline pressure for months.
A Tableau or Power BI estate consists of workbooks, datasources, semantic models, refresh schedules, gateways, and the user permissions that link them together. The seller's tenant typically holds reports for finance, sales, supply chain, HR, and operations. Some workbooks serve only the carved-out business. Others serve the seller's retained business and the Newco jointly. A meaningful share of workbooks were built by line managers and have no formal owner.
The carve-out has to address each workbook individually. A workbook that serves only the Newco can be exported, imported into the Newco's tenant, and repointed at the destination datasource. A shared workbook has to be cloned, simplified, and rebuilt to remove the seller's dimensions. The work has to be planned with the report owners on the Newco side because business logic sits inside calculated fields and report filters in ways that are not always visible from the metadata.
A pre-signing inventory of workbooks in scope, owners on the Newco side, and the dependencies between workbooks and datasources is the first deliverable. Without it the cutover wave order is a guess. The pattern overlaps with the broader TSA Snowflake and Databricks separation playbook because the data layer drives the BI layer.
Every Tableau or Power BI workbook points at one or more datasources. The datasource may be a published dataset on the BI server, a direct connection to a warehouse, or a flat file refreshed by a scheduled flow. Behind the connection sits a gateway. The gateway authenticates against the source and runs the refresh. When the warehouse migrates to the Newco's tenant, every datasource has to be repointed and every gateway has to be reconfigured.
The sequencing matters. If the warehouse carve-out completes first, the BI workbooks can be repointed during a parallel running window. If the BI cutover completes first, the workbooks continue pulling from the seller's warehouse temporarily and are repointed when the warehouse moves. Most Newcos prefer the first sequence because it reduces the number of cutovers that affect the same dashboard. The seller's gateway servers stay in place until the last workbook has migrated.
Refresh cadence is the second discipline. Daily, intraday, and on demand refreshes all run on schedules that the Newco needs to recreate in the destination tenant. A missed refresh produces a board pack that pulls stale numbers and the controller spends the morning explaining the variance. The cutover plan should treat the refresh schedule as part of the dataset itself.
Tableau published datasources and Power BI semantic models hold the calculated measures, hierarchies, and relationships that turn warehouse tables into the metrics the business actually uses. Net revenue, gross margin, working capital, and headcount all sit in semantic logic that may have taken years to refine. The carve-out has to preserve those definitions intact. A small change to a measure produces a different number on a board slide and the audit committee will ask why.
The discipline is to document every measure on the Newco workbooks before migration, sign each definition off with the metric owner, and reconcile the destination tenant against the source for at least one full reporting cycle. Tableau XML and Power BI model files can be exported and inspected. The work is tedious. It also catches differences that no testing approach short of cycle level reconciliation will catch.
Where the Newco moves to a new BI platform during the carve-out, the semantic layer has to be rebuilt. The work is heavier and the risk is higher. The choice between rebuilding on the existing platform under the Newco's tenant and migrating to a new platform should be made on the value creation case, not on a technology preference. The pattern overlaps with the broader carve-out business intelligence tools playbook.
Every workbook has a permission map. Users, groups, and row level security rules control who sees what. The Newco's user list will look different from the seller's. Identity will move to the Newco's directory under the carve-out. Permissions have to be remapped before the workbook can serve the right audience. A missed permission produces either a CFO who cannot see a report or a former seller manager who can.
Embeds are the harder problem. Tableau and Power BI dashboards are routinely embedded in operational portals, customer facing applications, and the seller's intranet. Each embed holds a reference to the source tenant. The Newco has to find every embed, redirect it to the destination tenant, and validate that the access tokens still work. A register of embeds maintained through the TSA period is the only defence against missed references.
Downstream API consumers also need attention. Some teams pull Tableau or Power BI data through the REST API into other applications. Those calls hold tenant identifiers that have to be updated at cutover. The auditor will ask which reports source from where and the register answers the question without delay.
Tableau is licensed by user role and Power BI is licensed by user and capacity. The seller typically holds an enterprise contract that covers the combined business at a discount. During the TSA the seller allocates a share of the licence cost and the capacity cost to the Newco. The allocation methodology needs to be transparent. Buyers that accept an opaque allocation often pay a share that does not reflect their actual consumption.
The Newco's own BI contract is procured during the TSA period. Some Newcos stay on the same platform under their own tenant. Some move to alternatives such as Looker, ThoughtSpot, or a finance led replacement built on the warehouse semantic layer. The choice depends on the level of investment in existing reports, the user count, and the value creation plan around analytics speed.
When the cutover completes the seller's invoice should drop to zero on the BI line. The pattern overlaps with the broader TSA invoice validation process playbook.
A clean Tableau or Power BI exit closes three records. The seller's tenant retains no active Newco workbooks, no active Newco datasources, and no active permissions for Newco users. The Newco's destination tenant holds the full operational view of the business. The cutover documentation supports the broader TSA exit certificate.
Open items, typically the long tail of legacy workbooks no one currently owns, are tracked under a short post-close services agreement with a hard end date. Most of those workbooks are quietly deprecated. A few are rebuilt by the new owner. The decision is made workbook by workbook with the business sponsor in the room.
Specialist support across the BI workstream is part of the TSA Exit Acceleration service when the milestone is at risk. The work coordinates with the Newco's CIO, CFO, the seller's BI administrators, and the report owners on both sides.
How buyers carve out data platforms during a TSA without breaking analytics.
Read the article →How buyers carve out FP and A models from the seller's Anaplan tenant.
Read the article →How buyers carve out identity and access during a TSA without breaking access.
Read the article →The 90-day governance, IT, finance, HR and procurement separation plan we run on live carve-outs. Get the playbook plus the bi-weekly Day One Letter — short, signal-heavy, buyer-side.
No spam. Unsubscribe in one click. · Read the overview first →
Fixed-fee proposal in 48 hours. Senior team on day one. The first conversation is always free.
Seven buyer-side moves to exit a Transition Services Agreement on time and below budget. The mark-up, the extension-fee curve, exit sequencing, and the 11-month calendar.
One tactic, one benchmark, or one pattern from a recent buyer-side engagement. Short. Signal heavy. Free.
Subscribe to The Day One Letter →Scope the data split tightly. Every table you over-migrate and every pipeline you leave on the seller is a line on next year's TSA invoice.