Migrate from IBM Cognos Analytics to Apache Superset

The count of active reports and packages is usually the largest driver, because every item that stays in scope must be inventoried, rebuilt, tested and signed off. Effort grows with the number of distinct outputs, not with the number of files in the content store. Before asking for a quote, pull usage logs from the Cognos content store and classify each item as used monthly, used occasionally, or unused. Retiring unused content and merging near-duplicate reports often reduces the work more than any technical choice. Ask the provider to price the inventory and the retirement decision as separate phases.

Framework Manager packages hold business logic, such as joins, calculated items, filters, security rules and prompts, and each piece must be recreated as Superset datasets, virtual datasets, metrics and row-level rules. Simple star schemas with few calculations map quickly, while packages with heavy conditional logic or many overlapping query subjects need careful translation and validation. To reduce this cost, document each package's calculations and prompts in a short spreadsheet before the engagement starts, and note which logic already exists in the source database as views. Moving logic into governed database views usually makes the Superset side simpler and easier to test.

Pixel-formatted and burst output is a separate workstream, because Superset builds interactive dashboards and charts rather than print-exact documents or bursting to thousands of recipients. Each report that needs exact layout, page breaks, or per-recipient distribution requires a different output path, such as a reporting tool, a document generator or a scheduled export pipeline, and each path needs its own build, testing and operations work. Count these reports separately during inventory. Ask any provider to quote them as a distinct line item, and confirm which recipients, file formats and delivery channels must be preserved.

Identity integration determines how users sign in and which groups map to which access, so it often adds time when the current setup uses custom namespaces or complex group nesting. Superset supports database, LDAP and OAuth2 authentication through Flask-AppBuilder, and it maps LDAP or OAuth groups to roles with AUTH_ROLES_MAPPING, so the work is mostly in defining the group model and testing it. Confirm with your identity team which directory is authoritative, how groups are maintained, and whether service accounts are needed for scheduled jobs. Settling these questions before design begins prevents rework.

Each database or warehouse connection adds drivers, network paths, credentials, query tuning and validation, and the cost rises with every additional source and with each one that uses a different dialect. Superset connects to any SQL datastore that has a Python DB-API driver and a SQLAlchemy dialect, so most connections are feasible, but each still needs a tested driver version and a service account with the right read permissions. List every data source referenced by the Cognos packages, note the owner and the refresh pattern for each, and identify any sources with restricted network access, because those often need firewall changes that lengthen the schedule.

Reconciliation is the process of proving that a rebuilt dashboard produces the same numbers as the Cognos report it replaces, and its cost depends on how many figures finance, compliance or operations rely on. Rigorous reconciliation means defining test cases, running both systems against the same period, investigating every variance and recording sign-off. Lighter reconciliation for exploratory dashboards takes far less time. Decide in advance which reports are regulated or financially material, since those need documented evidence, and ask providers how they record variances and approvals in the written scope.

Internal staff time is often the most underestimated cost, because report owners must confirm requirements, review rebuilt dashboards, approve retirements and sign off reconciliations, and these people have other jobs. Platform staff must also run the Superset infrastructure, which includes the metadata database, caching and worker components used for alerts and scheduled reports. Your team can often handle inventory, usage analysis and business sign-off, while an implementation partner handles configuration, integration, hardening and testing. DeployOpen plans, implements, integrates, hardens, backs up and documents self-hosted platforms in your environment, and it delivers fixed-scope engagements agreed in writing. Quotes follow from that written scope, and this page does not state a price.

One-time costs include the inventory, model translation, dashboard rebuilding, reconciliation, identity setup, documentation and staff training, while recurring costs include the infrastructure that runs Superset, database backups, upgrades, security patching and ongoing support. Superset has no license fee because it is Apache-2.0 licensed with no paid edition, so the recurring cost is the infrastructure and the people who operate it. Ask any provider to split the quote into these two groups and to state which upgrade and support activities are included. Upgrades require backing up the metadata database, running the documented database upgrade commands, testing in staging and checking the release notes for breaking changes.

No. Apache Superset is released under the Apache License 2.0 and is an Apache Software Foundation project with no paid edition, so the software itself carries no license fee. The costs that remain are the infrastructure to run it, the staff time to operate it, and the implementation work to rebuild your content. Cognos charges per user, so the comparison is between a recurring seat fee and a one-time migration plus recurring operating cost. Ask any vendor, including DeployOpen, to separate those components in a quote so you can compare them directly.

The cost is driven by the size and complexity of the estate, not by the software. The main components are the inventory of reports, packages, Framework Manager models and schedules; the rebuilding of models as Superset datasets and dashboards; the infrastructure for the metadata database, caching layer and workers; the handling of burst and pixel-formatted reports; testing with business owners; and ongoing operation. Costs rise with the number of reports you keep, the number of data sources, custom security rules and the volume of formatted output. Ask for a quote built from a written scope that lists each of these components separately.

The timeline depends on how much content you keep and how it is used, and no fixed duration applies to every estate. Inventory and retirement decisions usually come first, because every report you remove shortens the rebuild. Timelines then depend on the number of distinct data models, how many dashboards depend on them, how many scheduled outputs must be replaced, and how many rounds of business validation are needed. Parallel running with Cognos during validation extends the schedule. Ask for milestones and acceptance criteria in the written scope rather than a single end date.

Your team provides the knowledge that no export file contains. That includes the business owners who approve each report, the data owners who confirm which sources and fields are authoritative, the people who understand Framework Manager filters and security rules, and the platform staff who provide database access, network paths, identity integration and hosting capacity. Someone also needs to own the decision to retire content. Without these contributors, rebuilt dashboards tend to be validated late, and the scope grows. Confirm the named contributors and their time commitments before work starts.

Yes, a capable internal team can do it, provided it has skills in SQL, data modeling, Python-based deployment, PostgreSQL administration, identity configuration and report validation. The harder parts are usually not the software installation but the inventory, the rebuild of security and model logic, and the handling of formatted and scheduled outputs that Superset does not produce. Many teams use a partner for a defined slice, such as the platform build or the hardening, and keep content work in-house. DeployOpen plans, implements, hardens, documents and supports self-hosted platforms in your own environment, and works from a fixed scope agreed in writing.

Superset does not produce burst reports or pixel-perfect formatted documents of the kind Cognos generates from authored report layouts. Superset dashboards, alerts and reports cover many scheduled delivery needs, but a regulatory filing or a bank statement style output needs a separate home. Framework Manager packages do not transfer as-is, because Superset uses datasets rather than packages. Cognos-specific features such as certain authored report types, bursting by recipient and some scheduling patterns need a documented alternative. Confirm each output type in your inventory against Superset's documented capabilities for your version.

Superset needs a metadata database, which is tested with PostgreSQL and MySQL, and for alerts, reports, async queries and thumbnails, a caching layer and Celery workers, with Redis commonly used. Operators must back up the metadata database before each upgrade, run database upgrades and initialization steps, test in staging, and read the upgrade notes for breaking changes. The SECRET_KEY must be set and managed, and identity is configured through Flask-AppBuilder with OAuth2, LDAP and group-to-role mapping. Ask for a runbook covering these tasks, a monitoring plan and a defined patching cadence.