Prashanthi Kolluru is the founder of KloudPortal. Helping global capability centers (GCCs) hire product-ready engineering pods.
I’ve seen data migration projects begin with weeks of discussion about platforms, architectures and migration tools, while the people responsible for making the migration work are brought into the conversation much later.
That order is easy to understand. Technology decisions are tangible. Teams can compare platforms, estimate infrastructure costs and build migration plans around technical dependencies. The harder questions are less visible: Who owns the data today? Who understands the exceptions in the legacy environment? Who will validate the migrated data? Who makes the call when the business cannot agree on what “correct” looks like?
Those questions often determine whether a migration succeeds.
After working with organizations on data and technology modernization, I’ve become convinced that the platform is rarely the first decision that should be made. The first decision should be how the organization will bring its people, knowledge and accountability together around the migration.
Why A Migration Moves More Than Data
A data migration is often described as moving information from one system to another. That description misses much of the work.
Enterprise data carries business rules, historical decisions, ownership structures and exceptions that may never have been documented. A customer record may have a different meaning across two systems. A finance team may rely on a field that doesn’t appear important to the engineering team. A reporting process may depend on a manual reconciliation that nobody thought to include in the migration plan.
The platform can move the data. It cannot explain why those exceptions exist.
This is where experienced people become critical.
The people who work with the data every day often know things that aren’t captured in schemas, architecture diagrams or technical documentation. Leaving them out of early planning can turn a straightforward migration into a series of discoveries during testing and cut over.
The Cost Of Finding Out Too Late
There is a financial reason to take this seriously.
McKinsey’s research on cloud migrations found that organizations in its survey spent an average of 14% more than planned on migration each year, while 38% experienced delays of more than one quarter. The study also found that only 15% of the organizations surveyed were able to migrate more than 60% of their IT-hosting spend to the cloud within their planned timelines.
Those numbers are from a cloud-migration study rather than data migration alone, so I wouldn’t use them to predict the cost of a specific data program. But they illustrate a pattern I’ve seen repeatedly: Migration costs grow when organizations underestimate the coordination, decision-making and organizational work around the technology.
A platform does not create alignment. A migration plan does not create ownership. People do.
The People Who Know Where The Data Breaks
Before choosing a migration approach, I would bring together the people who understand the data from different perspectives: engineers, business owners, analysts and compliance teams.
Each sees a different part of the problem. Bringing these perspectives together early can prevent expensive decisions later. That is the reason I prefer joint ownership over a handoff model.
Why The Migration Team Needs A Clear Owner
Large migrations can easily become everyone’s responsibility and therefore nobody’s responsibility.
I’ve found that a dedicated leader with authority to make decisions is one of the simplest ways to avoid this. That person doesn’t need to make every technical decision. The role is to maintain priorities, resolve conflicts and make sure business and technology teams are working toward the same definition of success.
AWS’s migration guidance makes a similar recommendation: Establish a single-threaded leader with clear accountability and visible executive sponsorship, while involving application teams in validation and cut over decisions.
That structure becomes particularly important when a migration crosses business units. Without clear ownership, teams can spend weeks debating priorities, dependencies and exceptions while the migration schedule continues to move.
Don’t Treat Change Management As A Training Exercise
Training is necessary, but it isn’t the whole people strategy.
A migration can technically succeed and still create operational problems if employees don’t understand what is changing, why it matters or what they are expected to do differently.
The people who use the data need to know where it will come from, how definitions may change and what they should do when something doesn’t look right. The teams supporting the new environment need the skills and access required to operate it. Leaders need a mechanism for hearing concerns before they become cut over problems.
Choosing The Platform After You Understand The Work
This doesn’t mean platform decisions are unimportant. They are critical.
It means the platform should serve a migration strategy that has already established the business objectives, data ownership, dependencies, quality requirements and operating model.
I would start by asking five questions:
1. Who owns each critical data domain?
2. Which business processes depend on that data?
3. What knowledge exists only inside the current teams?
4. Who will validate the data before and after migration?
5. Who will operate the environment once the migration team moves on?
The answers can influence architecture, sequencing, tooling and even the migration approach itself.
The Real Measure Of A Successful Migration
I don’t consider a migration successful simply because the data arrived in the target platform. The more important test is what happens afterward.
Can people trust the data? Can teams find what they need? Are critical reports producing the expected results? Are business processes running without manual workarounds? Does the organization have the skills and ownership to operate the new environment?
Those are business questions, not platform questions.
For technology leaders planning a major migration, I would make one change to the usual starting point: Assemble the people who understand the business and the data before finalizing the technology plan.
The right platform can move data efficiently. The right people determine what should move, how it should be validated and whether the organization can actually use what arrives.
That is why, in my experience, the strongest migration strategies start with people and use technology to carry the plan forward.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?