From resistance to advocacy: driving data management adoption

Well done, you’ve got a strategy to scale your data management. The data landscape has been analyzed and prioritized and you are ready to start building the technical components to put that data management in place.

But here’s the catch: the technical elements and a value roadmap are just part of the foundation. The real magic? That happens when people across the organization are truly motivated and equipped to contribute.

In a world where data drives AI and analytics, data management has evolved far beyond centralized models. We’re talking decentralized, self-serve platforms—concepts like data mesh and data fabric are leading the charge, promising scalability and flexibility by giving data ownership to individual domains.

Regardless if you choose a fully decentralized model or end up choosing a different route to scalable data management, it will impact the day-to-day of your people. For example:

• From short-term fixes to scalable solutions: we aim to move away from developing point-to-point connections between teams/domains that are effective in the short run but difficult to maintain and that don’t scale. Instead, we encourage teams/domains to think holistically and develop products designed to serve a broad range of potential consumers, following Data Products Architectural principles.
• Making data discoverable and ready for scale: merely transforming and processing data is no longer sufficient. To prepare for scale, people must describe their data and make it easily discoverable, allowing potential consumers to locate and understand the data’s value and context.
• And there is more…

All this means change, and change is hard. So how do you encourage people to embrace data management and make it something they want to contribute to, not just something they have to use?

Let’s bring this question home with a concrete example:

Meet Data Scientist Sarah. Her organization is transitioning from a Central IT team managing data requests to a federated model. During the transition, a new ‘Data Catalog’  tool, designed for finding fully described data products, was launched, but Sarah was hesitant to use it. She felt it added complexity to her daily tasks and preferred turning to the Central IT team for her data needs, as she has always done.

Her perspective changed, however, when Alex, a trusted colleague and early adopter, started sharing how the catalog significantly improved his workflow. Inspired by Alex's success, Sarah decided to give it a try. After attending engaging, hands-on (and fun!) training sessions, she realized the catalog’s potential. Aided by the data management accelerator team, she started changing her way of working. The catalog saved her time, made data easier to find, and ensured accuracy.

As Sarah experimented, she saw further tangible improvements in her work, such as identifying data gaps and collaborating with data owners to enhance her analyses. She finally opened up to all the tools on the data management platform. Today, Sarah is not just a user but a passionate advocate who shares her success with others. An unstoppable feedback loop has been set in motion.

The question then is: "How do we make sure every initiative to scale data management has its Alexes for its Sarahs?"

Consider the innovation diffusion model (Rogers, 1971), where an innovation propagates from the innovators and early adopters to the late adopters and laggards. Alex typically falls among the early adopters—the enthusiasts who embrace and experiment with new technologies. Sarah belongs to the late majority, as does most of the organization, and requires more persuasion to get on board.

It's a common pitfall when transformations are rushed: skepticism and resistance take root, undermining adoption.

We suggest a two-stage approach, which promotes collaboration and creates a flywheel where everyone is equipped for - and eager to contribute to - data management:

• Phase 1: obtaining critical mass with the core transformation team. Develop your technical components and onboarding process to a point where the larger organization can get value from it.
• Phase 2: the hard part: scaling adoption across the organization. Make Data Management the new standard practice.

Let's look at these two stages in more detail.

Phase 1: Obtaining critical mass with the core transformation team

Critical mass is the goal here. If we open up new data management capabilities to the organization prematurely—without having robust technical capabilities or proven processes in place—we risk rejection before the initiative gets off the ground. It's a common pitfall when transformations are rushed: skepticism and resistance take root, undermining adoption.

To prevent this, we first need to lay a solid and tested foundation, both technical and organizational. Once this foundation is of sufficient maturity, we can achieve a critical mass: the point at which adoption gains momentum naturally, pulling the rest of the organization along with it.

So, how do we get there? Let’s break it down.

Next to a "Core Team" that develops the required capabilities to scale data management, start with a team of technically-skilled enthusiasts that apply these newly developed capabilities to realistic use cases. We refer to the latter group as the champions. This is where Alex is.

The champions work closely together on the platform, testing the newly developed capabilities and refining them with real-world feedback. The goal is to ensure that the tools, operating model, ways of working and trainings developed by the Core Team are validated by actual users, not just by technical experts. This reduces the risk of creating an engineering dream that does not meet business requirements. Value is the end-goal, technology is the means.

Value is the end-goal and a necessity - from the start - to keep senior stakeholders on board and make the transformation tangible for the rest of the organization. Without demonstrating value early on, the initiative risks losing senior stakeholder confidence during what would otherwise be an “underwater period” of capability development.

The champions showcase this value by delivering tangible results from their use cases, directly addressing the priorities of the project sponsors and senior stakeholders. Their success stories act as proof that the data management can deliver on its promise. Once other frontrunners start to see its potential and want to join in on the action (like Sarah did), you might be approaching that critical mass. Get ready for phase 2.

Phase 2: Scaling Adoption across the organization

Now, the real test: scaling data management across the organization. This goes beyond foundational work aimed at a handful of enthusiasts. Instead, it’s where organizational change takes centre stage. To succeed, you need to move beyond the early wins and convince the majority to embrace this new way of working. 

This transition is often more difficult than it seems, but with a carefully crafted strategy, it can be done. The goal is to create a flywheel effect, where early wins snowball into scalable, repeatable practices that become the new standard across the organization.

How can we achieve that? It’s another two-tiered approach:

1. Make data management something people want to contribute to
2. Empower people to actually contribute

Let’s dive in.

Make data management something people want to contribute to

A successful data management model inspires participation. It’s not enough for anyone working with data to simply understand the benefits of scalable data practices—they should actually enjoy and value the process. Here’s how to foster this:

1. Facilitate trainings that are actually fun. Make learning about data management engaging. Offer interactive, hands-on sessions that highlight real-world applications tailored to one's role. When people see how data management can make their work easier, it feels less like a chore and more like a benefit.
2. Appoint trusted role models. Nothing motivates like seeing the colleague you value most succeed. Identify and empower front-runners who can demonstrate the value of data management in their roles, creating a wave of enthusiasm across teams.
3. Ensure small, tangible wins. Start with gradual, incremental changes that solve real pain points. When employees see their daily tasks getting easier (and less painful) through small adjustments in data handling, they’re more likely to embrace the shift. 

In a well-designed data management model, people want to partake because they see its impact and enjoy the process. It’s the difference between pushing and pulling—when employees feel the benefits directly, they’ll naturally gravitate towards the new model.

Empower people to actually contribute

Creating a desire to participate is one thing, but equipping people with the skills and confidence to do so is just as important. Here’s how to make sure your organization can dive into data management with ease:

1. Make trainings ongoing and supportive. Go beyond the basics with deeper, targeted training sessions that build confidence. Training isn’t a one-off event—it’s an ongoing investment in your team’s capabilities.
2. Put accelerator teams in place for real-time help. Ad hoc support is essential. With dedicated accelerator teams ready to jump in on a case-by-case basis, employees can get the help they need whenever they need it, building confidence in using the tools independently - maybe you keep some champions on for this.
3. Create self-serve tools that grow with the organization. A good self-service platform is intuitive and designed to scale alongside your organization’s needs. As employees’ skills grow, the platform can support more complex use cases, creating a synergy between user capability and platform functionality.
4. Ensure clear onboarding pathways. Structured onboarding gives employees a roadmap for their journey in data management. Start simple, build up, and introduce new capabilities over time as confidence grows. This approach minimizes overwhelming feelings and encourages engagement.

When your data management system empowers users on all levels, it propels a flywheel of improvements. As people contribute, their skills grow, which further enhances the value and usability of the data products they’re working with, which in turn drives people to contribute more. The flywheel!

Wrapping it up

Data management isn’t just about setting up the right infrastructure and assigning roles—it’s about creating a culture that values data as an organizational asset. By inspiring your team to engage with data management and making it easy for them to contribute, you’re setting up a sustainable, scalable model that grows with your organization's needs.

In the end, it’s about more than just a strategy or a platform; it’s about building a community of data champions within your organization. When data management becomes something that people enjoy, trust, and understand, it transforms Data from practically every organization's frustration into a powerful ingredient for success.

---

This article was written by Job Hölscher, Data Scientist and Freek Gulden, Lead Data Engineer at Rewire.