During a recent conversation on The Executive Outlook, Rebecca Alexander shared a truth that many data leaders learn only after building dashboards, pipelines, and models that nobody fully uses.
A data strategy is not successful because the data team built it.
It is successful when the business trusts it, uses it, and makes better decisions because of it.
Rebecca’s own journey into data did not begin inside a startup, a dashboard project, or a gaming company. It began in science. She came to France because of her passion for physics and mathematics and completed a PhD in nuclear physics, where she studied the effect of radiation on different kinds of materials at a nanoscale level.
During that time, she realized something important about herself. What excited her most was the process of analyzing data, finding patterns, and using evidence to reach conclusions. In one research moment, an experiment in Japan helped validate what her team had simulated theoretically. That feeling of seeing analysis connect with reality stayed with her.
She wanted to experience that kind of impact more often.
In industry, data gave her that opportunity. Instead of waiting for rare experiments to validate years of theoretical work, she could see the impact of analysis on a daily, weekly, and monthly basis. That is what eventually led her from nuclear physics into data, mobile gaming, and data strategy leadership.
Today, as Head of Data at Oh BiBi, Rebecca brings a rare combination of scientific thinking and practical business understanding to one of the fastest moving environments in technology.
Her message is clear.
Data strategy must serve the business, not just the data team.
Rebecca describes building data infrastructure in a startup with a metaphor that captures the pressure perfectly.
It is like building a staircase while climbing it.
You know where you want to go, but the structure that will take you there is still being built. The company needs answers now. The business needs speed now. Teams are making decisions now. But the data infrastructure that will make those decisions scalable, reliable, and repeatable takes time to build.
That is the tension every startup data leader understands.
Move too slowly, and the business loses patience. Move too quickly, and the foundation becomes fragile. Rebecca believes the answer is not choosing one side. It is learning how to balance short term value with long term capability.
In a startup, leaders must keep showing what data is doing for the business today while also explaining why certain investments will pay off later. That may involve cleaner data, faster access, better self service, stronger infrastructure, and less dependence on analysts as gatekeepers.
But none of that works without communication.
Rebecca is clear that communication matters when things are going well. But it matters even more when things are not going according to plan. When assumptions change, when timelines shift, or when unexpected complexity appears, business stakeholders should not be left guessing.
For her, data strategy should never be a private conversation inside the data team. It should be something the whole company understands, especially the stakeholders who will make decisions from the infrastructure being built.
One of Rebecca’s strongest examples came from a situation many organizations will recognize.
The company had dashboards. The data team had built them. The numbers were there. But business stakeholders were not using them to make decisions.
They were still relying on gut feeling.
Instead of assuming the business was simply not data mature enough, Rebecca went closer to the problem. She spoke with stakeholders to understand what was stopping them from using the dashboards. What she discovered was not a technical failure. It was a language failure.
Some dashboards were outdated compared to the work teams were actually doing. Others did not speak the language of the business function they were supposed to support.
One example was especially clear. The dashboard showed gross revenue, but the finance team needed net revenue. Because the metric did not match how finance made decisions, the dashboard was abandoned.
That single mismatch revealed a deeper lesson.
A dashboard can be technically correct and still be useless if it does not match the decision language of the team using it.
Rebecca and her team responded by running specific workshops with finance, product, and marketing. They asked each team what decisions they made daily and weekly, what KPIs they cared about, and what language they used when discussing performance.
The dashboards were then rebuilt around those real decision patterns.
The impact was measurable. After the dashboard migration, usage increased four times. More importantly, dashboards started appearing inside business meetings. Teams opened them during discussions. They used them when presenting decisions to senior leaders. They stopped treating dashboards as reporting artifacts and started using them as decision tools.
That is what a strong data strategy framework should create.
Not more dashboards. Better decisions.
Rebecca also shared how she translates data strategy into work that engineers, analysts, and machine learning specialists can actually execute.
Her approach begins with alignment. When there is a strategy or roadmap, the data team is involved so they understand what the company is working toward. The work is not handed down as disconnected tasks. It is connected to the bigger picture.
She also tries to align projects with individual growth goals. If an analyst wants to grow in data engineering, Rebecca looks for a project that helps them build that skill. If a machine learning engineer can contribute to a data engineering or analytics project, she looks for ways to make that possible.
This matters because motivation and ownership are easier to build when people can see how the work supports both the company and their own development.
From there, she makes the work specific. Team members create the first draft of deliverables. Then the discussion becomes practical. What should the dashboard contain? Which filters matter? Which charts are needed? What is the expected outcome? What exactly does this project mean?
The roadmap then moves into two week sprints.
Rebecca also tries to reserve part of the team’s weekly time for long term projects. In theory, this might be around one day a week. In reality, startup life brings what she calls phantom tasks. Unexpected urgent work appears. Some weeks the long term focus may shrink. Other weeks it may expand.
The important part is that progress does not disappear completely.
Even small weekly progress keeps the long term strategy alive. That is how a data strategy moves from vision to execution without being buried under urgent requests.
Rebecca’s gaming experience adds another layer to her data strategy perspective.
In gaming, machine learning is not just about building advanced models. It is about improving player experience, retention, and monetization in ways that feel relevant to the user.
One example is personalization. A game can adjust the player experience based on level, behavior, and engagement. In multiplayer games, matchmaking algorithms can help ensure players are matched with others at a similar level. If the match is too easy, players lose interest. If it is too difficult, they become frustrated. The goal is to keep the experience challenging without making it discouraging.
Machine learning can also support monetization.
For games with in app purchases, offers can be personalized based on spending behavior. For games monetized through ads, ad frequency can be adjusted based on user behavior and churn risk. If a user can tolerate more ads without leaving, the game may increase ad exposure. If a user shows signs of churn, ad frequency may be reduced to protect retention.
The important point is that optimization depends on the business goal. Sometimes the priority is retention. Sometimes it is monetization. Sometimes it is balancing both.
A good data strategy in gaming must understand that context before applying machine learning.
Gaming data comes with scale. Millions of users can create huge volumes of events, revenue data, engagement signals, and behavioral patterns. That creates opportunities, but it also creates risk.
Rebecca’s approach to data quality begins with first principles.
Where can quality be lost?
For her team, transformation is one of the key points. They use DBT modeling and apply data quality checks at each transformation step to make sure data is not lost or distorted as it moves through the pipeline.
They also use monitoring dashboards to compare KPIs against the source of truth. If a KPI crosses an acceptable threshold, the team receives alerts through Slack and investigates.
But Rebecca makes an important distinction. Not every discrepancy is a crisis.
For example, revenue data may naturally have around five percent discrepancy because of currency exchange rates. The exchange rate at the time of purchase may differ from a daily exchange rate used later in reporting. In that case, demanding one hundred percent matching data all the time would create unnecessary noise.
The team accepts a threshold, monitors it, and investigates only when the difference crosses the expected range.
This is a mature approach to data quality strategy. It avoids constant firefighting while still protecting trust. It also recognizes that quality is not only about perfection. It is about knowing what level of variation is normal, what level is risky, and when the business needs to act.
Rebecca’s team also monitors major KPI movements. If organic installs are usually around fifty thousand a day and suddenly fall to five thousand, that is a signal that something needs immediate attention.
The goal is simple.
Keep data reliable enough for people to trust it without turning every small difference into an emergency.
One of the strongest stories from Rebecca’s conversation involved an A/B test in a game.
Two versions were tested. One version clearly performed better than the control. The dashboard showed it. The data supported it. But the result felt counterintuitive to the team.
Some people were ready to trust the data. Others were hesitant. They did not want to make a decision only because the dashboard said one version was winning.
Rebecca’s team did not respond by pushing the numbers harder. They went deeper.
They analyzed player behavior to understand why the winning version was performing better. They checked whether the result was consistent, whether it was being driven by outliers, and whether the dashboard was reflecting real behavior.
Even after that, some hesitation remained.
So Rebecca reframed the analysis in the language that mattered to the stakeholder. One of the unconvinced team members had a strong finance background. Instead of presenting only game metrics, Rebecca’s team broke the result into revenue, retention, potential risk, and possible loss.
That changed the conversation.
The data did not change. The way it was communicated changed.
Once the decision was explained in terms the stakeholder could connect with, the team reached alignment and made a data driven decision.
For Rebecca, this became one of the clearest lessons in data leadership. It is not enough for the analysis to be correct. It has to be understood by the people who need to act on it.
Across the full conversation, Rebecca Alexander is not presenting data strategy as a document, a roadmap, or a set of dashboards.
She is presenting it as a business discipline.
A strong data strategy connects infrastructure with decision making. It connects machine learning with player experience. It connects data quality with trust. It connects dashboards with the language of the teams who use them. And above all, it connects the data team with the business instead of leaving them to operate in separate worlds.
Her journey from nuclear physics to gaming data leadership makes this perspective even more valuable. She understands the analytical depth required to work with complex data. But she also understands that analysis creates value only when it changes how people decide.
That is the real test of data strategy.
Not whether the data team built something.
Whether the business used it.
Whether the dashboards entered the meeting room.
Whether stakeholders trusted the numbers.
Whether machine learning improved the experience.
Whether data quality created confidence instead of noise.
Whether communication turned resistance into alignment.
Rebecca’s message to data leaders is practical and clear. Build for the business. Speak in the language of your stakeholders. Keep the long term vision alive, even when urgent work appears. And remember that the success of a data strategy is not measured by what is delivered to the business, but by what the business can finally do because of it.
Want to hear more conversations with leaders shaping data strategy, gaming analytics, and business decision making? Explore more on The Executive Outlook.
Listen in for business strategies and interviews with business experts to help you build a business and a life that you love.
About Us
