WTF: Why That’s Fantastic
There is something wonderfully meta about being asked to develop a strategy for communicating the work of an office devoted to Data Science Strategy.
Strategy for strategy.
That was essentially the assignment when I became the first Creative Director brought in to support communications for the National Institutes of Health’s Office of Data Science Strategy — ODSS.
The office itself was already doing extraordinarily consequential work. NIH was confronting an explosion in the amount, complexity, and diversity of biomedical and behavioral data available to researchers. Artificial intelligence was advancing rapidly. Data sharing was becoming increasingly important. Biomedical information was being generated across laboratories, health systems, repositories, wearable devices, clinical studies, and countless other sources.
ODSS had the responsibility of helping NIH determine how all of that data — and the tools, technology, policies, infrastructure, and people surrounding it — could be used more effectively to accelerate discovery and improve human health.
The strategy was substantial.
Our challenge was to make it understandable.
Fortunately, ODSS was well underway in developing what would become the NIH Strategic Plan for Data Science, 2025–2030. That gave us the strategic substance we needed.
What it needed next was narrative structure, creative direction, visual hierarchy, and a communications approach capable of turning an enormously complex institutional strategy into something people could actually navigate, understand, remember, and use.
In other words:
The scientists and data strategists had to design the future of data science.
We had to design the story of that future.
That is what Data Science By Design is really about.
Introduction: What “Data Science By Design” Really Means
Data Science By Design is not about making data look prettier.
It is about recognizing that some of the most complicated ideas in an organization become significantly more useful when they are intentionally structured, translated, visualized, and communicated.
The NIH Strategic Plan for Data Science is a strong example.
The 2025–2030 plan establishes a broad vision for how NIH can leverage data science to accelerate understanding of human health. It responds to an evolving technology landscape, dramatic increases in the quantity and diversity of research data, and the need to connect newly generated information with existing knowledge. The Office of Data Science Strategy leads implementation in coordination with NIH Institutes, Centers, and Offices and partners across government, academia, industry, and philanthropy.
That is an enormous strategic mandate. And enormous mandates create an equally enormous communications challenge. Because a strategy that only makes sense to the people who wrote it has limited value.
A strategy begins creating organizational impact when people outside the planning room can see where it is going, understand how the pieces connect, identify where they fit, and act accordingly.
That transition — from information to understanding — does not happen automatically. It has to be designed.
1. The First Challenge Wasn’t Data. It Was Meaning.
Data science itself can become intimidating almost immediately.
Metadata.
Machine learning.
Artificial intelligence.
Data repositories.
Interoperability.
Common Data Elements.
Computational infrastructure.
Privacy-preserving technologies.
FAIR principles.
Federated data systems.
Put enough of those terms onto a page and even highly sophisticated audiences can
begin searching for the nearest exit.
Yet underneath the technical language are fundamentally human ambitions.
How can researchers find information more easily?
How can datasets from different places work together?
How can a scientist reuse valuable research rather than unknowingly recreating it?
How can artificial intelligence accelerate discovery while remaining trustworthy?
How can sensitive human data be made useful for research while respecting privacy
and consent?
And ultimately:
How can better use of data lead to better understanding of human health?
That last question became the narrative anchor.
NIH describes data science as advancing the understanding of biomedical and
behavioral phenomena and contributing to breakthroughs that improve health and
wellbeing. Its strategic plan is intended to create a stronger biomedical data ecosystem
capable of turning increasingly sophisticated data and tools into discovery and care.
That gives all of the technology a reason to exist. Good communications strategy continually brings an audience back to the reason.
2. Five Goals Gave a Massive Strategy a Shape
One of the most useful characteristics of the plan was its underlying architecture. Rather than presenting data science as one enormous technology initiative, NIH organized the 2025–2030 strategy around five goals:
1. Improve capabilities to sustain the NIH Policy for Data Management and
Sharing.
2. Develop programs to enhance human-derived data for research.
3. Provide new opportunities in software, computational methods, and artificial
intelligence.
4. Support a federated biomedical research data infrastructure.
5. Strengthen a broad community in data science.
That structure did more than organize a document.
It created a narrative.
The first goal deals with how research data is managed and shared.
The second expands what researchers can learn from human-derived, clinical, health,
environmental, and real-world data.
The third addresses the rapidly evolving technologies used to analyze that information
— including software, advanced computational methods, and AI.
The fourth tackles one of the great challenges of modern research: connecting data that
exists across different systems rather than allowing it to remain trapped in silos.
And the fifth recognizes something essential that technology strategies sometimes
overlook:
None of this works without people.
Researchers need training.
Institutions need expertise.
Communities need access.
Teams need to collaborate across disciplines.
The plan’s structure gave us the raw material for a visual and linguistic hierarchy that
could help audiences move from one idea to the next without losing sight of the larger
mission.
That is one of the central principles of strategic communications:
Complexity doesn’t necessarily need to be removed. It needs to be organized.
3. Design Became Part of the Strategy — Not Decoration Applied Afterward
There is a persistent misconception about creative work in highly technical
environments.
The experts develop the substance.
Then, at the very end, somebody “makes it look good.”
That is not strategic design.
Strategic design begins much earlier.
It asks:
What does the audience need to understand first?
What information deserves prominence?
Which concepts belong together?
Where does the eye need to go next?
What can become visual rather than verbal?
Where does complexity require explanation — and where does explanation simply add
more complexity?
How do we create enough consistency that a reader can understand the system before
reading every word?
Those questions affect comprehension, not cosmetics.
For the NIH plan, the design challenge was particularly important because the strategy
itself describes an interconnected ecosystem.
Data management influences data reuse.
Standards influence interoperability.
Infrastructure influences accessibility.
Artificial intelligence depends upon the quality and availability of the data underneath it.
Training influences whether researchers can use any of those capabilities successfully.
The plan even connects its goals to NIH’s broader Digital NIH efforts, which
encompass technology capabilities spanning extramural research management,
intramural research, clinical research, and agency administration. The plan describes
common architecture, advanced computing and analytics, workforce capabilities,
workplace technology, and cybersecurity as cross-cutting components of that broader
digital environment.
Those are relationships.
And relationships are often easier to understand when people can see them.
That is where visual strategy earns its place at the table.
4. Language Had to Work Just as Hard as the Visuals
Designing information is only half the challenge.
The words still have to work.
Highly technical organizations naturally develop highly technical language. Specialists
need precision, and there are moments when simplifying terminology too aggressively
can actually make communications less accurate.
The answer is not to “dumb it down.”
The answer is to create pathways into the information.
A useful strategic document can operate at multiple levels simultaneously.
A reader scanning for three minutes should understand the big picture.
Someone responsible for implementing a specific initiative should be able to go deeper.
A scientific expert should recognize the rigor behind the strategy.
A collaborator, policymaker, communications professional, or potential partner should
not need a doctorate in computational biology to understand why the strategy matters.
The NIH plan accomplishes this partly through its progression from broad goals to strategic objectives and then to implementation tactics — moving audiences from why, to what, to how. The plan describes those implementation tactics as a roadmap whose specific priorities can continue to be assessed and adjusted as technologies, capabilities, and stakeholder needs evolve. That adaptability is important. Because particularly in data science, a five-year strategy cannot pretend the technology landscape will stand still for five years. The communications shouldn’t suggest that either.
5. The Story Wasn’t Really About Data. It Was About Connection.
One theme kept surfacing as the plan took shape: connection. Connecting new data to existing knowledge. Connecting repositories that historically operated separately. Connecting researchers to data. Connecting software and computational resources to the people who need them. Connecting different scientific disciplines. Connecting NIH Institutes and Centers. Connecting government with academia, industry, philanthropy, and the broader research community. And connecting increasingly powerful technologies to the ultimate NIH mission of advancing health. That is why Goal 4 — supporting a federated biomedical research data infrastructure — is such an important concept.
A federated model does not necessarily require every valuable dataset to be moved into one enormous central location. Instead, NIH is working toward an ecosystem in which researchers can more effectively discover and work across disparate resources while data providers retain appropriate control and privacy protections. NIH identifies greater interoperability, improved search and discovery, enhanced metadata, standardized access, and initiatives such as its Researcher Auth Service as pieces of that evolving ecosystem.
That is a technology story. But it is also a communications story. Because the fundamental idea is surprisingly simple: The value of knowledge increases when the right people can connect to it. The same is true of strategy.
6. AI Was Part of the Story — But It Wasn’t Allowed to Become the Whole Story
Today, it would be nearly impossible to publish a major data science strategy without artificial intelligence dominating the conversation. AI is unquestionably important to NIH’s vision. Goal 3 specifically addresses software, computational methods, and artificial intelligence, including responsible AI, advanced analytics, privacy-preserving computation, visualization, sustainable research software, and new ways for researchers to search, discover, access, and analyze information.
But what makes the strategy more interesting is that AI is not treated as an isolated miracle technology. It sits within an ecosystem. AI needs data. Data needs standards. Standards enable interoperability. Sensitive information requires governance and privacy protections. Tools require infrastructure. Infrastructure requires investment. And every one of those capabilities requires people who know how to use them. That systems-level perspective has become even more relevant as organizations everywhere race to adopt artificial intelligence.
Too many AI strategies start with: What AI tool should we buy? A better strategic question is: What ecosystem must exist for AI to create trustworthy, sustainable value? NIH’s data science strategy provides an instructive answer. Technology works best when the surrounding strategy is designed just as intentionally.
7. A Strategic Plan Is Not Finished When the Strategy Is Finished
This may be the most transferable lesson from the entire engagement. Organizations spend enormous amounts of time creating strategic plans. Leadership retreats. Stakeholder interviews. Research. Working groups. Drafts. Revisions. Board presentations. Approvals. And eventually someone declares: The plan is done. Except it isn’t. The thinking may be done. The communications work has just begun. A strategic plan has to travel. It needs to move beyond the people who developed it and reach the people expected to understand it, support it, fund it, implement it, collaborate around it, or act because of it. That means the final artifact matters. The narrative matters. The hierarchy matters. The language matters. The visual identity matters. The digital experience matters. And distribution matters.
The NIH Strategic Plan for Data Science was conceived as more than a repository for strategic thinking. We helped turn that thinking into a designed communications asset that could be distributed digitally and understood across an unusually broad and sophisticated stakeholder universe. [PORTFOLIO LINK TO COME: Explore the NIH Strategic Plan for Data Science, 2025–2030 in Daniels By Design’s Our Work section.] That distinction applies far beyond government.
If your organization has invested months developing a strategy but hands audiences a dense 75-page document and assumes the job is finished, you may have developed a strategy without ever developing a strategy for communicating the strategy. Meta? Absolutely. Necessary? Even more so.
8. What Data Science Taught Me About Communications Strategy
The irony of the ODSS assignment was ultimately its greatest lesson. Data science is about deriving meaning from information. Strategic communications is, in many ways, about doing exactly the same thing. Organizations already possess enormous amounts of information. Research. Plans. Customer insights. Institutional knowledge. Program results. Leadership priorities. Market intelligence. Data.
The challenge is rarely simply acquiring more. The challenge is identifying what matters, recognizing patterns, establishing relationships, creating hierarchy, and translating all of it into something people can use. That is what data scientists do with datasets. It is also what strong communicators do with organizations. Both disciplines ask versions of the same question: What does all of this information actually tell us — and what should we do because of it? That is where communication moves beyond promotion. It becomes strategy.
That is a technology story. But it is also a communications story. Because the fundamental idea is surprisingly simple: The value of knowledge increases when the right people can connect to it. The same is true of strategy.
Conclusion: Making Complexity Useful By Design
The NIH Office of Data Science Strategy did not need us to teach its scientists about data science. They were the experts. Our job was different. We needed to understand their expertise well enough to help shape how other people experienced it. To identify the narrative inside the complexity. To create order around an ambitious five-year vision. To give five major goals a coherent visual and linguistic framework. To help technical concepts coexist with human purpose. And ultimately, to create a strategic communications asset worthy of the strategy it represented.
That experience continues to influence how Daniels By Design approaches complex assignments today. The more sophisticated the subject matter, the greater the temptation is to communicate everything. But communicating everything is not the same as communicating effectively. Effective strategy makes choices. Effective storytelling establishes hierarchy. Effective design creates pathways. And effective communications connects information with the people who need to understand it. NIH’s Strategic Plan for Data Science is ultimately about building a biomedical data ecosystem capable of connecting information, technologies, researchers, and communities in ways that advance discovery and improve health.
Our role was to help make that ecosystem understandable.
Because even the smartest strategy in the room cannot create its full impact if the
people outside the room cannot see what it means.
Data can reveal the story.
Strategy can determine where that story needs to go.
Design helps people get there.
That’s Data Science By Design.
And why that’s fantastic.