Aligning platform strategy with advanced analytics needs
Enterprise Technology | Data Platforms
As McKinsey expanded its advanced analytics capabilities, questions arose around platform alignment: could the firm consolidate infrastructure and workflows onto a single cloud provider for client work? Our research-led evaluation explored the needs of advanced analytics consultants across McKinsey and QuantumBlack to inform strategic decisions between AWS and Azure.
As Senior Manager, UX Research, I co-led the initiative alongside the two cloud PMs and infrastructure leads—guiding a 14-day sprint focused on cloud workflows, tooling preferences, and platform constraints. Due to the project’s technical nature, it was a deeply collaborative effort that brought together UX research, cloud product managers, and advanced analytics practitioners.
While the result revealed no clear platform winner, the research surfaced nuanced insight into how AA teams actually work—insight that later supported the development of McKinsey’s custom AA platform.
CLIENT
McKinsey & Company | Client Technology
01. Situation
The firm’s infrastructure team needed to assess whether McKinsey could streamline client-facing advanced analytics work onto a single cloud environment (AWS or Azure). The challenge was complex:
-
Client platforms and preferences varied widely
-
Some projects ran on McKinsey infrastructure, others on client environments
-
Consultants brought their own preferences and tools
-
Advanced analytics workflows (from model dev to deployment) were rarely standardized
-
Technical nuances and edge cases weren’t fully understood by decision-makers
This made it difficult to recommend one cloud provider without a deeper understanding of actual working conditions.
02. Task
Conduct focused research to:
-
Map the end-to-end advanced analytics workflow, from ingestion to deployment
-
Identify what matters most to AA consultants in tooling, environment setup, and feature usage
-
Compare AWS and Azure based on real-world workflows—not assumptions
-
Inform leadership’s cloud consolidation decisions with grounded, user-informed insights
03. Action
-
Co-scoped research plan with product and engineering leads from both AWS and Azure teams
-
Led 14-day deep-dive into AA workflows across McKinsey and QuantumBlack
-
Conducted in-depth interviews with advanced analytics consultants, data scientists, and infra specialists
-
Mapped common and edge-case workflows to visualize tool usage, friction, and flexibility needs
-
Captured preferences around features, performance, portability, and client-side integration
-
Synthesized insights into themes, trade-offs, and infrastructure implications
-
Presented findings in a workshop that surfaced key tensions—and set the stage for strategic next steps
04. Result
-
Revealed a 50/50 split in cloud platform preference—confirming there was no one-size-fits-all solution
-
Surfaced actionable insight into how advanced analytics consultants actually work, including workflow friction and decision logic
-
Informed platform development efforts, including McKinsey’s custom-built AA tooling platform
-
Created a shared understanding across infrastructure teams, replacing assumptions with grounded insight
Illustrating the process
Artifacts and patterns from the research sprint—highlighting the power of tightly scoped, technically informed discovery in infrastructure strategy.
AA consultant workflow mapping
Capturing complexity across client, cloud, and firm environments
We mapped the full end-to-end advanced analytics journey—revealing touchpoints across McKinsey infra, client infra, and both cloud platforms. This journey visualization helped technical and non-technical stakeholders grasp the real complexity behind “choose one cloud” questions.
Team synthesis workshop
Bringing assumptions to the surface through cross-functional alignment
A collaborative session with product, engineering, and platform leads helped surface unstated assumptions and competing priorities. Together, teams debated trade-offs and built a shared mental model of what was really at stake—setting the tone for deeper collaboration.
Cloud trade-off framework
Making platform decisions visible and evidence-based
.png)
We synthesized platform pros and cons based on actual consultant workflows—spanning portability, dev environments, deployment models, and compliance constraints.The resulting comparison helped teams weigh decisions based on real-world behavior.
Workflow themes & tool usage patterns
Understanding how advanced analytics consultants really work
From Jupyter to Databricks to Snowflake, we visualized which tools showed up where—and why. Patterns emerged across skill levels, project types, and cloud access limitations, revealing insight far beyond initial expectations.
Key takeaway
This sprint confirmed that platform decisions can’t be made in a vacuum. While some stakeholders initially believed “we already know what advanced analytics teams need,” the research revealed deeper nuance and unexpected complexity. Despite a 50/50 cloud preference split, the process gave teams a shared, grounded view of how analytics work really happens—paving the way for smarter infrastructure investments and more tailored platform development.
Navigator methods & frameworks used
Design & Technical Research Foundations
-
Interview techniques for UX
-
Journey mapping
-
Design sprints for research
-
Jobs to be done (JTBD)
Strategy & Alignment Tools
-
Workshop facilitation
-
Lean canvas
-
Google's HEART framework
I was resistant at first, but I learned more in these 14 days than I did in a year working on this project.
– Senior PM, Azure platform team