The PACE Framework: How Great Analysts Actually Drive Business Impact

In my recent Horizons interview with Sundar Swamanathan, former head of US/Canada performance marketing analytics at Uber, he revealed a practical framework for turning data into business impact. Here’s the complete breakdown of how successful analysts navigate from insight through to action.

Imagine spending weeks on a detailed analysis, creating the perfect deck, and proudly presenting your findings… only to have them gather dust in a shared drive somewhere. If you’re an analyst, you’ve probably lived this story. If you’re a leader working with analysts, you’ve probably wondered why some analyses drive action while others fade away.

Here’s the reality: technical skill is just the beginning. During his time at Uber, Sundar saw analysts regularly influence multi-million dollar decisions - including one analysis that led to turning off $35 million in annual Facebook spend. But these wins didn’t come from SQL mastery alone.

“There are four steps in an analysis that most people don’t realize,” Sundar explained in our conversation. “Prioritization, analysis, communication and execution.”

The Foundation: Data-Obsessed Culture

Before diving into the PACE framework itself, we need to understand the cultural context where it thrives. At Uber, this wasn’t just about having good analysts - it was about building what Sundar calls a “data-obsessed culture.”

“I truly think it has to be founder-led,” Sundar emphasized. “A lot of it is because data-driven data obsession, whatever you want to call it, what it results in is an ego-less decision making.”

This manifested in several key ways:

1. Junior people could influence major decisions with good data
2. Failed experiments were seen as valuable learning opportunities
3. Data access was democratized through tools like Query Builder
4. Leaders consistently asked for data to support decisions

The culture was so strong that Sundar saw “queries go viral” at Uber - analysts would share SQL queries via URLs in Query Builder, and these would spread throughout the organization as others built upon and modified them. This created a collaborative environment where data insights could come from anywhere in the organization.

P is for Prioritization: The First Hour That Changes Everything

Most analysts dive straight into the data, but the best ones start somewhere else entirely. They begin by deeply understanding the business context and stakeholder needs.

“The only way to know what the highest priority things are is to talk to your stakeholders,” Sundar emphasized. At Uber, he instituted a powerful practice: mandatory weekly meetings between analysts and stakeholders at the start of each week.

Why was this so important? Because it fundamentally shifted the relationship. “I got really sick very quickly of people thinking that data science was a service function and that we were just there to pull queries,” he explained. “If you want us to work with you, we’re not working for you, we work with you. And we reestablished the relationship as partners, not support.”

This mindset shift unlocks several benefits:

• Better project prioritization
• Deeper understanding of business context
• Stronger relationships with stakeholders
• More influence on decision making
• Greater likelihood of implementation

Time investment: 1 hour per week maximum

Working with Marketing Stakeholders

Sundar shared specific insights about analysts working with marketing teams:

• Give marketers context about what you’re seeing in the data
• Never fudge numbers, even if pressured
• Help stakeholders understand the uncertainty in statistics
• Focus on finding actionable insights, not just reporting numbers

How to Make the Most of Prioritization Time

1. Schedule regular check-ins with key stakeholders
2. Ask about their current challenges and priorities
3. Share what you’re working on and get feedback
4. Discuss recent wins and learnings
5. Align on next steps and timelines

A is for Analysis: The Core of the Work

With clear priorities established, it’s time for the technical work. But here’s a counterintuitive insight from Sundar: he recommends spending no more than 3-4 days on any single analysis.

Why? “Because what happens is you just go down this rabbit hole of chasing things. And then your analysis ends up being pretty complicated and messy. And then also stakeholders are waiting for an analysis, and now you’re like, wait, I just need one more day, I need one more day.”

The Power of Simple Tools

Throughout his five years at Uber, Sundar primarily used just two tools: SQL and Google Sheets. This wasn’t a limitation - it was a choice that enabled faster, more focused analysis.

“Bankers who are making multi-billion dollar mergers and acquisition deals are just using Excel,” he points out. “And they’ve been using Excel for hundreds… of years. That’s enough for them.”

This focus on simple tools helped in several ways:

• Faster iteration on analyses
• Easier collaboration with stakeholders
• More time for insight generation
• Less distraction from tool complexity

The Query Builder Effect

While the tools were simple, Uber did build one internal tool that transformed how analysts worked: Query Builder. This web-based SQL interface did something remarkable - it turned queries into shareable URLs.

“You could run a query and it would create a new URL for that query,” Sundar explained. “And that was awesome because I could go share that URL with anyone… and then people would tweak that query run for their own purposes. And all of a sudden you have this virality of a SQL query.”

This created a culture where:

• Best practices spread organically
• Analysts built on each other’s work
• Data insights could be easily shared
• Collaboration happened naturally

The key is to start with clarity about what you’re trying to answer. “Write out what you’re actually trying to answer first,” Sundar advises. “Then it becomes a lot easier to go figure out the queries that you need to write.”

Time investment: 3-4 days per analysis

C is for Communication: The Make-or-Break Moment

This is where many analysts stumble. After doing brilliant technical work, they fail to communicate their findings effectively.

“I can’t count the number of times where I’ve just been in a room and I’m like, I don’t understand what this deck is,” Sundar shared. “When your stakeholder feels like you’ve wasted their time, it really leaves a bad taste in their mouth.”

The $35 Million Analysis

Consider Sundar’s Facebook analysis at Uber. Despite potentially turning off $35 million in annual spend, he presented his findings in just seven slides. The key elements included:

• Clear data showing CAC volatility
• Connection to saturation analysis
• Simple proposal: 3-month test
• Clear expected outcomes

The brevity wasn’t accidental - it forced clarity of thought and focused the discussion on the key decision.

Understanding Statistics and Uncertainty

One crucial aspect of communication that Sundar emphasized was helping stakeholders understand uncertainty in data:

“There’s a misconception that data science is designed to create perfection by logic. It’s statistics, which has inherent uncertainty into it. In fact, everyone like every statistician publishes things with confidence intervals, 95% margins of errors… trying to find uncertainty on something uncertain.”

Time investment: 1 full day (spread across the week)

E is for Execution: The Often Forgotten Final Step

Many analysts consider their job done after presenting their findings. But the best analysts know that impact only happens through implementation.

“The only way you get promoted is if you make impact,” Sundar explains. “Well, you can only make impact if your recommendations get absorbed, if they get implemented, people go take actions on them.”

Building Trust Through Follow-Through

At Uber, this execution mindset led to significant shifts in how marketing budgets were handled. “I know what it’s like as a marketer to be concerned over budgets and giving back budgets and saying, well, if I give this back, I’m going to lose it,” Sundar noted. “And that’s the kind of stuff that I didn’t see [at Uber]… It wasn’t the sense of, I’m going to lose my budget. It was, this is the right thing for the business.”

Time investment: 1-2 hours per week

The Experimentation Connection

The PACE framework ties directly into successful experimentation. As Sundar explained, “90% of experiments fail… What that means is you have to be in such a psychologically safe environment that allows you to fail 90% of the time and people still think that’s a good thing.”

This high failure rate makes the framework even more crucial:

• Prioritization ensures you’re testing the right things
• Analysis needs to be rigorous but time-boxed
• Communication must focus on learnings, not just results
• Execution includes planning the next experiments

The Seven Steps of Experimentation

Sundar outlined a specific process that complements PACE:

1. Start with customer insights
2. Generate multiple hypotheses
3. Prioritize experiments
4. Design rigorous tests
5. Ensure quality data
6. Analyze thoroughly
7. Create a perpetual cycle

Measuring Success Through PACE

How do you know if you’re successfully implementing PACE? Sundar suggests looking for these indicators:

• Stakeholders proactively seek your input
• Analyses regularly drive concrete actions
• Failed experiments lead to new insights
• Communication becomes more efficient
• Impact can be clearly measured

The Budget Test

One clear sign of success? When returning unused budget isn’t seen as a risk. As Sundar explained, at Uber “you knew you could give it back, but you could go then redeploy it somewhere else. You could go try maybe another channel… It wasn’t the sense of, I’m going to lose my budget. It was, this is the right thing for the business.”

Getting Started with PACE

Whether you’re an analyst looking to increase your impact or a leader trying to help your team succeed, here are three ways to start applying PACE:

1. Audit Your Time

• Track how you currently spend your time across these four areas
• Look for obvious gaps or imbalances
• Make small adjustments week by week

2. Start with Communication

• Review your last three analyses
• Ask stakeholders for honest feedback
• Practice explaining complex findings simply

3. Build Relationships

• Schedule regular check-ins with key stakeholders
• Ask about their priorities and challenges
• Share your work in progress and get early feedback

A Note on Team Structure

Sundar suggests an interesting model for smaller companies: consider fractional or interim analytics leadership while maintaining full-time analysts. “For the leader, especially early on, if you’re a smaller company, you don’t need a full-time analytics leader. It’s just overkill and too expensive. But you do need someone to mentor, prioritize, coach analysts, work with leadership.”

The Future of Analytics Leadership

The role of analysts is evolving. As Sundar notes, “The problem with analysts has never been the lack of tooling or tools to make their jobs easier. It’s always been [that] analysts are really not great at being business savvy.”

This is where PACE becomes crucial - it’s not just a framework for doing analysis, it’s a framework for driving business impact through data. It helps analysts:

• Build business acumen
• Develop stakeholder relationships
• Drive concrete actions
• Measure real impact

Even as tools evolve (including AI), these fundamental skills remain crucial. As Sundar puts it, “AI will make smart people smarter… The analyst has to have had the potential to be 10X in the beginning.”

Building Your Data Career Through PACE

The framework isn’t just about individual analyses - it’s about building a successful career in data. Sundar’s journey from the US Treasury to Uber to advising startups demonstrates the power of combining technical skills with business impact.

The key is remembering that data work isn’t about the tools or the techniques - it’s about driving real business impact. As Sundar emphasizes, you have to “transform the way they look at a loss or a failure and really convert it into this is a learning opportunity.”

Want to hear more insights about driving impact through data? Listen to my full interview with Sundar, where he shares his journey from managing $19 trillion at the US Treasury to leading Uber’s brand data science team.