• LinkedIn
• Twitter
• Instagram
• YouTube
• RSS Feed

Apple •Spotify• Pocket Casts •Youtube •Overcast •RSS

What’s up everyone, today we have the pleasure of sitting down with Sundar Swaminathan, author of the experiMENTAL newsletter and part time Marketing and Data science advisor.

Summary: After leading Uber’s Marketing Data Science teams, Sundar shares insights that work for both tech giants and startups. Beyond uncovering that Meta ads generated zero incremental value (saving $30 million annually), they mastered measuring brand impact through geo testing and predicting LTV through first-week behaviors. Small companies can adapt these methods through strategic A/B testing and simplified attribution models, even with limited sample sizes. Building data science teams that embrace business impact over technical complexity, and maintaining curiosity, like when direct driver engagement revealed that recommending Saturday afternoon starts over Friday peak hours improved retention.

In this Episode…

• Marketing Incrementality Testing Reveals Meta Ads Ineffective at Uber
• How to Run Marketing Experiments With Limited Data
• Best Ways To Measure Marketing Campaign ROI
• How to Choose the Right Marketing Attribution Model for Multiple Channels
• Using Propensity Matching for Marketing ROI Analysis
• A Guide to Customer Lifetime Value Prediction Methods
• Proving Brand Campaign Impact Using Geographic Testing
• Why Marketing Faces Unfair ROI Pressure and How to Embrace Measurement
• How to Build Strategic Data Science Teams for Marketing

About Sundar

• Sundar started his career as a software developer at Bloomberg before managing $19 Trillion at the US Treasury as a Debt Manager
• He pivoted to growth marketing and data science consulting where he worked with DirectTV and an ed-tech AI startup
• He then made the mega move to Uber where he spent 5 years building Brand, Performance, and Lifecycle Marketing Data Science teams
• He moved over to a travel tech startup and helped them go from $0 to $100K MRR
• Today, Sundar is a marketing and data science advisor, he helps B2C founders and marketers 
• He’s also working on an upcoming podcast and has a newsletter where he shares frameworks, how-to guides to help B2C marketers

Marketing Incrementality Testing Reveals Meta Ads Ineffective at Uber

Performance marketing often reveals surprising truths about channel effectiveness, as demonstrated by a fascinating case study from Uber’s marketing operations. When confronted with unstable customer acquisition costs (CAC) that fluctuated 10-20% week over week despite consistent ad spend on Meta platforms, Uber’s performance marketing team, led by Sundar, decided to investigate the underlying causes.

The investigation began when the team noticed significant volatility in signup rates despite maintaining steady advertising investments. This inconsistency prompted a deeper analysis of Meta’s effectiveness as a primary performance marketing channel. The timing of this analysis was particularly relevant, as Uber had already achieved substantial market penetration eight years after its launch, especially in major urban markets where awareness wasn’t the primary barrier to adoption.

Through rigorous data analysis, the team implemented a three-month incrementality test to measure Meta’s true impact on user acquisition. The test utilized a classic A/B testing methodology, comparing a control group receiving no paid ads against a treatment group exposed to Meta advertising. The results were striking: Meta advertising showed virtually no incremental value in driving new user acquisition, a finding that was validated by Meta’s own data science team.

The outcome of this experiment led to a significant strategic shift, resulting in annual savings of approximately $30 million in the U.S. market alone. While this figure might seem modest for a company of Uber’s scale, its implications were far-reaching when considered across global markets. The success of this experiment also highlighted the importance of data-driven decision-making and the willingness to challenge assumptions about established marketing channels.

Key takeaway: Established marketing channels should never be exempt from rigorous effectiveness testing. Regular incrementality testing can reveal unexpected insights about channel performance and lead to substantial cost savings. Marketing teams should prioritize data-driven decision-making over assumptions about channel effectiveness, even for seemingly essential platforms.

Back to the top ⬆️

How to Run Marketing Experiments With Limited Data

Most companies don’t have the volume of signups or users that an Uber does. Marketing experiments require a mindset shift when working with small data samples. While A/B testing remains the gold standard for measuring marketing effectiveness, Sunday thinks that companies with limited data can still validate their marketing efforts through strategic pre-post testing approaches.

Pre-post testing, when properly implemented, serves as a valuable tool for measuring marketing impact. The key lies in isolation: controlling variables and measuring the impact of a single change. For instance, a marketplace company successfully conducted a pre-post test on branded search keywords in France by isolating specific terms in a defined region. This focused approach provided reliable insights despite not having the massive data volumes typically associated with incrementality testing.

That being said, Sundar adds that early-stage companies should prioritize high-impact experiments capable of delivering substantial results vs testing tiny changes that will barely have detectable effects. With small sample sizes, tests should target minimum detectable effects (MDE) of 30-40%. These larger effect sizes become measurable even with limited data, making them ideal for fundamental changes such as exploring new ideal customer profiles (ICPs) or revamping core value propositions, rather than pursuing minor optimizations.

An example that Sundar recalls while working at a travel tech startup demonstrated the value of running A/B tests even with limited data. Despite having only 100-200 weekly signups, they detected a 40% conversion drop after modifying their onboarding flow. While the test might have been considered “poorly powered” by strict statistical standards, it successfully prevented a significant negative impact on the business. This illustrates how even small-scale testing can provide crucial insights; it’s better to have 60% confidence in a positive change than to miss a catastrophic drop with 95% confidence.

The confidence level in marketing experiments operates on a spectrum, with A/B tests providing the highest confidence and pre-post tests offering valuable but less definitive insights. Success depends on maintaining experimental discipline, carefully controlling variables, and understanding the tradeoffs between confidence levels and the humbling reality of practical constraints. Marketing teams must balance their confidence requirements against their risk tolerance when designing and interpreting tests.

Key takeaway: Companies with limited data should focus on measuring high-impact marketing changes through carefully controlled pre-post tests. Success comes from isolating variables, targeting substantial effect sizes, and maintaining experimental discipline. This approach enables meaningful measurement while acknowledging the practical constraints of smaller data sets.

Back to the top ⬆️

The Difference Between AB Testing and Incrementality Testing

Marketing experimentation terminology often creates unnecessary complexity in what should be straightforward concepts. The fundamental structure of both A/B testing and incrementality testing follows the same principle: comparing outcomes between groups that receive different treatments.

Statistical analysis remains consistent across both testing approaches. Whether using Bayesian or frequentist methods, the underlying comparison examines differences between groups, regardless of what those groups receive. The statistical calculations remain indifferent to whether one group receives no treatment (as in incrementality tests) or a variation of the treatment (as in traditional A/B tests).

Incrementality testing extends beyond simple presence versus absence comparisons. For example, marketers can test spending incrementality by comparing groups receiving different budget allocations ($100 versus $200 spend). Some experiments even incorporate three variants, such as testing zero spend, standard spend (1X), and double spend (2X) simultaneously. This approach maintains budget neutrality while mapping the response curve across different spending levels.

Both A/B testing and incrementality testing serve as tools within the broader framework of experimental design. While incrementality testing typically focuses on validating channel effectiveness, and A/B testing often examines creative or tactical variations, they share the same statistical foundation. The key difference lies in their application rather than their fundamental methodology.

Key takeaway: A/B testing encompasses all experimental comparisons between groups, including incrementality tests. Marketers should focus less on rigid definitions and more on clearly defining test objectives and measurement approaches. This simplified understanding enables more flexible and practical experimental design while maintaining statistical rigor.

Back to the top ⬆️

Best Ways To Measure Marketing Campaign ROI

Marketing measurement techniques fall into distinct categories based on their scope and application. While Multi-Touch Attribution (MTA) and Marketing Mix Modeling (MMM) serve as comprehensive measurement frameworks, they differ fundamentally from campaign-specific measurement methodologies.

The five core methods of measuring campaign effectiveness (pre-post analysis, difference-in-difference, causal inference, A/B testing, and baseline analysis) operate at a granular, campaign-specific level. These approaches can be arranged on a spectrum, balancing accuracy against implementation complexity. Pre-post analysis, for example, functions as a simplified A/B test, comparing performance before and after a campaign implementation.

MTA and MMM, by contrast, operate as holistic marketing measurement tools. These methods evaluate cross-channel effectiveness and model overall marketing ROI rather than isolating individual campaign performance. While MMM has demonstrated value in campaign measurement scenarios, its primary strength lies in providing a comprehensive view of marketing effectiveness across channels.

Campaign-specific measurement tools offer direct insight into the effectiveness of individual marketing initiatives. These methodologies enable marketers to isolate and evaluate specific changes, making them invaluable for tactical decision-making and optimization efforts. The choice between different measurement approaches depends largely on the specific questions marketers need to answer and the level of granularity required.

Key takeaway: Choose your measurement methodology based on your analysis goals. Use campaign-specific methods (pre-post, A/B testing, causal inference) for evaluating individual initiatives, and reserve comprehensive tools like MTA and MMM for understanding broader marketing effectiveness. This targeted approach ensures you’re using the right tool for the right analytical job.

Back to the top ⬆️

How to Choose the Right Marketing Attribution Model for Multiple Channels

Attribution modeling becomes more nuanced as companies grow and marketing strategies evolve. The debate between channel-specific attribution versus unified attribution approaches presents an interesting challenge for marketing teams of all sizes. When discussing attribution methods at Uber, Sundar reveals compelling insights about the practical application of attribution models at scale.

Maintaining consistency in attribution methodology serves as a cornerstone for effective marketing measurement. Sundar emphasizes that while last-click attribution has its limitations, its consistency provides valuable insights over time. The framework follows the MECE principle (Mutually Exclusive, Collectively Exhaustive), which becomes compromised when different channels employ varying attribution methods. Using multiple attribution models for different channels creates potential overlap, undermining the fundamental purpose of attribution tracking.

The complexity of marketing demands a comprehensive approach to measurement. At Uber, the team employed a combination of last-click attribution, Multi-Touch Attribution (MTA), Marketing Mix Modeling (MMM), and incrementality testing. This multi-faceted strategy allowed them to triangulate data points and develop a more complete understanding of marketing performance. Rather than relying on channel-specific attribution methods, this unified approach provided consistent, trackable results while acknowledging the inherent limitations of any single attribution model.

Changes in attribution patterns often signal significant shifts in marketing effectiveness. When introducing new channels, marketers can leverage the consistency of a unified attribution model to identify meaningful changes in performance. This approach helps teams recognize when new channels impact existing marketing efforts, providing valuable insights into the overall marketing ecosystem. The predictability of certain patterns, such as Google’s tendency to receive higher attribution credit or non-digital channels showing minimal direct attribution, becomes a useful framework for analysis.

Key takeaway: Implement a consistent attribution framework across all channels while using multiple measurement methodologies (last-click, MTA, MMM, incrementality testing) to triangulate true marketing impact. This approach provides more reliable insights than channel-specific attribution methods, which can lead to overlap and inconsistent measurement.

Back to the top ⬆️

Using Propensity Matching for Marketing ROI Analysis

Propensity matching offers marketers a powerful alternative to traditional A/B testing when measuring program effectiveness. This statistical approach compares customers who participated in a program with similar customers who did not, creating a natural experiment that reveals true program impact. Sundar shares insights from implementing this methodology at Uber, particularly during the Uber One loyalty program launch.

First-party behavioral data forms the foundation of effective propensity matching. Instead of relying on demographic data, which can be expensive and privacy-sensitive, marketers can leverage rich behavioral signals they already possess. These signals include usage patterns (time of day, day of week), product preferences (service types selected), location characteristics (urban vs. suburban), and customer tenure. At Uber, these behavioral patterns even enabled the team to identify specific customer segments, such as government employees based on their regular commute to particular areas.

Communicating the value of propensity matching to stakeholders requires framing it as a collaborative solution. Sundar emphasizes approaching stakeholders with the mindset of finding alternatives to controlled rollouts, which can disrupt network effects in platforms like Uber. The methodology appeals to marketers because it eliminates the need for holdout groups while still providing robust impact measurement. This approach helps compare similar customers (apples to apples) rather than dissimilar ones (apples to oranges).

Behavioral segmentation through first-party data often yields more accurate insights than traditional demographic segmentation. While demographic data might seem essential for customer understanding, behavioral patterns provide direct insight into customer actions, which marketers ultimately aim to influence. This approach proves particularly valuable in privacy-conscious environments where demographic data collection faces increasing restrictions.

Key takeaway: Build propensity matching models using first-party behavioral data instead of demographic information. Focus on actual customer behaviors like usage patterns, product preferences, and location data to create accurate matches. This approach provides reliable ROI measurement while respecting privacy concerns and eliminating the need for controlled rollouts.

Back to the top ⬆️

A Guide to Customer Lifetime Value Prediction Methods

Short-term indicators provide surprisingly accurate predictions for long-term customer behavior metrics. Drawing from his experience at Uber, Sundar explains how rapid business growth and frequent product launches influenced their approach to predictive modeling. The company discovered that customer retention and churn patterns emerge quickly, often within the first few interactions with the service.

Customer decisions stem from immediate experiences rather than long-term patterns. A single negative experience, such as extended wait times or unexpected price surges, can trigger immediate churn. Similarly, positive experiences can quickly restore customer confidence and drive retention. This immediate cause-and-effect relationship makes short-term behavioral data particularly valuable for predictive modeling.

The limitations of long-term lifetime value (LTV) calculations become apparent when considering market dynamics. Sundar challenges the conventional wisdom of extended LTV predictions, suggesting that measurements beyond 18-24 months lose accuracy and practical value. External factors like economic cycles and unexpected events (such as the pandemic) can completely invalidate long-term models, necessitating frequent rebuilds.

Early customer behavior serves as a reliable predictor of future value. At Uber, the team found that customer actions within their first week strongly indicated their potential two-year value. This insight led to developing predictive models based on short-term behavior patterns, using ratios between six-month and two-year values as monitoring benchmarks. This approach balanced the need for quick insights with the reality that true long-term value calculations require waiting for the full time period to elapse.

Key takeaway: Focus on building predictive models using short-term behavioral signals (1 week to 6 months) rather than attempting to forecast years ahead. Monitor the relationship between short-term and long-term metrics to validate your predictions, and be prepared to rebuild models when significant market changes occur.

Back to the top ⬆️

Proving Brand Campaign Impact Using Geographic Testing

Marketing measurement becomes fascinating when examining brand awareness campaigns at scale. While national TV campaigns resist traditional measurement approaches like Multi-Touch Attribution (MTA) and incrementality testing, other brand initiatives offer rich opportunities for rigorous measurement and optimization. Sundar shares insights from his unique position bridging performance and brand marketing at Uber, where he built their brand data science team.

The foundation of effective brand measurement starts with a fundamental question: does the campaign move the intended metric? Surprisingly, many companies launch brand campaigns without establishing this baseline. The methodology begins by confirming consistent movement in awareness metrics through targeted creative and customer insights. Once this predictable movement is established, the team implements geographic lift tests, strategically holding out specific regions while running omnichannel brand campaigns in others.

These geographic experiments require patience and commitment, typically running for 12 weeks to 6 months to capture meaningful results. The process validates a crucial hypothesis: substantial increases in brand awareness metrics consistently drive downstream business impact. For instance, a 20-point increase in awareness naturally leads to increased session activity. This relationship extends through the funnel, from consideration to concrete business metrics like ride requests.

Post-campaign analysis revealed a critical insight about campaign duration. When brand activities ceased, awareness metrics consistently declined, demonstrating the ineffectiveness of short-term brand activations. This data-driven revelation supported the strategic shift toward “always-on” brand marketing, creating a synergistic relationship between brand marketers’ intuition and data science validation. Initially met with resistance to holdout testing, the approach eventually united teams around shared goals of proving and improving ROI.

Key takeaway: Brand campaign effectiveness requires geographic testing over 3-6 month periods, with consistent presence rather than short-term activations. Success depends on establishing clear metric movement, implementing proper control groups, and maintaining long-term commitment to brand building activities.

Back to the top ⬆️

Explaining Unmeasurable Marketing Results to Executives

Navigating measurement expectations in marketing requires a delicate balance of transparency, education, and proactive communication. Sundar shares insights about handling situations where traditional ROI metrics become impossible to track, particularly in broad marketing initiatives like national TV campaigns.

The most problematic scenario occurs when marketing leaders are caught off guard by the absence of measurement capabilities. Picture a CEO of a multi-billion dollar company inquiring about the results of a significant marketing investment, only to discover that no measurement framework was established. This situation creates unnecessary tension and undermines the marketing team’s credibility. The solution lies in setting clear expectations from the beginning and acknowledging measurement limitations before campaign execution.

Sundar emphasizes that while it’s acceptable to run campaigns without comprehensive measurement, the key is establishing this understanding upfront. Marketing operations teams should resist the pressure to manufacture data points retrospectively. This approach might initially face resistance, but it builds trust and credibility over time. Through experience, marketing teams learn to include measurement discussions in their planning phases, leading to more strategic campaign development.

The conversation around measurement limitations often becomes an opportunity for education and alignment. When teams want to execute large-scale campaigns that can’t be measured precisely, the focus shifts to understanding trade-offs. Sometimes, alternative measurement approaches can be explored, such as analyzing performance in smaller markets with fewer variables or conducting pre-post analysis. However, these methods come with their own limitations and should be presented with clear confidence levels and risk assessments.

Key takeaway: Success in marketing measurement hinges on proactive communication about limitations. Start measurement strategy discussions early, be transparent about what can and cannot be measured, and ensure all stakeholders understand these constraints before campaign execution. This approach prevents surprises and builds trust between marketing operations and executive leadership.

Back to the top ⬆️

Why Marketing Faces Unfair ROI Pressure and How to Embrace Measurement

Marketing faces unique scrutiny over ROI that other departments rarely encounter. While finance teams report on budget adherence and HR tracks employee satisfaction, marketing must justify every campaign, creative decision, and strategic initiative with revenue metrics. This disparity has complex historical roots and practical implications for modern marketing teams.

The dynamics of marketing measurement have shifted dramatically since the 1990s. During the earlier decades, marketing commanded respect without constant ROI validation, particularly in consumer packaged goods companies like Coca-Cola and Pepsi. These brands excelled at marketing and storytelling because they operated in commoditized markets where differentiation through marketing proved essential. However, the rise of technology companies introduced a new paradigm, where product teams established robust measurement frameworks early on. Sundar points to Facebook’s legendary growth team as an example of data-driven experimentation becoming the norm.

The emergence of performance marketing in the past 15 years has further complicated the situation. The ability to track every click and attribute specific actions created an expectation that all marketing activities should demonstrate similar measurability. This has put traditional brand marketing in a challenging position, competing for resources against both product teams and performance marketing initiatives that appear more quantifiable. The irony lies in the fact that brand marketing historically never needed to prove its ROI, leaving many teams unprepared for today’s measurement demands.

Progressive companies now invest in advanced measurement capabilities, bringing in experts in econometrics and data science to quantify brand impact. These efforts consistently reveal positive ROI for brand investments, gradually building a case for more nuanced evaluation of marketing activities. This evolution suggests a potential shift away from the strict performance marketing paradigm that has dominated recent years, toward a more balanced understanding of marketing’s multifaceted impact.

Key takeaway: Marketing teams should acknowledge the historical context of ROI measurement while developing sophisticated frameworks that capture both immediate performance metrics and long-term brand value. Instead of resisting measurement, focus on creating comprehensive evaluation models that reflect marketing’s full scope of impact across both quantifiable and qualitative dimensions.

Back to the top ⬆️

How to Build Strategic Data Science Teams for Marketing

Marketing teams often chase perfect ROI measurement at the expense of growth, particularly in consumer companies where rapid scaling is essential for survival. Sundar emphasizes that this pursuit of perfect measurement can actually hinder progress, suggesting a more balanced approach focused on strategic decision-making and clear communication.

The fundamental challenge lies in how data scientists perceive their role within marketing organizations. Rather than viewing themselves as technical specialists who happen to work in marketing, they should approach their work as marketers with strong data science capabilities. This mindset shift is crucial for effectively communicating complex trade-offs to leadership, such as when to pull back on Meta advertising despite potential negative impacts on quarterly numbers.

A compelling example from Sundar’s consulting work illustrates this principle. When working with an e-commerce company’s high-performing CRM team, he discovered they were running daily deal campaigns simply because “that’s how they’ve always done it.” Through a strategic test comparing different email frequencies, they proved that seven weekly emails provided no additional benefit over four, while increasing unsubscribe rates. Despite this valuable insight, the team hadn’t previously felt empowered to challenge existing practices or discuss optimization opportunities with leadership.

The irony of marketing analytics lies in its accessibility; while CEOs wouldn’t casually step in to handle financial reconciliation, they often feel qualified to direct marketing decisions. This dynamic makes it even more critical for data professionals to develop strong leadership communication skills and the confidence to push back with data-driven insights. As Sundar notes, the weakness in most marketing data stacks isn’t technical capabilities but rather the human element of strategic communication and decision-making.

Key takeaway: Focus on developing data professionals who think like marketers first and technical experts second. The key to successful marketing measurement isn’t perfect ROI tracking but rather building teams that can effectively communicate strategic trade-offs to leadership and confidently challenge assumptions with data-driven insights.

Back to the top ⬆️

Finding Technical Talent That Embraces Marketing Challenges

Marketing and technical roles often seem to exist in separate worlds. Technical professionals frequently gravitate toward product development, user experience, and coding while avoiding marketing-related tasks. However, Sundar’s experience at Uber reveals how curiosity and genuine interest in solving customer problems can bridge this gap.

Personal investment in a company’s mission can transform how technical professionals approach marketing challenges. During his time at Uber, Sundar’s dedication to the company’s success motivated him to deeply understand marketing’s role in achieving business objectives. This understanding wasn’t driven by an innate love for marketing itself but by a genuine curiosity about customer needs and business impact. He regularly engaged with Uber drivers during his commutes, gathering firsthand insights about their experiences and challenges.

This curiosity led to meaningful improvements in Uber’s driver retention strategy. After learning that new drivers struggled with their first trips during busy Friday nights, Sundar analyzed the data and discovered significantly lower retention rates for drivers who started during peak hours. By testing a simple change, recommending Saturday afternoon starts for new drivers instead of Friday nights, the team achieved a substantial improvement in driver retention rates.

The key to finding technical talent who embrace marketing lies in identifying individuals driven by curiosity and customer obsession. These professionals can successfully transition between different roles and teams because they’re motivated by understanding and solving customer problems, regardless of the technical or marketing context. This adaptability stems from their genuine interest in how their work impacts the end user, rather than just the technical aspects of implementation.

Key takeaway: Technical professionals most successfully engage with marketing when driven by genuine curiosity about customer problems and business impact. Organizations should focus on finding talent who demonstrate customer obsession and natural curiosity, as these traits often predict success across both technical and marketing roles.

Back to the top ⬆️

Managing Work Life Balance Without Burning Out

Life has a way of teaching its most profound lessons through unexpected moments. For Sundar, this revelation came through a series of personal losses, three deaths within three weeks, spanning ages from 33 to 99. These experiences fundamentally shifted his perspective on work, life, and the pursuit of happiness. Rather than focusing on managing stress after it occurs, he advocates for a proactive approach to prevent stress from taking root in the first place.

At the core of Sundar’s philosophy lies a simple but powerful distinction between controllable and uncontrollable elements in life. He emphasizes the importance of giving your best effort to things within your control while developing resilience toward factors beyond your influence. This approach has helped him maintain a positive outlook, viewing happiness not as a constant state but as a long-term trend, similar to investment strategy in stocks. Short-term fluctuations are inevitable, but consistent investment in oneself and meaningful relationships yields lasting returns.

This perspective led Sundar and his family to make bold life changes, including relocating to Amsterdam. Instead of overthinking decisions, they followed their energy and embraced new opportunities, discovering unexpected levels of happiness along the way. This willingness to challenge conventional wisdom and revisit fundamental assumptions has been crucial to their journey. Growing up in an Indian household where career traditionally took precedence, Sundar has consciously chosen to prioritize family over work, finding that this clarity of values enhances rather than diminishes professional performance.

The unpredictability of life has taught Sundar to be selective about where he invests his energy. Rather than viewing life as simply “too short,” he sees it as too unpredictable to waste on pursuits that don’t bring genuine fulfillment. This realization has led him to establish clear priorities and make deliberate choices about how he spends his time and energy, recognizing that unhappiness in one area tends to affect all aspects of life.

Key takeaway: Success and happiness stem from preventing stress rather than managing it, maintaining clear priorities, and investing energy in activities that truly matter. Focus on controlling what you can, building resilience for what you cannot, and making conscious choices that align with your values rather than societal expectations.

Back to the top ⬆️

Episode Recap

Effective marketing measurement is about the courage to challenge assumptions, the wisdom to interpret data meaningfully, and the strategic acumen to translate insights into action. Marketing thrives on assumptions, but what happens when data shatters them? At Uber, Sundar’s team uncovered a startling truth: their Meta advertising campaigns, despite consistent spend, generated zero incremental value in new user acquisition. This discovery, validated by Meta’s own data science team, led to $30 million in annual savings across the U.S. market.

This revelation catalyzed a deeper examination of marketing fundamentals. Traditional metrics like lifetime value calculations proved less valuable than simple behavioral patterns; a customer’s first week of activity reliably predicted their two-year value. Geographic testing demonstrated how brand awareness directly impacted business outcomes, while propensity matching offered new insights into customer behavior without requiring control groups.

The transformation extends beyond metrics. Marketing departments, once commanding respect without validation, now face unprecedented pressure to justify every investment. This shift, driven by tech companies’ data-focused culture, has paradoxically strengthened marketing’s position by revealing the quantifiable impact of both performance and brand initiatives.

Yet the most valuable insights often emerge from combining technical expertise with human understanding. When Uber’s team engaged directly with drivers, they discovered that recommending Saturday afternoon starts instead of Friday peak hours significantly improved retention, an insight no algorithm could have surfaced. This exemplifies modern marketing’s core challenge: balancing rigorous measurement with strategic thinking and customer empathy.

Listen to the full episode ⬇️ or Back to the top ⬆️

Follow Sundar👇

• LinkedIn
• Newsletter

✌️

—
Intro music by Wowa via Unminus
Cover art created with Midjourney (check out how)

Apple •Spotify• Pocket Casts •Youtube •Overcast •RSS

Related tags

See all episodes

Future-proofing the humans behind the tech

Apple •Pocket Casts•Google •Overcast •Spotify •Breaker •Castro •RSS

• LinkedIn
• Twitter
• Instagram
• YouTube
• RSS Feed