Should Liftoff Mobile prioritize expanding its machine learning capabilities for better ad targeting or focus on improving its real-time bidding system for faster ad delivery?

Introduction

The trade-off between expanding machine learning capabilities for better ad targeting and improving the real-time bidding system for faster ad delivery is a critical decision for Liftoff Mobile. This scenario involves balancing technological advancements with operational efficiency in the mobile advertising space. I'll analyze this trade-off by examining the product context, potential impacts, key metrics, and experimental approaches to inform a strategic recommendation.

Step 1

Clarifying Questions (3 minutes)

• Context: I'm thinking about Liftoff's current market position. Could you share how we're performing against our main competitors in terms of ad targeting accuracy and delivery speed?

Why it matters: Helps prioritize which area needs more immediate attention Expected answer: We're lagging in targeting accuracy but competitive in delivery speed Impact on approach: Would lean towards prioritizing ML capabilities

• Business Context: Based on our revenue model, I assume we charge on a cost-per-action basis. Is this correct, and how significant is the impact of improved targeting on our revenue?

Why it matters: Determines the financial implications of our decision Expected answer: Yes, CPA model with significant revenue impact from targeting improvements Impact on approach: Would strengthen the case for ML investment

• User Impact: Considering our advertiser segments, are we seeing any particular challenges or requests regarding targeting capabilities or ad delivery speed?

Why it matters: Aligns our focus with customer needs Expected answer: Enterprise clients pushing for better targeting, while smaller advertisers prioritize speed Impact on approach: Might suggest a segmented strategy or phased approach

• Technical: Given our current infrastructure, what's the estimated time and resource investment needed for significant improvements in ML vs. real-time bidding?

Why it matters: Assesses feasibility and opportunity cost Expected answer: ML improvements require longer-term investment, while RTB upgrades are more immediate Impact on approach: Could influence short-term vs. long-term strategy balance

• Resource: How does our current team composition align with these two potential focus areas? Do we have more ML experts or RTB specialists?

Why it matters: Evaluates our capacity to execute in each area Expected answer: Stronger in RTB expertise, but growing ML team Impact on approach: Might suggest leveraging current strengths while building capabilities