This time Uber After the first wave of OA in 2026 Summer Growth Intern was released, the overall feedback was actually a bit more "hardcore" than everyone imagined. Many people think that Growth is more about business intuition, but in practice, you will find that it is a combination of data analysis + business understanding + tripartite market logic. If I could sum up the intensity in one sentence: it’s not that the questions are tricky, but that the pace is fast, there are many dimensions of thinking, and the error tolerance is low.
Below is a complete disassembly process.
Online Assessment (75–90 minutes)
The entire OA is divided into two parts: Quantitative Analytics and Strategic Case, using Uber’s internal system. Time control is very important. The score will not be displayed immediately after completion, so stable output during the process is even more critical.
Funnel Analysis
The first question is based on NYC rider sign-ups data, asking you to find out whether the largest drop-off occurs in Acquisition or Activation.
This question looks like basic funnel analysis, but the real test is whether you can quickly build an AARRR framework and distinguish the essential difference between "registration churn" and "first order churn". Many people will directly calculate the ratio, but do not explain why the loss of a certain step has a greater impact on long-term LTV.
The Growth position pays more attention to your understanding of the "economic significance" of each step of the funnel, rather than simply calculating the correct percentage.
Market Elasticity
The question is set in a high-density market like San Francisco. If the booking fee is increased by 10%, how will it affect the balance between driver supply and rider demand. This question is essentially a deduction of marketplace equilibrium.
You need to discuss rider price elasticity, driver income changes, platform commission structure, whether surge is triggered, and whether long-term supply and demand will be rebalanced. Simply saying "price increases and demand decreases" is far from enough, because this is a three-party dynamic game, not a unilateral demand model. People who answer well will take the initiative to mention the difference between short-term shocks and long-term adjustments.
LTV&CAC
Taking London's new rider acquisition campaign as the background, given a monthly churn of 5%, it is required to calculate the payback period and LTV/CAC ratio.
This question is a standard unit economics test. The key points are whether to use 1/churn to approximate the life cycle, whether to consider contribution margin, and whether it is clear whether payback is calculated based on gross profit or revenue. Many people get stuck on formulas, but really high-scoring answers explain the assumptions rather than mechanical substitutions.
Experimentation
Design an A/B test of “Uber Green” loyalty reward and explain the North Star metric.
The focus of this question is not on writing the process, but on whether you understand the experimental goals. Is it to increase the proportion of green trips? Or improve overall retention? Will cannibalization occur? Do you need guardrail metrics to monitor overall order volume and driver experience? The experimental design of the Growth post must take into account both growth and health, rather than just focusing on one indicator.
Stakeholder Management
How to convince the Product team to prioritize referring features instead of UI redesign. This question tests influence, not debating ability. You need to use data to prove the potential of Referral to reduce CAC, use experimental results to reduce controversy, and use incremental lift to quantify the opportunity cost. Growth is essentially a battle for resources, and the logic must fall on ROI.
Virtual Technical Interview
The format is a 45-minute Case Interview, typically conducted via Zoom. The core revolves around the understanding of the three-party market of Rider, Driver and Eater. Many problems will start with data anomalies in a certain city, allowing you to investigate the cause. For example, if orders decline, will you look at the demand side or the supply side first? Is it a price issue or a matching efficiency issue? Is it competitor shock or seasonal fluctuations? The biggest fear here is that the ideas will diverge. The structure must be clear, otherwise time will be consumed quickly.
Final Round
Finally, there will be two rounds of interviews with the Growth Lead, focusing on behavioral and business intuition. I will dig deep into your past growth experiments and often ask: How would you optimize if you did it again? What if the data doesn’t support your hypothesis? If the interests of the driver and rider conflict, which side do you give priority to? This round tests maturity and judgment more than numeracy.
Write at the end
If you don't have confidence in OA now, or don't know where to prepare the system, you don't have to do it hard. It’s not that many people don’t know how to do it, but they don’t know how to examine the logic and output structure. Once the direction of practice goes astray, the more they practice, the more anxious they become. We have been organizing real question banks from major manufacturers for a long time, and we also provide Real-time assists The service provides you with thought reminders and structural guidance when you are doing OA or Case training, helping you to stabilize your rhythm instead of messing up on the spot.
