Answering PM Interview Question - Ola notices a decrease in driver availability in peak hours
Let's look at how we can answer a real product management interview question around execution and root cause analysis.
Today we will answer a product management interview question that was asked in the Senior Product Manager interview round for a ride-hailing company.
Before jumping to the solution, please also check ProductHope practice interview questions which has 50+ interview questions asked in real product management interviews across product design, strategy, execution and analytics question which can influence your thinking - Click Here
Question: Ola notices a decrease in driver availability in peak hours, what factors might be contributing to this and how would you address the issue?
“The interviewer also mentions that you are free to take some valid assumptions while going through the problem”
Step 1: Clarify the problem and approach
Let’s start by asking clarifying questions to understand the problem better and make some valid assumptions
To clarify, are drivers logging off during peak hours but remaining active during non-peak times, or is there an overall reduction in availability? - “Let’s assume the interviewer mentioned they are just logging off during peak hours”
Can you specify the exact time frames we're considering as peak hours for this issue? I assume it’s 9 - 10:30 am & 6 - 7:30 pm considering a significant % of users use Ola for office commute?
<pause, take time - “ask the interviewer if you can take a min or two to think & structure your answer>”
Now that you have understood the problem better and clarified your approach to the interviewer -
I’d first analyze the data to confirm the trend and understand its specifics of the problem and accordingly build my hypothesis.
Metrics: Data-related, attribution changes
Did we change the logic of attribution or definition of metrics of driver availability?
Timeline: Is it a sudden drop or trend? Does it look like a seasonality issue? What's the magnitude of the decrease?
Demographics: Is this trend consistent across all regions or specific to certain areas? Let’s assume it’s happening in Bengaluru.
Product or Tech: Are there any new releases? Did we make any new features or experiments live which could have impacted the availability of drivers?
Change in incentive structure or commissions: Did we make any changes in any of the programs?
After looking at all the data, I will state my assumptions -
Let's assume our data shows a 15% decrease in driver availability during weekday evening rush hours, from 5:30 PM to 8 PM, primarily in tier 1 cities like Bengaluru and we have not made any changes to incentive structure or commissions.
Step 2: Understand the potential root cause
Next, I will try to dive deep into potential factors that could influence this. For this, I'd like to survey and research either by speaking to some of the drivers or asking them to share their thoughts on a survey. I will carefully select the users who show the behaviour of not being available during peak hours but are available at other times to understand the root cause and build my hypothesis. Parallelly, I will also try to explore more external factors that could have influenced this.
Economic or Competition: Did any new competitors update their offering or incentive structure?
Driver frustration: Heavy traffic during these hours could be leading to fewer completed rides and lower earnings.
Step 3: Building Hypothesis
Post this, I will build my hypothesis based on qualitative analysis. In certain cases, I will also try to validate by qualitative analysis with quantitative data points as well. I believe based on my analysis from customer surveys, these are some of the potential reasons -
Competitors offering better incentives: New competitors are offering better incentives and commission structure which has shifted the driver behaviour of using that platform within peak hours while they are not able to provide a similar no of rides during non-peak hours that’s why overall availability has not decreased.
Heavy traffic leads to driver frustrations: Drivers are frustrated with traffic and prefer driving during non-peak hours. Considering they also mentioned that the price they get from rides during this time doesn’t even suffice their petrol or diesel cost.
Other factors:
Seasonal or Environmental factors: Extreme weather during typical peak hours or seasonal factors around new gig opportunities that lead some drivers to either shift or not be active on the platform.
Personal commitments and behaviour change: Some drivers have personal commitments or now preferring part-time option vs full-time. [Although we should not consider it since we have already clarified in the beginning that the overall availability of drivers has not been impacted, it just shifted from peak to non-peak hours]
<mid-check in: pause and check with the interviewer if you are on the right track or not. you can ask like - does it make sense?>
Step 5: Prioritise and start brainstorming initiatives
Out of these, let’s focus on one or two reasons - competitors offering better incentives or driver frustrations due to traffic.
Now, I will go deeper into each of these reasons and think of some solutions or initiatives and prioritise based on simple frameworks
Solving for a competitor incentive offering
Key Motivations for drivers to switch to competitor platform or what are the value propositions during peak hours
Higher earnings potential: Competitors might be offering significantly better commission rates or bonuses during peak hours.
More transparent incentive structure: Competitors may have clearer, easier-to-understand bonus systems for peak hours.
Lower platform fees: Rival apps might be taking a smaller cut of the fare during high-demand periods. For eg: Rapido introduced a new model where they made 0% commissions instead charge a weekly or monthly subscription fee.
Since our riders are not completely moving out from the platform, we could make certain improvements in our product or incentive structure to re-gain the active drivers.
<i’ve not worked on Ola so i’m not sure if these features already exists today on the Ola driver app>
Few ideas -
Dynamic Earnings Forecast: Implement a predictive earnings calculator that shows potential earning opportunities when a driver attempts to go offline during or just before peak hours.
This feature would display a compelling visualization of potential earnings for the next 2-3 hours, based on historical data, current or predicted demand, and predicted surge pricing.
The forecast would also highlight any special bonuses or incentives available during the upcoming peak period. This "last chance" reminder aims to persuade drivers to reconsider logging off by clearly demonstrating the financial opportunity they'd be missing.
Lowering commissions for peak hours: Introduce a tiered commission structure where Ola takes a smaller percentage of the fare as demand increases during peak hours. For instance, the normal commission might be 20%, but during moderate peak times, it could drop to 15%, and during the highest demand periods, it could be as low as 10% potentially close to competitor offerings. The reduced commission directly increases driver earnings without necessarily raising prices for riders, making it a win-win scenario that encourages drivers to stay active during crucial high-demand periods.
New incentive structure or reward program of peak hours: Develop a gamified rewards system specifically for peak hour participation. Drivers can earn a monthly incentive for completing certain number of rides during peak hours.
< Mid-check-in: Always try to do a quick check-in with the interviewer to keep them engaged with your ideas. It helps to keep your answer on track with the interviewer's expectations. In case of any follow-up questions, you can answer them there without losing track. Also, keep track of time while answering the question; that's why communicating your thought process helps.>
Evaluating the feature ideas before prioritisation
It's good to showcase that you are prioritizing the features not only based on effort vs. impact but also thinking deeply about potential upsides and downsides before finalizing the experiments.
Potential upsides
Dynamic Earnings Forecast
Increased driver retention during peak hours
Better supply-demand matching
Lowering commissions for peak hours
Significant boost in driver earnings and availability
Competitive advantage over other platforms
New incentive structure or reward program of peak hours
Flexible, cost-effective incentive structure
Increased driver engagement and loyalty
Potential downsides or risks
Dynamic Earnings Forecast
Risk of overestimation leading to driver disappointment
UX impact for users who genuinely want to go offline during peak hours.
Lowering commissions for peak hours
Direct impact on company revenue per ride
Risk of drivers only working during peak hours
New incentive structure or reward program of peak hours
Risk of drivers trying to game the incentive system (low probability).
Looking at both the upside and the downside, all three initiatives can be carried out as experiments in one city/area and expanded based on the success of the experiment.
Step 6: Summarise and answer follow-up questions
Once we get into the final set of initiatives, it's always better to summarize the approach. To summarize: we started by understanding the root cause of the drop by looking at key metrics. Looking at the data, it's assumed that there is a drop of around 20% in driver availability during peak hours in urban areas like Bengaluru. Then, we can conduct customer surveys to understand the exact reasons and potential motivations of drivers. While competitors offering better incentive structures seem to be the main factor, we listed some initiatives to tackle this issue. Lastly, the idea was to also look at some upsides and downsides of each of these initiatives and prioritize accordingly, focusing on the key metrics.
Follow up questions
Generally, if time permits, the interviewer wants to deep-dive into one of the solution ideas with key metrics to track while keeping a check on control metrics. It often turns into a product-design question like: "How would you design a dynamic earning forecast to minimize the downsides mentioned?"
We will try to cover that in an upcoming email newsletter
With this, we would also like to mention that we have launched a question bank for product manager interview questions - https://theproducthope.com/product-manager-practice-interview-questions/