PRD: Increasing Customers Reviews on Food Delivery

Team:Ratings & Reviews (Food Delivery)

Contributors: Mugdha

Status: In Development

Launch Date: TBD

PROBLEM DEFINITION

For any Reviews and Ratings play a huge role in deciding which restaurant to order from. Even though a customer might not go completely go in-depth with checking the same before ordering, they would hesitate ordering from a place with no or less reviews/ratings. 

Despite having a MAU base of 24Mn, and servicing an average of 1Mn orders/ day, the % of reviews/ratings left by the customers is as low as 5%, which averages only to around 50k reviews/ratings in a day. Amongst the rest, there is a substantial base of approx. 0.6Mn who engages with the review and rating feature on the app, but only before placing an order. 

User research suggests that the customers, who in their entire journey on the platform engage with the feature, find the process long and time-taking. 

If even 5% of the customers from each of these segments can be motivated to leave a review, that would mean an incremental of 12k reviews/day. 

This would help in the platform filter out restaurants offering poor service, which in the end will improve the overall experience that the customer has with their food. 

GOALS

• Functional Metrics:

1. Number of Orders/Day
2. Number of Unique Users Leaving a Review
3. Number of Reviews/Total Number of Orders in a Day
4. Click-through Rate (CTR) on the Prompts suggested
5. Click-through Rate (CTR) on Review Button

These metrics will help us understand the impact the reviews is having on the number of orders. 

• Non-Functional Metrics:

1. System Performance: Impact of increased reviews on system load and performance.
2. Review Processing Time: Time taken to process and display reviews.

NON-GOALS

We’re not focusing on onboarding new users as that is not the direct aim here. 

VALIDATION OF THE PROBLEM

To validate this problem, we have conducted user research and gathered insights from surveys and competitive analysis.

     User Research Insights:

1. Users find the review process long and time-consuming.
2. Customers are motivated to leave reviews based on strong experiences (positive or negative).
3. Some users feel their reviews may not make a significant impact.
4. Privacy and usability concerns deter users from leaving reviews.

    Competitive Landscape Insights:

1. Competitors like Swiggy, Amazon, and Flipkart have streamlined review processes that encourage user participation.
2. Zomato allows users to rate each item separately and rate restaurants even without prior orders, unlike competitors.

THE TARGET AUDIENCE

• Size: 0.6Mn customers who engage with reviews/ratings but do not leave them.
• Personas: User wants to find reliable restaurant information, and share their food delivery experiences. But they find the process time-consuming review process and are uncertain about what to write. Moreover, they are concerned about their privacy.

SOLUTION

We are implementing two main solutions to address the problem:

1. Suggesting Review Prompts:

• Implement AI-driven prompts to guide users in leaving reviews.
• Use machine learning to suggest contextual prompts based on order details.
• Introduce visually appealing prompts to make the process engaging.

    User Flows:

1. Initial Prompt: Users are prompted to leave a review after an order is delivered.
2. AI Suggestions: Contextual prompts are provided based on the order's cuisine, delivery time, etc.
3. Review Submission: Users can choose from suggested prompts or write their own reviews.

      2. Making the Review Button More Prominent:

• Enhance the visibility of the review button on the app and website.
• Use design changes to ensure the review button stands out and encourages more clicks.

LAUNCH READINESS

Key Milestones:

1. Design Complete
2. Development Complete

Launch Checklist:

• Support: Plan for customer support to handle new review features.
• Operations: Ensure operations team is ready for increased review volume.
• Stakeholder Communication: Keep all internal stakeholders informed.

Experimentation Plan:

• Conduct A/B testing to measure the impact of AI-driven prompts and prominent review button on review participation.

OPEN QUESTIONS & DECISIONS TAKEN

Open Questions:

• How to address privacy concerns effectively?
• What additional features can further enhance user engagement?

Decisions Taken:

• Focus on existing user base rather than onboarding new users.
• Prioritize development of customizable and contextual review prompts.

Descoped Items:

• Onboarding new users to the platform.

Trade-offs Made:

• Delaying additional features to focus on core review prompt enhancements and prominent review button implementation.