3. Case Study - Optimizing the Order of Service Requests at GetNinjas

This case study focuses on optimizing the order of service requests in the GetNinjas platform to enhance user experience for professionals and improve overall platform performance. The case provides insights into the challenges faced, hypotheses tested, data analysis conducted, and final solutions implemented to maximize the effectiveness of the marketplace.

Understanding the GetNinjas Platform

GetNinjas is a marketplace connecting clients who need various services with professionals who can provide them. The platform operates primarily through a mobile app for professionals. Here's how it works:

  1. For Clients: Clients submit a service request by providing details about the service they need and their contact information.
  2. For Professionals: Professionals view a list of service requests in an app-based "store." Each request is anonymized, with client contact details hidden. Professionals purchase leads (i.e., the right to contact the client) using virtual credits, similar to a prepaid model.

The challenge in this case was to determine the best way to prioritize and order the service requests in the store for professionals. By improving the ordering logic, GetNinjas aimed to increase lead purchases, improve service delivery, and boost overall platform performance.

Objectives and Hypotheses

The initial analysis revealed that 50% of revenue is generated through the store, with the remaining 50% coming from push notifications. This finding highlighted the importance of optimizing the store interface to improve the professional experience and increase revenue.

The goal of the optimization was to maximize the Gross Merchandise Value (GMV), which represents the total transaction value between clients and professionals. Simply increasing lead purchases was not enough; it was crucial to ensure that professionals purchased valuable leads that would result in completed services and satisfied clients.

Several key hypotheses were developed to guide the analysis:

  1. Professionals’ browsing behavior: How far do professionals scroll through the store? Are they only interested in the first few listings, or do they explore deeper?
  2. Factors influencing lead purchases: What factors make a professional more likely to purchase a lead? Hypotheses included distance between the professional and the client, the level of competition (number of other professionals who have purchased the lead), and the urgency of the service.
  3. Impact of urgency: It was believed that urgent service requests (e.g., needing a service the next day) would be more attractive to professionals than less urgent requests (e.g., a service needed in three months).

Data Tracking and Analysis

To validate these hypotheses, a detailed tracking system was implemented to collect data on professional interactions with the store. The tracking system captured information such as:

With this tracking in place, the team was able to gather actionable insights about professional behavior and preferences.

Key Findings

The data analysis provided several important insights:

  1. Scrolling behavior: Approximately 55% of all service request views occurred within the first 10% of the store's listings. This indicated that professionals typically did not scroll far down the list, reinforcing the importance of optimizing the order of requests to display the most relevant ones at the top.

  2. Frequency of visits: Most professionals visited the store no more than three times per day, with each session lasting around 20 minutes. This data helped the team understand when professionals were most active and what their browsing patterns looked like.

  3. Competition: Contrary to initial expectations, competition (i.e., how many other professionals had already purchased the lead) was not a significant deterrent. Most professionals were willing to compete with others, particularly if they felt confident in their abilities to close the deal. However, in certain categories, like technical assistance, competition was more of a concern.

  4. Urgency: The urgency of the service request proved to be a crucial factor in determining lead purchases. Requests with a higher urgency (e.g., needing a service within the next few days) had a significantly higher purchase rate. Around 70% of the requests were classified as urgent, and these requests saw a much higher conversion rate.

  5. Distance: The distance between the client and the professional also had a significant impact on lead purchases. Leads for clients located within 5 kilometers of the professional were 3 to 4 times more likely to be purchased compared to leads for clients located more than 30 kilometers away.

Final Solution: Two-Stage Ordering System

Based on these findings, the team developed a two-stage ordering system for the store:

  1. Stage 1: Grouping by Distance: Service requests were grouped based on the distance between the client and the professional. The closer the client, the higher the request would appear in the list.
  2. Stage 2: Ordering by Urgency: Within each distance group, requests were further ordered based on their urgency. Requests that needed immediate attention would appear higher in the list than those that could be delayed.

By combining distance and urgency as the primary ordering factors, the team was able to create a more relevant and attractive list of service requests for professionals. This approach ensured that professionals saw leads that were both geographically convenient and time-sensitive, leading to higher engagement and more successful service completions.

Conclusion

This case study demonstrates how a data-driven approach can lead to meaningful improvements in product performance. By analyzing professional behavior and testing various hypotheses, GetNinjas was able to optimize the ordering of service requests in its store, resulting in a better user experience and increased platform revenue.

Key takeaways from this case include: