Improving customer visits to co-living properties by OYO Life

A case study on how reducing the cognitive load to make a selection, lead to 100% plus uplift in conversion.

Arash Bal
5 min readFeb 2, 2021

--

About OYO Life:

OYO Life offers long-term affordable housing for millennials and young professionals. The properties are equipped with essential amenities like Wi-Fi connectivity, television, refrigerator, furnishings, AC, regular housekeeping, power backup, CCTV surveillance, and 24/7 caretaking.

What’re we trying to solve?

In India, the majority of people prefer visiting the property to verify the property, living conditions, observe neighborhood, amenities near-by the property, etc. The C1 metric (Unique Visits Scheduled / Online Users) i.e. the percentage of users scheduling visits to properties, was running below par even after easing of COVID restrictions. With thousands of users visiting the website daily, there is a tremendous potential for improvement.

My Role: Product Designer
Designed in August 2020 & launched in mid-October 2020
Geography: India

Let’s quickly go over the user journey:

Illustration by: https://www.shutterstock.com/g/shlyapanama

Fun Fact: A typical user spends around 20 minutes during the physical visit on the property.

Existing Solution

But what’s wrong with the existing solution?

The existing solution was a result of incremental changes since the launch of OYO Life. Few problems:

  1. In the SAV (Schedule a Visit) user journey flow, significant drop-offs were observed. Only 9% of the traffic that initiated visit creation ended up scheduling a visit.
  2. We were fetching mobile number post date/time selection. With the customer contact info, we had no option to follow up with the customer with alternative marketing channels (e.g. call center, offers through text messages).
    This was a bigger problem since 70% of the users are either new users or with no active session.
  3. For a new user, an additional option step popped up for his/her name & email.
  4. The humungous banners which were initially introduced to communicate the benefits of OYO Life were diluting the focus from the data input.
  5. There was a highly skewed date/time selection trend observed in the visit scheduling journey. For visit slot selection, we were offering a wide window of 30 days, 12-hour slots to choose from.

Solution 1: Reduce drop-offs

Let’s see where major drop-offs were happening:

Fetching the phone number of customer is a mandatory step in our flow. Why is it mandatory? To ensure the safety of existing tenants of the property.

We made a conscious move to fetch the phone number before the visiting slot. During the design testing phase, we saw high intent users went ahead with the further flow as well with only a few drop-offs, whereas low intent users dropped off before the first step only. The logged-in users automatically by-passed this step, reducing their journey.

Revised flow:

Solution 2: Simplifying visit date/time slot selection

In order to simplify slot selection, we wanted to reduce the cognitive effort on the users’ end. While talking to internal customers, we saw a pattern of responses to when can the customer visit the property.

Few Observations:

  1. 80% of visits were being scheduled for the upcoming 4 days. (57% for the same day)
  2. 73% of the visits were scheduled between 12 pm & 7 pm.
    From 9 am to 12 pm: 17% of visits
    From 12 pm to 4 pm: 35% of visits
    From 4 pm to 7 pm: 38% of visits

    Customers can schedule a visit from 9 am to 9 pm.

For date selection:
From showcasing the full month calendar, we changed the stance to bringing just the next week upfront.

For time slot selection:
From having 12 one-hour slots from 9 AM to 9 PM, we grouped the time slots by morning, noon, evening, and late-evening. Instead of a traditional dropdown with numerous time slots, we improved the visibility by bringing all 4 slots upfront in form of segmented control.

Solution 3: One-click visit scheduling

Based upon the trends of a city, we started to auto-assign a recommended default date/time slot for visit. e.g. in Gurgaon people prefer visiting properties during lunchtime, in this scenario we auto select 12–3 pm same day slot. Customers can easily select another date/time as per convenience.

This worked exceptionally and below are the results for the same:

Designs:

Results

(Changes were released over the fourth quarter of 2020)

  • The C1 metric percentage (Unique Visits Scheduled / Online Users) has seen a 100+% uplift owing to a host of concerted efforts with a significant contribution coming from these changes.
Screenshot of improvement from Google Analytics (C1 metric exact numbers are confidential)
  • The drop-offs post schedule a visit flow initiation has been cut significantly, with the incremental traffic at completion seeing an 85+% uplift in contrast to the earlier trend.
  • The page performance has seen marked improvements. For instance, the property detail page, as well as the home page load time, has improved by ~50%, the visits APIs’ performance has witnessed a ~3x improvement with latencies down significantly. Kudos to the backend team.
  • 58% of the converting users are going with a one-click schedule a visiting model by opting for the default slot.

Team:

Product: Shuja Naik, Vipul Srivastava
Tech: Sai Bharath Gutti, Priyanka Walia, Madhav Mehta, Siddharth Verma, Tharun Raj Soma, Shiva Reddy, Vishnu Vardhan & Vyomkesh Tiwari
Data Science: Aakash Mittal & Vipul T
Design: Arash Bal

Learnings

Sometimes even the smallest changes can create a big impact and this was one of those projects.

Thanks for reading.

Please leave your feedback :) Reach out to me on: hi@arashbal.com

--

--