Introducing search and improving discovery of doctors on MFine

Introducing search and improving discovery of doctors on MFine

💊

TL;DR

The project was intended at increasing the overall consult conversion by improving discovery experience of the doctor list page. We introduced search and previously consulted doctors for finding the doctor of choice. We also introduced capability to filter by experience, language, doctor fees, location, availability and gender, and sorting based on experience and doctor fees to narrow down on a doctor as per the user’s preferences.

Objective

Improve the consult conversion from the doctor listing page by helping users in finding the right doctor for their symptom(s) or condition, as per their preferences.

My Role

I collaborated closely with the product manager, UX writer, engineers, business analyst and stakeholders throughout the project. I conducted user research, competitive analysis and designed the entire user journey, various UI elements, interactions and micro-interactions.

What was done?

Based on the insights from user research, competition analysis and existing data, we made the following changes to the doctor listing page:

Improved the filters and sorters
Introduced doctor search
Introduced previously consulted doctors

Impact

Overall consultation conversions increased by 12p.p (from 40% to 52%).
Repeat user conversions increased by 5p.p (from 38% to 43%).
Filter abandonment reduced by 49p.p (from 55% to 6%) and sorter abandonment reduced by 20p.p (from 53% to 33%).

The Problem

The existing doctor list page consisted of a list of doctor cards, basic filtering (by speciality and by hospital) and basic sorting (by experience and by earliest available doctor). This made it very difficult for the users to find the right doctor for their consultation. To get a deeper understanding of this problem, we looked into existing usage metrics and talked to a few users.

Old doctor list page.

Insights from data

~60% of the users drop from the funnel at the doctor listing page.
Only 13% of the users clicked on existing sort option and 53% of them abandoned it without making a selection.
Only 17% of the users clicked on existing filter option and 55% of them abandoned it without making a selection.
People who used either sorter or filter had a higher conversion rate of 43% (vs 37% otherwise). People who used both sorter and filter had a conversion rate of 48%.
25% of the repeating users comeback for the same speciality. Within them, 35% of the users pick the same doctor, while the rest 65% picked another doctor. The 35% of the users who picked the same doctor had a click depth of 61 doctor cards while 65% of the users who picked another doctor had a click depth of 48 doctor cards in the listing page.

Learnings

💡

Improve filters and sorters

It is clear that existing filters and sorters doesn’t align with the user expectations.

💡

Better conversion with filter and sort

People using filters and sorter had a better conversion rate over those who did not.

💡

Consult same doctor again

There is strong intend from repeat users to consult with the same doctor.

Insights from user research

We talked to 20 users, homogenised between hyperlocal and non-hyperlocal locations. We validated the insights of this research through a survey that was answered by 400 users.

The major insights from the research were:

Years of experience and consultation price were the most important factors to decide on a doctor for consultation.
Language was a bigger criteria than location of the doctor. Users preferred doctors who speak their native language. In fact, location is often used as proxy for language.
People had varied expectations from doctor’s availability. While some users wanted to consult with a doctor immediately, other wanted to book an appointment for later in the day or next days.
Non-hyperlocal users showed strong affinity towards Hospital brands like Fortis, Apollo etc. These users were also ok with paying higher amount for consultation if the doctor is from a reputed branded hospital.
Repeating users preferred consulting with a previously consulted doctor.
Sometimes users come in search for a specific doctor through referrals or word of mouth.

‣

View select responses

Learnings

💡

Experience matters most

Experience and price are the most important factors.

💡

Previously Consulted Doctors

Repeat users preferred previously consulted doctors over new.

💡

Native language preferred

Users prefer doctors who speak their native language

💡

Brand Value

There is an affinity towards branded hospitals like Apollo, Aster, etc.

Based on these insights, and competition analysis, we brainstormed for solutions and ideal experience. The following features were introduced after several discussions, reviews and design iterations.

Doctor Search

Micro-interaction on search box

Users landing on the doctor listing page can now search for

Doctor names
Speciality
Language
Location
Hospitals
Conditions
Symptoms

You can view the user flow of search implementation on the video

User flow of doctor search

Other Scenarios

If the doctor searched for is currently unavailable, it is shown as greyed out card at the bottom of the search list

If the doctor searched for does not exist in the system, a no result page with other available doctors are shown.

If the doctor searched for is not available for the day, the CTA is different. List is sorted by earliest available.

When the searched doctor is available only in a different speciality, they are shown with a warning informing change in speciality

If user searches for a common word or symptoms, doctors from the relevant speciality are shown with an info card explaining the context.

If user searches for a language, the order of languages in the doctor card is dynamically sorted to show the searched language first.

Filters

Users landing on the doctor listing page can now filter the doctor list by

Experience
Language
Location
Availability
Speciality
Gender
Hospital

Characteristics

All filter sub-options show the number of available doctors even before applying. If there are no doctors available, the option will be disabled.
The options within a filter is sorted based on the number of doctors available.
When users applies two or more filters at once, the filters form an AND conditions. For eg: If I pick language as Bengali and location as Bengaluru, I will see doctors in Bangalore speaking Bengali. The count of doctors under the filter options will also update dynamically.
Filters with a long list of options like location, speciality, etc will have a search within the filter itself.

Experience Filter

Language Filter

Doctor Fee Filter ( currently removed from production)

Location filter with search

Doctor availability filter. This filter dynamically disables based on the time of the day.

Speciality filter with search

Gender Filter

Hospital filter with search

Confirmation pop-up if users closes filter without applying.

Sorter

Users landing on the doctor listing page can now filter the doctor list by

Earliest available near you
Earliest available across India
Nearest doctors
Most experienced
Doctor fees - high to low
Doctor fees - low to high

List of all sort options available

Some of the sort options will be disabled if conflicted filters are applied. For eg: If you are a user in Bangalore, applying location filter as Hyderabad, nearby doctor sorting will not be available.

Previously Consulted Doctors

Eligible users landing on the doctor listing page can now find a quick filter to see previously consulted doctors only. The position of this quick filter can be configured from the backend and be extended to any other filter as well.

Characteristics

Previously Consulted Doctor filter is available only for repeat users who have consulted previously.
The filter when applied will only show doctors from the chosen speciality, and not all the previously consulted doctors.
The doctor card has been modified with a tag for doctors who have been consulted previously.
After a lot of debate and brainstorming, we decided to keep previously consulted doctors as a filter instead of a section with in the page.