Product Teardown: Apna Appš±
šĀ Hi, Iām Harkirat Singh. I write Inside Startup, a newsletter to help Startup founders and Product growth managers to accelerate their product growth through GTM strategy, growth and marketing experiments.
š„Ā Donāt miss out on interesting product & growth teardowns in your inbox every Monday and Friday - just enter your email here to sign up:
šØImportant Message: This project was completed in Marchā22. And it didnāt reflect the new changes made within the Apna app.
ā¹ļø About Apna App
Nirmit Parikh (founder of Apna) first saw the problem in hiring skilled workers at his family-owned manufacturing businesses. Parikh spoke with employees and went undercover as an electrician and a floor manager to have a better understanding of the problems they faced.
Soon he realised that he could use technology to help Blue-collar workers find genuine work quickly and close to their homes. He left Apple and returned to India in 2019 to launch Apna, a professional networking and upskilling platform.
Once on the platform, job seekers enter their personal information which is converted into a virtual business card and distributed to potential employers.
The algorithm on their āJobsā tab shows openings with a short description and the skills that are required for relevant candidates. The candidate needs to score well in the screening assessment tests (Normally 2-4 questions) before the app allows them to contact the employer.
After the candidate successfully connects with the recruiter, the company claims to complete the hiring process in less than 48 hours. This way job-seekers can not only find jobs but also gigs relevant to their skills.
To date, Apna witnessed over 200 million professional conversations and 170 million interviews, with maximum interviews happening in Metro cities and tier-2 cities like Jaipur, Lucknow, Ahmedabad, Patna, and Ranchi.
By the way, I have already written in detail about How Apna became Unicorn in less than 2 years.
š¤ Problem Statement
(As given during Product Folks Teardown Event)
Apna is a jobs and professional networking product for the rising working class -
it uniquely has a jobs marketplace combined with the ability for users to discover, connect and network with similar professionals.
One of the skills that are required for most jobs in this space is the ability to speak English (proficiency level varies depending on the job type) and many times this is a skill that users struggle with.
How would you leverage Apnaās professional networking products (Groups, Connect, 1-1) to "help users learn English that is required for their current or aspirational job and deliver both learning and job outcomes for them"?
The approach + solution should ensure that:
1. Users achieve learning and job outcomes
2. Apna is able to drive engagement + user retention on networking products3. any other benefits for stakeholders.
š Research Insights
During the research, we conductedā 30 in-depth user interviews. Out of which 22 were Males and 8 were Females. These users were working full time or looking for a job in one of the following categories: Sales Executive, Telecaller and Delivery boy.
Let me share some important insights we got from the research:
Why users come back to the Apna app:
To apply for their aspirational Job.
To socialise on the "Groups" Tab. (All users joined Learn English group").
Connect with professionals and search for recruiters.
Why did Apna users want to learn English?
To find a better job with a higher salary.
To converse fluently with colleagues, customers and managers at the current company.
Thinks better English = More respect in society.
How Apna users are trying to learn English (Outside the app)?
Read books, watch English movies & youtube videos.
Users were using Halo app and Hello English app to learn English using their in-app chat functionality.
By talking to their friends and colleagues in English.
Challenges faced by Apna users while learning English?
Lack of English-speaking environment. (Most users come from Tier-2,3 cities where a regional language is spoken).
Not able to find a serious English-speaking partner (No accountability).
Lack of time after the day job.
Other important Insights
Almost all the users failed to clear an interview they got from the Apna app. On average, users got 1-2 interviews scheduled using Apna until now.
Users with <Rs 20,000b monthly salary were more open and serious to learn English but were not willing to spend too much time & money to learn English.
š„ User Personas (Post Research)
š©š» Persona Name: Minakshi
Minakshi is a 24 years old Operation executive from Nagpur, Earning less than 30K who is looking to learn English to get Telecaller or any back office job with a higher salary. She's been referred by her friends to use Apna app for job search.
She's been an active user for the past 20 days, spending maximum time on the "Jobs" & "Groups" tabs.
šÆMinakshi Goals
Looking for better pay job.
Trying to improve my spoken English to crack job interviews & talk to people fluently.
ā¤ļø Minakshi Motivation
Need a higher salary since it would help her fulfil his parentās dreams + would help her in getting married to a good family.
Social Status and Respect from being able to speak English.
Minakshi knows that both the above things will not come until and unless she upskill herself & become competitive
šØ Minakshi Challenges
Canāt confidently communicate in English (colloquial and fluency).
English words/ sentences she learned via movies & Newspapers havenāt been used in real conversations yet.
Lack of a structured online course, time & work commitments to learn English through courses.
No proper feedback loop was built for the hiring process from application to interview to the job.
š”UX Enhancements & Feature Suggestions
We decided to map Minakshiās user journey, pain points and relevant solutions on 3 screens within the app āJobs Tab, Group Tab and Connect Tab.
Jobs Tab- Problem & Solutions
Pain Point 1āThe user hesitates to apply for a Job because of 2 main reasons:
Lack of TRUST because the company may be a scam/fraud.
Unsure whether sheāll get a call back from HR or not
SolutionāRich Job Cards
Recruiters on App with Identifiers Tags to solve for Recruiter Identification (Trust). Job Seekers can now ask questions on Open Job Positions Best questions get upvoted basis of ranking logic (#of claps).
The Recruiter gets to answer queries pre-job application questions such as:
Clarity of role and skill set required
Clarity on career progression
Current employee comments to solve for the authenticity of the job posting.
Metrics
Users/Job Application.
Clicks/Job Application
Questions/Job Application
Recruiter Rating (basis quality of job of posted + Questions Answered
Pain Point 2āThe user feels demotivated after failed first attempt of assessment and eligibility test.
Solutionā Tweaking the "Job Application" Flow
Instead of asking the user to give an assessment test first and then apply, Let users first apply & then ask them to give an assessment test for "Increasing their selection chances by 70%".
This wouldn't hurt the candidate's motivation (as their job application has been sent) to apply for multiples job, even if he/she fails to score well on an assessment test.
Once the user clears the assessment test & successfully applies to the job, Nudge users with the "Similar job that matches your skill" feed. This ensures that users stay on a Job application streak and apply to at least 5 jobs to land an interview.
Metrics
Success Metric: New Job Applicate rate vs Old job application rate
Assessments Taken/User (Post Application to the job)
Funnel drop off at Assessment vs application before after
Group Tab- Problem & Solutions
Pain Point 1ā The user is not clear about the value proposition of the āGroupsā as a product and how it can help them in their journey to get a job.
One of the reason for this is Lack of major incentives for users to stay active and return back on this āGroupā tab.
SolutionāHigh Visibility Filters
Adding high visibility filters in the āGroupā tab to help users discover & structure relevant posts based on their intent.
For example: English Learning group will have filters like āEnglish Seekhoā , āWord of the dayā , āQuizā , ājobsā etc.
Posts within groups will have feed ranking logic based on usersā engagement metrics (Posts interacted, No. of connections etc.) + Personalisation logic.
Metrics
Average Time spent on App
Groups Open Rate
Filter utilisation rate
The incremental increase in Group Engagement (Claps, replies, Posts, shares)
Pain Point 2ā Majority of the group posts is not relevant to the user. Content discovery is a bit challenging for the user.
The user needs to scroll through to find the relevant learning posts (Videos, quizzes, polls etc).
Solutionā Daily Quiz/ Contest/ Skill Up Games
Users can enter the daily quiz challenge on the "Learn English group" where they can get a chance to earn points & English skill medals.
Once a user achieves a certain milestone of points & badges, they'll get better job opportunities from recruiters on the app. The daily challenge includes a 10-questions English assessment quiz that helps candidates to improve their grammar and vocabulary.
After the contest, show the leaderboard of participants based on their points. Community Champions and Influencers to broadcast winners in Groups to ensure high social status.
Metrics
Average Time spent on the app
Badges earned/user
Feature Usage = # of users opening feature/total app openers
Card Completion Rate = No. of users successfully completed or reached the end state of card / no of the unique card opens
Connect Tab & 1:1 Message- Problem & Solutions
Pain point 1ā The user is looking for a English speaking partner.
Solutionā Find your language buddy/teacher
The idea is to connect users who are in need of learning English to experts or teachers who wouldnāt mind spending a little time guiding users.
The Apna community managers and other experienced members can act like mentors & would be able to help out users in need.
The user needs to answer just two questions, What topic do they need help with and their location? And the app will suggest experts/Mentors to connect with.
Metrics
No. of mentors connect with and reached out daily/weekly
Time spent with each mentor
No. and Kind of messages exchanged with each mentor
Pain Point 2āThe user feels lost while taking the conversation forward with an unknown person.
The user thinks that other person might judge you based on conversations and the way of approaching on DM's.
Solutionā English first smart replies to assist users
Using machine learning, The Message screen will show users suggestions that are more contextual and relevant to the conversation theyāre having.
The smart replies suggest at most three responses based on the message user received. This will help users easily respond to DM messages in English without much thinking.
š¢Feature Prioritization (using ICE method)
Note: In the prioritization table, youād find a few solutions that I havenāt discussed within this blog but Iāve included them within my original deck!
Thatās all from this Teardown folks!
If you like what you read, please share this post within your company Slack or Whatsapp groups. It means a lot to me :)