Job Alert from TimesJobs

(Case Study)

Job Alert are daily mails sent to jobseekers comprising matching jobs based on users’ profile and other criteria’s selected by the job seekers. 2 million job alert mails were sent everyday.

Product

TimesJobs.com, the premium recruitment portal that meets all jobs needs whether it is from the perspective of the jobseeker or the recruiter. It is Times of India Group’s (biggest media organization in India) product and has been in existence since 2004 and has established a strong competitive position in the online recruitments space amidst tough competition.

The jobsite TimesJobs caters to the needs of the job seekers in India whereby the job seeker create their profile on the website ( by selecting industry, functional area, years of experience, location they prefer to work, setting the job alerts etc) and apply to the jobs by searching the same on the Timesjobs website.

My Role

As a product manager I was responsible for new product development, problem/issues to be fixed in the current product, competitor analysis, data analysis, web analytics, preparing product requirement document, coordinating with cross functional teams (design, engineering, testing, sales, marketing).

Project Name

Job Alert (daily mail comprising matching jobs based on users’ profile and other criteria’s selected by the job seeker) mail sent to job seekers.

Problem

 Following were the problems in job alert:

·    Only 70,000 (23%) jobs were applied everyday through the job alerts (i.e., the daily mail consisting of the jobs matching basis the user selected criteria in their respective profile) by job seekers although the daily job application across the website was 3,00,000 per day.

Goal/ Target

·       Increase the percentage share of job application through job alert from 23% to 40% approx. of the total applications per day.

Root Cause Analysis

 (1)      Internal factor analysis:

• Worked with the engineering team to understand the issues/bottlenecks in processing 2 million job alerts daily (single user was allowed to make multiple job alerts) and sending the same in mailers to the jobseekers.

• Analysed the job alert data (delivery, open rate, ctr etc) and confirmed if the data was implemented, tracked and stored correctly or not by the analytics team.  Also analysed the data with respect to application percentage across 2 years through job alert.

• Worked with the design team to create multiple personas (industries, sector, age group, experience, location, job level etc) for performing user interviews and field studies on job alert mailer.

• Conducted user interviews & field study with customers/job seekers of different personas (diverse industries, sector, age group, experience, location, job level (executive, manager, senior manager etc).

•  Analysed mails/chats with support team for issues/complaints related to job alerts by the customers/job seekers.

• Analysis of receipt of job alert mailer in different 3rd party service providers inbox (Gmail, Yahoo, Outlook, etc) i.e. whether the mail is going to spam or not, if yes? then why is it going in spam.

(2)   External factor analysis:

o   Competitor analysis with respect to job alert:

•  Profile creation in other job portals along with job alert setting.

•  Alert settings with multiple different personas.

•  Analysis of the jobs sent by the competitor in their job alerts in different mailer  services (Gmail, yahoo, outlook etc.)

• Design of mailer and ease of applying.

• Study of the changes/updates in mailer guidelines by 3rd party applications (Gmail, Yahoo, Outlook etc) 

Findings & Solution

o   Finding:

• The customers/job seekers were not getting the matching jobs based on their profile/persona.

o   Solution:

• The algorithm (for matching jobs against the customers/job seekers job alert settings) was analysed thoroughly from different perspectives.

• The algorithm (for matching jobs against the customers/job seekers job alert settings) was revamped entirely in order to match the correct jobs against the job profile/personas and also to accommodate multiple scenarios and wider number of jobs against each job alert mail so that the jobseekers were able to apply to more jobs.

o   Finding:

• Job alert was sent in one go to all the users and so if a jobseeker in desperate need would not get the job alert earlier and on the same time every day, then he/she would apply maybe the same job on another job portal or may not also visit the site daily.

o   Solution:

• To increase the application rate the priority was given to job applicants with higher application rate (i.e., those users who applied > 5 jobs and daily as well). They were sent the job alert mail earlier than other users/jobseekers. Those who were not opening the mails or were not applying frequently were sent the job alert mailer later in the day.

o   Finding:

•  Design changes & A/B Testing.

• Job alert mailer was redesigned after field studies with our job seekers and

• competitor job alert design analysis.

• A/B testing was done with respect to the Job Alert Title i.e., the job alert title which

• would lead to higher open rates for the mail.

o   Finding:

• Each time the users clicked on the job title in the job alert mail then they were asked to login and then they were taken to the particular job details. Hence, the user didn’t apply to multiple jobs from job alert.

o   Solution:

• Autologin feature was introduced. Upon click of any job in the job alert mail, the user was automatically logged in to his/her account and the job was applied.

o   Finding:

• We were not following the 3rd party email (gmail, yahoo, Hotmail, outlook) service providers rules and hence lots of job alerts were landing in users spam folder.

o   Solution:

·        All rules were studied for rectifications were implemented in the job alert.

 Metrics

• Jobs applied through job alert before the launch: 70,000 per day

• Jobs applied through job alert after the launch: 1,60,000 per day

• Job Alert contributed 53% (instead of 23% earlier) of the 

Total Job Applications on the website.

• Job Alert Delivery before the launch: 78 %

• Job Alert Delivery after the launch: 91 %

• Job Alert Open Rate before the launch: 18 %

• Job Alert Open Rate after the launch: 31%

• Job Alert CTR before the launch: 8%

• Job Alert CTR after the launch: 23%

Achievement

• Appreciation mail from the Vice President (Mr. Gautam Sinha) on the success of the project.

Key Takeaways/Learnings

•  Learned how much as a product manager it is important " to be in regular contact" with all possible user/customer types.

•  Helped me in building up my confidence that I can work on a large volume of data from multiple sources and across multiple cross functional teams.

• Brainstorming sessions with cross functional teams helped me as a product manager in understanding others view points and how to make others understand my view with the help of data and analysis.

• Learned A/B testing for the first time by implementing it in this complex project.

• It helped me to learn and build empathy as a product manager towards other teams like the engineering team, when I understood the complex task of sending 2 million emailers daily with matching jobs.

·