India has a hiring problem, which is ironic for a job-starved economy. Over 60% of the 1.3 billion population of the country falls in the working age group (15-59), and more than 300 million of this group are blue-grey collar workers. Yet, businesses find it difficult to tap into this talent pool and hire staff for their needs.
What causes this conundrum?
Most popular online job portals are not designed to solve problems for blue-grey collar workers at the bottom of the job market value chain. These platforms simply do not address the socio-economic realities of India’s low-income group with less-than-average literacy levels.
So, job openings for the blue-grey collar segment remain invisible and are relegated to social media posts, word-of-mouth referrals, or flyer-based advertisements. This creates a discovery problem for both job seekers and job providers when it comes to finding their match.
"There is an artificial dearth created by the information asymmetry in the job-market supply chain," explains Moiz Arsiwala, Co-founder and CTO of WorkIndia, one of India's largest online job portals for hiring blue-grey collar staff.
Moiz Arsiwala co-founded Bangalore-headquartered WorkIndia with Kunal Patil (Chairman/Advisor) and Nilesh Dungarwal (CEO). Their platform connects 30 million job seekers with 100K businesses monthly, and this figure is steadily increasing.
In a candid chat with Team ProdWrks, Moiz gives insights into the harsh socio-economic realities of the blue-collar segment. He explains how these insights helped him build a platform to enhance job relevancy, expedite recruitment processes, and bolster fraud detection, ultimately facilitating better job discoverability by leverage technology.
White-collar workers seek a career, blue-collar workers look for a job
Moiz says, "White-collar workers view jobs as a career to pursue and grow. But blue-collar workers are not looking for career progression. A delivery agent would not want to become a senior delivery agent. They would be willing to shift to a tele-calling job if it offered a better salary."
"In a single day, blue-collar job seekers would attend multiple interviews in different companies in the same locality to increase their chances of getting a job. They want instant gratification from their job search. Their economic situation does not allow them the luxury of waiting for ten days to get a call from HR."
MSMEs want to hire quicker without involving middlemen
Around 95% of WorkIndia’s revenue comes from micro, small, and medium-sized businesses (MSMEs). Unlike large businesses with a team of HR recruiters, MSMEs usually have a one-man army—a senior manager or the proprietor who juggles multiple duties like hiring, business development, and accounts.
WorkIndia’s research showed that these recruiters do not want to waste time looking at a list of candidates sourced from job placement agencies and figuring out who is interested. They preferred getting inbound calls directly from candidates showing a high intent to join the job.
"Based on these hypotheses, WorkIndia pioneered a completely automated candidate discovery feature. It removes middlemen or any human involvement in the interview selection process. Candidates seeking work can discover jobs on the platform and connect with employers directly over the phone after passing an eligibility test for the job. We were the first platform to enable direct calling,” says Moiz.
When WorkIndia started operations eight years ago, the simple candidate discovery process wasn’t the industry norm. Back then, middlemen charged a fee and dictated who could be hired, exploiting both employers and employees.
"Our candidate discovery process has become an industry standard today for all small and medium-sized businesses who now believe that hiring happens when you get more calls. We made this mind shift happen in eight years," Moiz proclaims proudly.
WorkIndia claims that its platform today speeds up the entire hiring process by 85.7% and reduces hiring costs by 53.8%. Its success is reflected in numbers – WorkIndia caters to the critical hiring needs of over 15 lakh SMBs in India.
Recruitment platforms that tackle job frauds can build trust, scale quick
During the early years, when Moiz and his team conducted user interviews with security guards and delivery agents, they encountered a common question – sir, fraud tho nahi hai na… paisa kitna loge? (I hope this isn’t a fraud… How much will you charge to give me a job?)
Moiz discovered that fraud and exploitation were rampant in the recruitment industry, and workers did not trust job postings easily. Platforms that build trust with users by eliminating fraud posts can scale quick in India.
Moiz lists the three major job scams that are digitally invading most recruitment platforms:
- Job posts from manpower consultancies and middlemen seeking money for jobs.
- Fake job posts that are disguised to sell their own products, services, or agendas.
- Vulgar jobs posts hiring for personal favors, massage parlors, and prostitution.
Being one of India’s largest recruiting platforms, WorkIndia gets 1.87 lakh job listings on the platform and it isn’t easy isolating fraudulent job posts. Raw job datasets can be ‘layered’, so fraudulent jobs and profiles can be hard to track in large recruitment platforms, such as WorkIndia.
You may find individual fake jobs that appear innocuous but share the same address as a dozen others, which in turn are intermediaries for other entities of interest. The ultimate profile owners or beneficiaries of the fraud posts are unclear unless you resolve all the links in this chain.
To best decipher this maze of relationships and curb fraudulent job posts, Moiz has built a three-pronged fraud detection engine that operates with 9.8 million data points collected over a period of time.
Here are the three major elements in WorkIndia’s fraud detection engine:
- A business rule management system (BRMS) identifies fake jobs by categorizing data and recognizing patterns fed into the system with preset rules.
- An NLP-based text analysis uses natural language processing and profanity filters to identify and eliminate vulgar job posts on the platform.
- A proprietary graph analytics system finds interlinked fake job posts. WorkIndia developed this system by drawing inspiration from researchers and journalists who used graph analysis to uncover the Panama Papers scam!
Explaining the inner workings of their proprietary graph analytics engine, Moiz said, "We build metadata profiles of employers using their IP addresses, devices, fingerprints, and various other attributes, and perform graph analytics to try and find deep connectivity between two employers who look very similar to each other. If we find a relationship, then that employer will be blocked on the platform."
Better job relevancy enhances discoverability for job seekers
WorkIndia’s commitment goes beyond curbing fraud; Moiz and his team are continuosly working to perfect the relevancy of the jobs shown to the users and ensuring that job seekers and employers find mutually beneficial matches.
Their highly sophisticated self-learning recommendation engine, with over 10.3 billion data attributes, ensures employers and employees get relevant matches. It deserves a seperate case study.
To explain it to you in brief, here are the six pillars that support the accuracy of the recommendation and relevancy of jobs shown to candidates in WorkIndia:
- User preferences: Jobs are shown based on search history and filters applied by the user.
- User eligibility: Matches qualified workers with the right jobs based on eligibility criteria like education level, language proficiency, etc.
- Job proximity: Distance between candidates’ residence to job location.
- Seriousness of the employer: Filters to identify and remove jobs where positions have been filled and filters to remove jobs posted by dormant accounts.
- Popularity: Jobs that are in high demand or from popular employers are pushed to the top of the list for better discoverability.
- Centroid relative distancing algorithm: A logic-based algorithm that helps the system show jobs that are close to railway lines that connect the user’s residence.
The centroid relative distancing mechanism is unique to WorkIndia and is not found in other recruitment platforms.
"It was born out of a hypothesis that job seekers would be ready to travel 15-20 minutes extra if they can find a job on the same train line without switching their mode of transportation that connects their residence," explains Moiz.
Blue-collar workers are a mobile-first user segment
It’s easy to get mesmired by WorkIndia’s technological prowess, but the real magic is how they’ve packaged their deep tech expertise within an easy-to-use interface for low-income segments who may be digitally challenged.
Moiz says the current success of the platform comes from the hard lessons learned from one of the earlier versions of their application, which failed because of its complexity.
"The initial version of our platform had a fancy map that showed balloons that users could click to look for jobs in a particular area. We internally called it an Uber for jobs. Obviously, the blue-collar workers did not understand how to work that application. So we had to return to the drawing board and make it really simple."
"The blue-collar segment did not progress the traditional way from using a desktop to a smartphone. They moved directly from no device to using a smartphone. So, the way they access technology is very different and demands a mobile-first approach. They can do much better work within their smartphones' small six-inch screens."
Blue-collar workers are cost-consious internet users
"We started our journey with an app. But it was a big app, and the first thing we noticed was that blue-collar job seekers did not want to download it as they could not access broadband or high-speed internet. So we made our app available as an APK, and they shared it among themselves with Xender and SHAREit. Over time, we reduced the file size extensively by making it barebone, which, when we look back today, is the right decision."
Internet stinginess was a common theme that Moiz observed in this target group. Since the price of data packs was high, the blue-collar workers accessed the internet only from the free community WiFi networks installed in their residential localities by the government and social welfare groups. They used their 2G packs only for WhatsApp and other essential tools. So, Moiz decided to build an offline app.
"We built an offline app that works without the internet. New jobs will be synced in the background when their phone connects to WiFi. So they will always see jobs to apply for when they open the app, even if they don't have internet."
Moiz underscores that the way the rest of the world accesses and understands technology is fundamentally different from the way the socio-economically marginalized do. Understanding the subtle nuances in their thought process and designing to their needs decides the success or failure of your apps.
Build 'Hinglish' apps for users who speak 'thoda' English
Product leaders often assume product localization is just a translation of their product features. However, when building products for segments with low-literacy levels in a diverse country like India, localization takes on a whole new level of complexity.
Moiz recounts anecdotes of workers who found seemingly simple tasks, like searching the WorkIndia app on the Play Store, challenging due to spelling errors.
"You'd be amazed," Moiz chuckles, "When we asked people to search for WorkIndia, they'd often misspell it as W.A.R.K I.D.Y.A. Basically, what they speak is what they write. It showed us the importance of designing interfaces and features that align with users' English language proficiency and literacy levels."
Recognizing this linguistic gap was pivotal for WorkIndia in delivering intuitive user experiences to the less privileged. To bridge this gap, they incorporated Hinglish – a blend of Hindi and English – into the platform, making job searching more accessible to those with limited English proficiency.
Moiz has also introduced the ‘Thoda English’ option in job posts – a feature to test the English language competency of candidates and let employers know that it lies somewhere between basic (thoda English) and fluent. Employers can choose ‘Thoda English’ as an option in the jobs they post, making them open to applications from candidates who are not proficient in English.
"Thoda English feature was born out of an insight from the ground. Users did not know how to assess their own English language proficiency. They thought they spoke good English even if they could speak only a few words. They said, 'humko thoda English aatha hai'. This phrase resonated with us, and we added it to the application as an option. When candidates select 'Thoda English, ' it lets the recruiter know that they can speak a few words and read a bit."
Apart from English, WorkIndia’s mobile application is available in 10 regional languages, covering over 85% of languages spoken across India. However, Moiz believes converting everything to the local language is not practical.
User research insights showed that they don’t need to be very local to the language; they just need to do a good job of making users understand what’s shown.
"We convey the essence of the job in the headers. We made job listings as simple as possible, with large texts mentioning the job type, salary, and distance from their current location. This was enough for users to understand and click on the job. No one wants to read addresses in Hindi, Telugu, or Kanada. They are fine reading basic descriptions in English. This is our approach to solving the localization problem from a regional language standpoint."
Our conversation with Moiz Arsiwala highlighted the fact that technology is perhaps the best tool there is to solve the unemployment crisis looming large and addressing gaps in the informal sectors.
However, WorkIndia’s success lies not just in its technological innovations but in its unwavering commitment to bridging the gap between technology and the socio-economic realities of blue-collar job seekers.
Product leaders like Moiz Arsiwala are important parts of the puzzle to bridge those gaps.
