Navigate back to the homepage

Change what you can change

Kamlesh Chandnani
December 29th, 2020 · 7 min read

I usually talk about the interview process in our industry(especially for the frontend) and how it’s broken. I had shared my perspective about the gaps and how we can possibly fix them sometime back.

The reason why I think the interview process is of utmost important because that’s the only source of connect between you and the person applying for the job. It’s not only about the person finding the right team and people for them to work with but also about you finding the right person for your team. This means it’s a two-way process.

There’s this famous quote

Be the change you wish to see in the world.

This surely sounds aspirational but practically it can’t happen overnight and all alone by just a single person.

But that doesn’t mean we can’t change anything at all. At least, we can try to change what we can.

When I had joined Razorpay last year I saw gaps in our interview process since I had been part of the same interview process. At that time we had started growing our frontend team and I realised this is the high time to fix those gaps because we can’t hire and build a good team with those gaps.

I can’t change everyone’s interview process because

  • Literally, I can’t do it alone 😄
  • I don’t know their problems 🧐
  • I don’t know if they realise there exist some problem 🧐

But at least I knew problems around me so I thought I can at least start with changing what I can change.

It’s been a year and I wanted to reflect my thoughts about how I tried to change things and what are the lessons learnt on the way.

Phase 1 - Fixing the Process

It all began with taking opinions from multiple folks inside Razorpay and their view of an ideal interview process. We literally asked people to share the ideas, putting them in the shoes of the interviewee and asking them to imagine if they were interviewing for a job then what would have be an ideal interview process according to them.

The Efforts

Here’s what we came up with

  • Number of rounds and types of rounds for each role/level.
  • The mapping of interviewers to type of round they can conduct for each level.
  • Onboarding new folks to conduct interviews.
  • Asking people what they know and driving interview based on that rather than what as an interviewer we know.
  • Creating decision-making matrix considering both subjectivity and objectivity angle(since everything can’t be subjective practically)

After that we documented the entire new process in detail capturing all the above things.

This is how our new process panned out


At this point, you might be thinking that does documenting a process means that you’re done? Nah. You can’t bring in all the folks on the same page about the new process. In today’s world who prefers to read a big fat document? We all need practical demos.

  • As a next step we planned to train the interview panelists on our new refactored interview process so all of us are on the same page and we can give a consistent and great interview experience to everyone regardless of who takes the interview.
  • The training included how each round needs to be taken, how to write feedbacks, how to make decisions etc.
  • We conducted training for the Talent Acquisition/HR teams as well and trained them on how to screen the profiles - what to look for, how to communicate with the candidates about the expectations for the number of rounds and brief about the expectations from different rounds.
  • The training was not just one-time training. It is a recurring one.
  • The training was conducted in batches of small groups to have focussed discussions rather than a boring classroom session.

Once we were all set, it was time to act and perform things on the ground in reality. We tested the above process in reality on the ground and all of us witnessed significant differences.

The outcomes

  • The panelists were aware in advance about what round they are going to conduct next, which helped them to drive the discussion accordingly.
  • A well-defined round structure helped the candidate as well as interviewers to set the expectations for each round.
  • Clear process for different roles helped to drive the discussions more contextually which was a win-win for both the candidate as well as the interviewers.
  • The Talent Acquisition(a.k.a TA) team became comfortable about the clear communication to the candidates about expectations for each round.
  • And our conversion(number of candidates interviewed : offers rolled out) numbers went up by 5% approx.
  • We used to take candidate feedbacks as well for each round and each interviewer which helped us to fill the gaps.
  • Proper detailed feedback on the recruitment portal for each round helped us to make better decisions

The realisations

Even though things had improved but few things didn’t work out and there was still room for improvement. Here are some of those things

  • Our way of conducting interviews improved drastically but the profiles of candidates that we were getting had an expectation mismatch because of which the conversions were impacted.
  • There’s no one to blame here because the profiles are generally scanned by TA team who usually have very less understanding of the actual technical expectations and they end up looking at just keywords(which is a usual process).
  • Because of the above gap we(interviewers) were conducting more and more interviews but the outcome was not that great.

Training also didn’t help much here because the only way this can be fixed is by training the TA team to code 🙈

Phase 2 - Fixing the Sourcing and Filtration

We looked back and re-iterated to fix the above gaps. At this stage we looked at the gaps from a different perspective all together. We wanted to try something modern than just boring job posts on traditional hiring platforms.

Our base problem here was to fix the filtration process before the interview even starts to save everyone’s(candidates as well interviewers) time upfront.

The efforts

  • Varun came up with a great idea of building a creative jobs page because we were hiring for frontend which means the jobs page can serve the purpose of showcasing the work we do.
  • As a next step we built this amazing jobs page which is less boring and more interesting with all relevant information minus the extra noise.
  • The page just accepts very basic information(No brainer 3-4 fields that you can even submit it from your mobile). Less hassle more submissions.
  • We then signed up with Sid’s tiny react newsletter with a link to our jobs page.

The profiles started flowing in after that. But what’s next?

  • Couple of folks(from frontend team) volunteered to help and we started screening all the profiles one by one. Yes, we literally scan each and every profile that comes through our jobs page manually as of today. I’ll speak about the reason of manual screening in a while.
  • After we saw the amazing response through our jobs page, we started sharing our jobs page on all social media platforms.

The outcomes

  • After the screening process, ~25% profiles were shortlisted and we scheduled interviews for all those shortlisted ones.
  • Surprisingly, all of the 25% shortlisted folks turned up for the interview. 💪🏻
  • Out of all the above who appeared for the interview ~30% of them made it through and we rolled out offers to them. 🔥
  • All of this happened in the span of just 3-4 weeks. 🚀

The realisations

Again at every stage we look back and see what worked out and what didn’t. Few learnings on the way.

  • We had literally tried a lot of tools out there in the market for screening but none worked out since no one can mimic a human while screening. All the tools out there are literally keywords based. 😔
  • We realised we wanted a personal technical connect with the candidates(about their work and skills) and that can only happen with us manually screening the profiles(even though it’s a tedious effort but the outcome is worth the effort). 🤓
  • The traditional way of sourcing and screening works out if you just care about masses and not specifics. For us we were chasing the latter. 😇
  • It’s not an optimal solution but I can consider it as a sub-optimal one because we invested time upfront while screening(though it’s a tedious process but still worth it because you want to work with the people who are suitable for your team) ahead of time and this eventually helped us for better conversions later on without draining ourselves and the candidates. 😄

Phase 3 - Bringing out the creativity

At this point we are trying to be more and more creative on social media. Some glimpse of it

Recap 👀

  • Until this point we have been able to fix our interview process which has helped us to give a better interview experience to the candidates.
  • We have been able to fix the sourcing and filtration criteria which has helped us to have better conversion rates.
  • We are trying to make our job postings more creative and interesting. Enough looking at those thesis sized job descriptions. 😄

It’s not a deal done. We still regularly keep a check on our process and iterate to fill the gaps to make it smoother and better. It’s an evolving process.

Wrap up 📝

  • Driving any change is a long process. It can’t happen overnight, it’ll take time.
  • If you want to change something start small, take baby steps, start with changing things in your surroundings rather than taking a big shot.
  • Patience is the key, since processes involve humans and convincing humans ain’t easy.
  • Getting feedbacks regularly and iterating on them is necessary.
  • Building a great team is hard but identifying and fixing the gaps make it simpler.
  • If nothing works out it’s still okay at least you tried to fix things. And you never know you might have discovered some better way or method to fix things.

If you want change something and need ideas to start with or if you have changed certain things and want to share your stories you can write it to me or you can DM me on Twitter.

If you like this then don’t forget to
🐤 Share
🔔 Subscribe
➡️ Follow me on Twitter

Join the Newsletter

Subscribe to get the latest write-ups about my learnings from JavaScript, React, Design Systems and Life

No spam, pinky promise!

More articles from Kamlesh Chandnani

Service Worker and Server Side Rendering, hmm? 🤔

Create an offline experience for server side rendered applications

October 30th, 2020 · 4 min read

Server Side Rendering, the better way!

Set up Server Side rendering from scratch the easier and the scalable way

September 22nd, 2020 · 8 min read
Link to https://github.com/kamleshchandnaniLink to https://twitter.com/@_kamlesh_Link to https://www.youtube.com/playlist?list=PLpATFO7gaFGgwZRziAoScNoAUyyR_irFMLink to https://linkedin.com/in/kamleshchandnani