Can I recruit 10x better?
As I reflect on my time at Swiggy, I remember that building a team was an important aspect of my role. But my memories of the recruiting experience were far from pleasant — countless hours of CV screenings, followed by interviews where I realized early-on that the candidate is a poor fit for the role, the weekly chore of moving-around my calendar to accommodate hiring slots, and talking to HR about “who will interview the candidate next?”. To top it all, there was no guarantee of a light at the end of the tunnel — the candidate could drop-off if they received another offer … or even worse, we end up hiring the wrong candidate only to realize it a few months later!
And, as I support my portfolio companies in building teams, I realize that this is the reality of recruitment across the board. Overall, I feel the “average recruitment experience” today, even with the best applicant-tracking system (ATS), still performs poorly on several dimensions.
Fortunately, there are strong tailwinds to suggest that things are about to change.
Tailwinds for the RecruitingTech segment
- The pandemic has broken several barriers when it comes to recruiting: openness to hiring through video interviews unlocks several possibilities by using data which never existed before; moreover, remote working has suddenly exposed organizations to a much wider talent pool for which current CV screening & interview/evaluation processes are not equipped.
- RecruitmentTech (traditionally considered a cost-centre enabler) usually lags SalesTech (revenue-centre enabler); and sales teams have had access to technology for automated pipeline generation, cold-emailing & client engagement, sales-call observability & feedback through NLP (Gong) since some time now.
- Recruitment Software is already a $2.2B business — which indicates that the market for specialized vertical/functional software is also getting big-enough. This should mean that one-fits-all recruitment systems (e.g. ATS, job marketplaces, interview/assessment tools) should give way to more specialized tools enabling different vertical-focused processes.
- There is stronger awareness & intent among organizations to improve on their Diversity & Inclusion initiatives — and recruitment has widely been considered the primary beachhead to solve this.
A typical recruitment process, and pain-points
Let us use the chart below to discuss the key issues with a typical recruitment process. Note that this is a very linear flow, which is a very simplistic view of the process — real life recruiting projects have various breaks and feedback loops, making the process fairly non-linear. Still, this should help us analyze the layout
But as we discussed earlier, a typical recruitment process has several problem areas, which technology can readily fix! You can also view the chart on this spreadsheet here
Where we are, and aren’t
In the past 5 years, we have moved beyond ATS and job-marketplaces to already see several solutions to tackle some of these issues. The schematic below outlines a few notable players and their current stage of maturity. You can also find a spreadsheet version here.
Brighthire, for instance, leverages NLP on video interviews to provide pointed feedback to interviewers on how they can improve their performance, and the candidate’s experience (kind of like a Gong for recruiting). On the other hand, platforms like HireVue and Sparkhire cuts out the need for screening interviews or CV shortlists altogether through automated assessments of video interviews. TestGorilla is increasingly gaining acceptance as a better addition to the recruitment process for many roles currently relying solely on interviews; and we are even seeing some verticalized-specific players like ByteBoard (for developers) & Enigma (for cybersecurity). Lastly, the likes of Kula take away the chore of cold emailing and follow-ups by using bots instead (sales teams have had this since at least 4–5 years).
However, despite the flurry of players in this space now, I believe there are still some areas which remain largely green, and with tremendous potential for new meaningful solutions.
- A Chief recruiting officer (CRO) dashboard: despite all the modular intelligence systems already available, the CRO does not have a single view that enables her/him to identify & improve existing recruitment processes. E.g. if a candidate dropped off, was it because of a poor interview experience, or delaying in scheduling? Should I increase the interviewer panel, or replace one round with automated tests? And most importantly, no solution accurately goes so far as to analyze recruitment processes with actual on-job performance outcomes — “are we hiring the right candidates”? I see potential for such a solution — which can plug into the ATS, HRMS, and other modular software like automated test-libraries, but act as the key intelligence layer, almost an OS for the CRO.
- More data-driven sourcing/matching tools: savvy recruiters don’t look at just LinkedIn/Resume data anymore. Especially in fields like cybersecurity, or development — there is much richer data available, and being used in unscalable ways by tech hiring managers (e.g. looking at GitHub contributions). I can see opportunities and tremendous value for tools which can build on these vast databases, to provide sharp and vertically-focused candidate sourcing/matching.
- Automated Tests/Assessments: I feel like we have only scratched the surface with automated tests — and here again, I believe the theme will be “verticalized”. They are invaluable in ensuring calibration across all candidates, removing biases, and engaging candidates at-scale and in-time. I’m excited to see how this model evolves further, but I do know that this will unlikely be a winner-takes-all — there should be room for at least 2 large horizontal, and a few vertical players.
- Interviewer Intelligence: If believe that for most management hiring, automated tests could pare-down the shortlist, but actual interviews are not going away anytime soon. In such times, and when there is as much pressure on the interviewer to make an impression, as on the interviewee, solutions like Brighthire will gain further popularity. In fact, they could go a step further and not just enable the interviewer, but also supplement their analysis with candidate data-points of their own. The next gen interview intelligence tool would also challenge the interviewer’s scores — with calibrated data, and playbacks of specific parts of the interview as evidence.
The (ARR 100) million dollar question — Defensiblity
As many solutions relying solely on building data-analytics & NLP layers on existing processes will see — true defensiblity would be hard to find for many business models. I envision three sources of defensiblity in the opportunities I have listed:
- Vast individual-level usage data: especially for interview intelligence & CRO dashboard, the highest form of defensibility is the histories & trajectories that these solutions will build at an individual user-level (interviewer, HR POC, hiring manager, …) and the resulting dependence on them
- Large user-base: making it difficult to replace existing software due to retraining needs. Especially in real-time interviewer intelligence, there would be resistance to change existing solutions.
- Deep-integrations with several systems: I see this as a weak moat at best, but could still provide some defensiblity on edge-cases
- Superior-product and constant innovation: Especially in automated test-libraries, being able to build a solid content & delivery team could be the real differentiator
Overall, I am very excited for what this space has to offer in the coming 2–3 years. I have spoken to a few promising companies that are building on the lines I mentioned above — would love to hear more thoughts and leads on any other startups in this space.