TurboHire completes 3 years, a year in pure R&D work, and 2 years as a company, with COVID eating away 12 months of the economy but not our enthusiasm or growth. This article has been penned down by Deepak Agrawal, CEO TurboHire, talking about the journey of building a product that is holistically impacting Recruitment globally, at the same time bringing together a team embarking on a path that is culturally progressive.
As I start writing…
I cannot help but mention two guard-rails, I have resonated with. One is the movie “The Englishman Who Went Up A Hill But Came Down A Mountain”, the story of an English cartographer who must tell a Welsh village that their mountain is only a hill, the offended community sets out to change that. The second is the book “Let’s Build A Company: A start-up story minus the bullshit” by Harpreet Grover. I hope the learning from both keeps me grounded.
The Early Days Of TurboHire
I had worked in ISB for about 8 years with a deep understanding of the hiring process conducted by 200+ variety of companies on campus. They were there predominantly to hire MBA talent with lateral hiring flavor. I was more intrigued by why some candidates were successful with certain companies for certain roles, much less by “process”. I also knew the majority of candidates and understood their past work experience.
8 years were an iterative learning process, which built a conviction on the need for “Augmented Intelligence” and “User Defined Search” eschewing traditional “machine learning” and “match finding” approaches. I had a belief that humans when writing a CV/ resume, they describe themselves in complex idioms and linguistics forms, which, unlike traffic or credit card transactions, simply cannot be/must not be subjected to machine learning. The new approach was essentially:
- Take any resume prepared manually and extract only the essential information for the hiring process.
2. Build datasets encompassing educational institutions, academic degree/ background, companies, and a deep linguistic model for capabilities a person describes through sentences. We call it the “Real World” data scaling, going beyond simply roles and designations.
3. The “Real World” data is then laid in a semantic form over the relevant information extracted from the resumes.
4. In this human-machine collaboration, we give the real power at the hands of the user as they can search for the “most relevant” (not necessarily the best) candidates for their organization and the role.
Conceiving all this required knowledge, I ended up learning Chris Manning and Dan Jurafsky’s Stanford NLP course, which in turn made me re-learn matrix algebra from scratch.
There were some core decisions that I had to take to build the venture.
1. Building the team – I decided to reach out to co-founders with complementary skills. An avant-garde approach, but 4 amazing gentlemen Abhishek Kabra, Gaurav Kumar, Aman Gour, and Rakesh Nayak came together to build a Tesla when the world was still busy with Diesel Engines. At this moment, sometime in mid of 2018, “I” became a “We”. (There will be another story on – How To Build A Startup Core Team, soon.)
2. We decided to bootstrap the venture.
We got together to start our R&D work on extracting data, building “real-world data”, semanticizing entity data, building algorithms, checking the accuracy, reworking but every fortnight reaching a place better than before. The R&D work quickened with immense contributions from our advisors Edward Rogers, Suman Gopalan, Amit Nankeolyar, and Ayaskant Sarangi.
After 12+ months of labor, we felt we had an MVP and foundation to register a company, which we did on 12th March 2019, and from the next day, we started our sales outreach.
Untested Hypothesis In Hiring
The hypothesis was that large enterprises receive a large number of CVs and they get piled up as unstructured data inside the organization and TurboHire could be the best tool to make them structured, augmented, and searchable through human-machine collaboration.
We believed then that TurboHire will get plugged onto the enterprise’s current ATS or HRIS to provide “Reusability of CV Database” and “Relevant Screening” value. However, selling a transformative solution to large enterprises proved to be a Sisyphean task (Larger enterprises and their Big4 consultants are so proudly busy in tinkering optimization of existing process and technology and innovator’s inability to deal with their multi-headed hydra-like decision-making that we still shy away from very large enterprises). Within 3 months, we decided to build an entire Talent Acquisition platform now catalyzed with “Augmented Intelligence” in every step of the hiring process cutting hiring time drastically and saving millions in hiring cost while also improving Hiring Quality.
By the end of the sixth month from incorporation, we had our first paying customer. We made our target segment as medium enterprises and went into an iterative loop of improving the product and getting more customers on board.
The customer acquisition momentum was building up as well as bootstrapped finance was running out. The lesson learned is to never underestimate time and expenses on an R&D-based product.
The Blackswan Crisis
Buoyed by our momentum, we had Sunil Roger join us as Director of Sales. His first day was also the first day of lockdown in India. The tail-spin had started.
Our customer acquisition slowed down. Customer payments started to get delayed and by the second month of lockdown, customers started dropping off altogether.
During this time, we decided on a few good things. We decided neither to reduce the team nor to do salary cuts.
The customer leads who were coming were basically curious lot, who were trying to feel newer players and spending available time while their business activity had come to a grinding halt. The culture we set was to perform demos and sell enthusiastically.
This interaction did 3 things. One, we got valuable product inputs; two we improved sales processes, and; three, traction kept the overall morale high.
Early 2020, we had decided to raise money and made a reach out to possible angel investors and VCs. However, before the VCs could even warm up, the COVID crisis was set-in. Rain or sunshine, Aman Gour, leading our Product and Strategy, vigorously articulated our value-proposition and future potential. 2 early angel investors converted to give us a breather for the COVID induced a lean period of limited revenue and static expense.
Most of our earlier VC connections floundered and we again started from scratch at the beginning of the lockdown period. We worked relentlessly and a quarter later we had primed 2VCs and the second round of Angels to make up 1 million USD Seed Round. We were also fortunate that former entrepreneurs in the B2B space came in as our investors.
The due-diligence started in the second quarter of lockdown but by the end of the quarter, we had the majority of the money in the bank.
3 things happened in the later 6 months of 2020. One, we kept a steady pace on product development. Two, we continued to sell enthusiastically and actually brought additional sales teams and cool tech folks onboard. Three, we continued to improve our pricing and sales tools.
Glimmer Of Hope And Path To Future
The start of 2021 was soothing sunshine. The CFOs in mid-enterprises started to allow new purchases. We had a truly mid-enterprise-worthy product. We had digital marketing optimized and giving a steady flow of MQLs and SQLs. We had learned to convert customers and we had a great Customer success vertical under Gaurav Kumar. We had existing customers giving us excellent customer references to help acquire new ones. Our Pricing has remained premium but we could now demonstrate value.
We had been enrolled in the Microsoft for Startup program, so we had better Azure infrastructure, with better stability and performance for our SaaS offering. We had a bigger and deeper handle on our engineering under Abhishek Kabra and Rakesh Nayak.
Every coin has two faces and I would want to recount blessings we received:
1. COVID lean period made us better on the product, platform, company processes and made us go deeper into data.
2. We could start selling remotely saving on travel and time-wasting meetings.
3. It made us setup automated invoicing and payment infrastructure.
4. We crafted a complementary strategy of integration with leading HRMS platforms.
5. It brought us together in a tighter embrace.
What we achieved in the leaner customer acquisition period, is now making us succeed faster and more confidently.
Today, TurboHire has been launched in the US and is winning globally recognized enterprise clients, allowing us to aspire for a sizable Series A in the near future, to prove the impact of “Augmented Intelligence” on the global Talent Industry.