We went from an almost bankruptcy to $1M in revenue in 2.5 years

I started Cogno AI right out of college, back in April 2017. At that time, we were an Enterprise Artificial Intelligence startup that developed custom AI/ML-driven solutions for large enterprises.

Services businesses are almost always unscalable. They are great cash-cows for people who are just starting up. However, scaling them is really difficult. We were one such services business — no wonder, my Co-founder left 1.5 years after we started.

This Co-founder's exit happened in August 2018. It was one of the hardest phases of my life — intense pressure, no team, lack of money, lack of support, and lack of business. I had to swallow the reality and find a way to move ahead. When my Co-founder left, I even had to pay some amount to buy out his stake in the company. I was left with Rs. 1.0 Lakhs ($1,400 approximately) in the company’s bank account.

It was a terrible situation to be in. When you are in such a phase of life, everyone walks away. No one wants to help you when you are going through a bad phase of your life.

I invested a small amount of money from my personal savings. Rs. 5.0 Lakhs (about $7,000) that I had saved in the college days by doing freelancing. Coming from a small, middle-class family, it was a difficult choice for me. But I was out of options — take this plunge or shut the shop. What else could I do?

I decided that we already have 4 paying clients, so why not take the plunge. I explained the situation to my parents. They were totally freaked out. I literally had to calm down my parents because they were worried about my career.

Fast forward 2.5 years later in March 2021, we closed Rs. 7.25 crores of revenue (~$1Million), staying completely bootstrapped. Not just that, we built a team of ~80 team members, all of them working remotely in this pandemic.

How did this sudden change happen?

We focused.

It obviously wasn’t that simple. However, if you ask me what took us from $1,400 in the Bank to $1M in revenue in 2.5 years, I’d say that the biggest contributing factor was “focus”.

Before my Co-founder’s exit, we were trying to dirty our hands on so many things:

  1. An AI-driven ChatBot platform.
  2. A chat-driven cognitive data analysis platform.
  3. A document search and data aggregation platform.
  4. An email complaint automation system.
  5. An e-learning platform for competitive exam preparation.
  6. A platform for enterprises to build reports collaboratively.
  7. A face recognition platform to personalize the shopping experience.
  8. A generic AI/ML platform for large enterprises.
  9. A lending platform for farmers based on their produce.
  10. An algorithmic trading platform for crypto-based assets.

And whatnot.

We were trying hands in all 100 different directions, not focusing on anything. No wonder nothing worked out. It was a mistake that me and my previous Co-founder jointly made. And everything went wrong.

When in August 2018 he left, I sat down with a sheet of paper and noted down the mistakes we made — every single mistake. I wanted to reflect upon what went wrong and how can such mistakes be avoided in the future.

I realized that the biggest mistake we were making is that we were not focused at all. We had a limited amount of time and bandwidth and we had to focus on depth rather than breadth. 10 different ideas won’t take us anywhere. It is easier to build 1 product and take it to 100 customers than to build 10 different products and take them to 10 different customers.

When I write this article, I feel stupid that we didn’t realize this horribly silly mistake back then. It indeed was quite stupid of us to focus on so many things.

I decided to cut off all the random ideas and focus on just 1 thing that was working — the ChatBot solution.

At that time, we had an intern and a full-time Software Engineer along with my new Co-founder. So, including me, there were 4 people. I discussed this idea with the 3 of them and we all aligned ourselves in 1 single direction — build a fantastic ChatBot platform for large enterprises.

We spent about 3 months working on this platform and it shaped really well. We knew what was to be built based on the 1.5 years of services business experience. One of our existing clients called us for a meeting with their Recruitment Head, who was exploring collaboration with startups for their problem statements around Recruitment.

I showed him our platform and he really loved it. He was quite impressed and immediately asked his team to onboard us. We spent time building and perfecting the product with his feedback.

(Honoured by one of our Clients)

We showed our platform to a couple of other prospects also, and they liked the platform. Things seem to be going in the right direction now. We realized that focusing on 1 thing was really working well for us and we spent time really building the best ChatBot development platform.

With the reference of our existing clients came more clients. With the help of one of our mentors, we built a subscription-based pricing model which was suited to large enterprise clients. We realized that charging a bomb on day 1 makes no sense and that the land and expand strategy would work the best.

BTW, if you're wondering how we acquired such large Enterprise clients, I have written about it separately.

We priced the product in such a way that a VP-level person can approve it within her budget. Once she gets sold, she can share references for other people within her organization for other use-cases.

(Honoured by the Ministry of Information Technology)

ChatBots have a bunch of use-cases in Financial Services — the sector we operate in.

  1. Customer Support Bot
  2. Employee Support Bot
  3. Recruitment Bot
  4. Field Agent Support Bot
  5. Lead Generation Bot
  6. WhatsApp Bot
  7. Voice Bot on Alexa and Google Assistant
  8. Bot for Privilege Banking Agents
  9. Bot for Relationship Managers
  10. Bot for Call center agents

These are just a few. A typical Financial Institution has at least 50 business teams and you can sell a Bot to each of them. This means that 1 large enterprise client is essentially equivalent to 50 customers.

We further realized that for selling 1 implementation, we need to get buy-in from 6 people — Business, IT, Information Security, Legal, Compliance, and Procurement.

This obviously is a tedious process. However, we figured that if we have the clearance from the last 5, we don’t need it again and again for selling to the other business teams within the organization.

For instance, if you have the SLA signed (Legal), you don’t need a separate SLA for your other implementations. A short 1-page addendum can be added to the existing SLA. Similarly, once the price has been agreed with the procurement team, it can be used for all the implementations.

This works not only across the business teams within a Financial Institution, but also it works across the group companies. For instance, ICICI Securities would easily accept the Service Level Agreement for ICICI Bank. The same holds for an Information Security clearance and procurement clearance. Obviously, some extra formalities would be required. But you can save time from the negotiation of terms and that helps you move deals faster.

We realized this early in the journey and it worked like a charm. Customers came flowing in like crazy. People absolutely loved our product. I spent my full-day meeting customers and understanding their challenges and adding features to our product.

(Select clientele)

We focused heavily on building things that our customers love. References and revenue came as a by-product. Our ChatBot platform has some fantastic features:

  1. Extremely easy to customize and build, even for Business teams.
  2. Can be integrated with almost all types of backend systems.
  3. Built for enterprise scale.
  4. Strong NLP which can interpret and understand a wide variety of customer queries.
  5. A deep analytics layer that shows actionable reports

On top of that, we provided fantastic support and SLA. I personally took customer calls when needed to have a full assurance that we are always by their side, be it a Saturday or a Sunday.

(We help enterprises manage customer experience at scale)

Later when COVID-19 happened, we added 2 new modules to our Product:

  1. Live Agent Chat: so that Bot can transfer the chat to Live Agents.
  2. Cobrowsing: so that Sales agents can help the customers make a purchase easily.

Both these solutions got a wide acceptance among our customers. In fact, these solutions came from our customer’s problem statements. We co-created our products with our customers.

At the time of COVID, our products turned out to be a boon for our customers. The usage shot up significantly. We helped our clients digitize their contact centers in a smooth manner to Bots and Live Chat.

Today, as of May 2021, we are ~85 members in the team and growing fast. While initial customer acquisition happened through the contacts and network of our advisors, we are now onboarding new customers because of the merit of our products.

During these whole 2.5 years, my team members really worked hard and I give full credit to them for where we are today. Personally, for me as a young entrepreneur, bootstrapping to $1M at the age of 25 years is a celebration-worthy achievement. However, we have a long way to go in building a global SaaS company from India.

I’d like to take the opportunity to sincerely thank the Cogno AI Team members, Customers, Mentors and Advisors, and also my parents. They’ve always been by our side and supported us in this journey. Without them, nothing would have been possible.

I believe that while we should celebrate the small wins, we should always keep the big picture in mind — a vision of helping Businesses manage Customer Experience at Scale.

I will conclude with a quote I love.

Hard work beats talent when talent doesn't work hard.