Guesswork - Product Recommendation Engine for New Users in eCommerce Sites

An entrepreneur and a developer, whose sole motivation is to build – that is Mani Doraisamy for you. His endearing humility and risk-taking spirit, combined with the zest for creating a level playing field for companies makes for a riveting read. Co-founders, Mani Doraisamy and Boobesh Ramalingam, have known each other since college days and have worked together for five years. One thing we can be sure of is that there is much more to come these stables of Guesswork. Excerpts from my interview with Mani.

1. What gave you the idea for Guesswork?
In our early days, Boobesh and I had built a tool to analyse social profiles of users. No one wanted to buy it. But, a marketing head of a large e-commerce company said they could use it to recommend products for their new users, whose preferences were unknown.

We weren’t sure if it was a big enough problem. So, we asked e-commerce companies about the percentage of new users they had, but no one wanted to reveal that number. To demystify the secrecy, we prepared a presentation that said 30% of registered users in e-commerce sites were new users. We showed it to an e-commerce company and asked if it matched their number. They said, “Yes.” Then, we increased it to 40% and showed it to the next company. They, too, said, “Yes.” This went on till 80% and every company said yes.

When we showed 90% as new users, they stopped us and said that it too might be true. But, if they agreed to it, it meant that only 10% of their users were active; while their valuation was based on total registered users. At that moment, it dawned to us how big this problem was and how hard it was for e-commerce companies to even talk about it openly. Since then, we have been super focussed on solving this new user problem for e-commerce companies.

2. I love the name ' Guesswork '. How did you come up with the name?
The purpose of AI is to take guesswork out of the decision making process. We thought, “Why not name the product after the problem it solves?” Hence the name.

3. Tell me about the TechCrunch Hackathon in the Bay Area that you won.
I like Twitter because it lets me follow someone I met in a tech conference without requesting for her/his permission. But, I really don’t want to know if his dog pooped today morning. Irrelevant tweets like these fill our timeline. I thought it would be cool if I could follow someone’s persona, someone I am interested in, rather than following everything they do. That's what I built at TechCrunch Hackathon.

4. Can you elaborate on the algorithm that you use to understand your consumer preferences?
The base AI algorithm comes from the hackathon I built at TechCrunch. Since then, our algorithm has become very sophisticated by crunching huge amount of data. So far, we have analysed more than half a billion users across the world. Based on that, we understand what kind of product an urban millennial browsing on a MacBook will buy versus a rural woman browsing on an entry level smartphone.

5. What sets Guesswork apart from other players in the big-data/recommendation engine space, who are working on understanding consumer preferences?
It’s about focus and data.

Existing recommendation engines work well for users with purchase history. We focus on new users, who have no history in e-commerce sites.

Over the last 3 years, we have built one of the largest consumer persona databases. This data gives us substantial advantage over competition and our accuracy is very high.

6. What are the major challenges that you faced or are facing?
Before we found the e-commerce market, we were really struggling. We were trying to sell this AI tool to CRM companies and it just wasn’t working. We were running out of money because living in the Bay Area was super expensive.

At that time, a product head at Google asked me if I wanted to join Google. I felt insane rejecting it, especially when there was no paying customer in sight. We came back to India to save cost, rebuilt the product for e-commerce companies, got customers and became profitable.

That evolution from a generic tool to a compelling value proposition was very challenging.

7. If not an entrepreneur, what wold you be?
Developer. I still am, in fact. I don’t see myself as an entrepreneur. I love creating products and entrepreneurship is a means to that end. It is the shortest path to go from idea to reality.

8. What can we expect to see in the future from Guesswork (as a company) and from you (as an entrepreneur)?
We will be launching our mobile version soon. We have already rolled out our early version at Babyoye and the results have been incredible. I am super excited about that.

9. How has the consumer feedback been?
We are excited about the results. We are on the same team as our customer and we get paid only when we make a sale through our recommendation. On multiple occasions, customers come back and increase our pay-out. Action speaks louder than words, they say.

10. Which has been the most rewarding moment in your journey of building Guesswork? In 5 words, define the culture of Guesswork?
The day we went live with Zalora - Asia’s largest fashion e-commerce company and increased their conversions by 5 times. I was in tears that day!

We have a never give up attitude. On a fun note, we are like a star trek enterprise - to boldly go where no man has gone before. In the last 3 years, we have lived in 3 different continents (Bay Area, Bangalore and now Paris) to seek new opportunities and to grow Guesswork.

11. Which is the one value in your firm that you are proud of and which value you still wish you could instil in your employees?
We will stand for fairness in the e-commerce marketplace.

We think monopolies are unfair, but have been made heroic by Silicon Valley folklore. Monopoly is the single biggest threat that can hold both consumers and sellers hostage to one company’s profit-making objective. We want to enable every e-commerce company with same data insight and technology as monopolies. We want to be known for creating a level playing marketplace.

12. What is the craziest thing you have done?
Before Guesswork, I co-founded another start-up and put my heart and soul into it for 10 years. A year before I left, we raised money. But as most of my shares were unvested, I lost most of it when I came out. It was all paper money. But still, it was my life’s hard work and it was worth millions of dollars.

At that time, I had a new-born baby and I had just moved to the US with my wife and baby. To lose everything and to start again from scratch, knowing it might take another 10 years to stabilize my life was crazy.

13. If you had a week with unlimited money what would you do?
In my personal life, I don’t make decisions with money as a constraint. So, unlimited money wouldn’t change a thing.

My co-founder (Boobesh) and I have eerily similar views about money. Although, we don’t make personal choices based on money, we embrace it as a constraint for business. We love it because it makes us more resourceful. That’s why, I think we have been profitable without raising any money.

14. What are you afraid of?
I became a founder because promotions and salary hikes were a rat race. Now, I worry if start-ups have also become a rat race like ‘raise money, exit and become an investor.’ I am afraid of getting sucked into it.

15. Tell me something you wish you hadn't done.
Being a developer myself, I get impatient when a product gets delayed or is subpar. Often, I end up seizing the work and doing it myself, while hurting the people who look up for my guidance. I wish I could be more patient.