Thought Leaders in Artificial Intelligence: Pavan Sondur, CEO of Unbxd (Part 2)

Sramana Mitra: Walk us through some use cases of what you are able to do. What kind of personalization are you able to achieve?

Pavan Sondur: One is pure relevancy. When somebody is searching for a black leather jacket, we want an engine that can understand and tries to emulate an offline sales person. It needs to understand, at scale, what it is that I’m looking for and dig for it in the catalog even if the catalog has bad data in it. It will give me back leather shoes even if two out of the three words match. That’s one use case. It needs to work for every single query. It needs to work all the time.

The second key use case we’ve been able to solve once we were able to show the right sets of product is what amongst these products makes sense for you, as a visitor, at this point in time depending on what you’ve done before. If you’ve given a signal to say you are a size six, then we look for an availability in size six. We see trends across. The East Coast and West Coast do not buy in a similar way. A whole bunch goes in to determine what would be a good assortment of products that’s tailor-made. That’s what the engine computes at that point in time.

Sramana Mitra: What kind of lift are you able to get? What is the result of plugging your system in?

Pavan Sondur: There are different kinds of measurements of success. I would also agree that increase in revenue is the most important metric. When we plug in, we have demonstrated a lift of up to 70% to 80%. On average, we consistently do 30% to 40% increase in dollar revenue. Somebody doing $10 million a month is now able to do $13 million or $14 million in actual revenue increase.

Other things that contribute to all of this is an increase in engagement. People see more products. The click-through rates of those products go up. Exit rates go down significantly. You have more and more repeat users. Once we plug in, we’ve seen increase in search traffic as well as repeat users who are more likely to use search. We’ve seen three times increase in repeat users.

Not only is there an increase in loyalty but the customer acquisition cost for such visitors is essentially zero because they keep coming back. The last thing is the increase in the average order value. All of these parameters contribute to the ultimate goal which is increase in revenue.

This segment is part 2 in the series : Thought Leaders in Artificial Intelligence: Pavan Sondur, CEO of Unbxd
1 2 3