Optimizing Websites With Machine Learning: BloomReach CEO Raj De Datta (Part 6)

Sramana: What rates of improvements do your customers see when they deploy your solution?

Raj De Datta: They will see results as soon as one week. At the one-year point they should see an 80% improvement, especially in the areas of natural search.

Sramana: What types of companies are your typical customers?

Raj De Datta: Primarily large brand websites. Williams-Sonoma, Guess Jeans, and Drugstore.com. About two-thirds of our customer base is in e-commerce. We have a lot of travel websites as well. We will continue to move into other verticals.

Sramana: How do you sell? How do your customers find you?

Raj De Datta: If you have a really great value proposition, people will find you. That has happened in this case. There were six people in the company when we started. We did get a call list, and we did start networking. We used our investor networks and we went to conferences. We sell to CMOs. One we got them engaged in a conversation, we were able to get them interested.

Sramana: What are the deal sizes you work with?

Raj De Datta: They are big deals. Philosophically we have a pay for performance model. We get paid based on the incremental benefits we bring. If our solution does not work, then the client is out a couple of days of integration time. If it works and they convert traffic, then we generate a lot of money. Our business model is CPC based for BloomSearch. It measures the incremental traffic that it drives from search engines and charges on a cost-per-click basis for that. We want to be the most profitable ROI channel for our customers.

Sramana: What kind of CPC rates are sustainable? That depends on what products you are selling and what the price points are, correct?

Raj De Datta: It varies with each customer, and we have to negotiate each deal accordingly. There is publicly available data on the cost of AdWords and the return on ad spending and affiliate pricing, which we can use as a benchmark.

Sramana: What you have done requires a sizeable amount of development work. How long did it take you to develop the product before you started generating customers?

Raj De Datta: We started the product in February 2009. We had an alpha version of the product in July 2009. By October, the product was working well and we were testing with a few different customers across a few different sites. The first time we generated revenue was in the beginning of 2010.

Sramana: How many customers did you have involved in your design phase?

Raj De Datta: There were six or seven.

Sramana: What was special about these six or seven customers?

Raj De Datta: They had a burning need. They were focused on how they drove traffic to their website. We were straightforward about the fact that what we were doing was experimental; they just wanted to be on the bleeding edge of technology. Our first customer was Bellacore.com, and they are a reference customer for us to this day.

This segment is part 6 in the series : Optimizing Websites With Machine Learning: BloomReach CEO Raj De Datta
1 2 3 4 5 6 7