ChaCha’s Dan Hodges Expects Guided Answers to Rival Web Searches

When “guided search” platform ChaCha hired Dan Hodges in March to be its president the company also set up a New York office for him to head up its growth.

ChaCha is a mobile Web platform for getting questions answered with the help of human guides. Users can submit questions via text message, the ChaCha mobile app, or toll free calls. The website also features new and trending questions and answers. ChaCha runs on ad revenue (and is backed by venture funding), keeping its services free for users.

The company is headquartered in Carmel, IN, and counts and Quora among its rivals, though they use different approaches for furnishing answers. With updates in the works for ChaCha’s platform and mobile app, Hodges, previously chief revenue officer with Verve Mobile, spoke with me about the direction he sees the company taking since he came onboard.

Xconomy: How does ChaCha differentiate itself in the online question-and-answer sector?

Dan Hodges: There are basically three ways to find answers for people. There’s algorithms, there’s crowdsourcing, and there’s human-assisted guides. When you use algorithms to ask questions and get answers, you’ll only get it right about half the time because algorithms just aren’t at that point yet where they understand human intent. When you crowdsource, you only get an accuracy level of about 55 percent, maybe 60. When you use human-assisted guides, which ChaCha’s done since day one, our general accuracy is around 73 percent. From what we understand, it will take 10 to 15 years before algorithms are able to understand human intent.

The way intent works is if you misspell or mispronounce something along the lines of “How tall is the Umpire State Building?” our technology will ask “Are you saying ‘How tall is the Empire State Building?’” Our database is comprised of fully formed questions and fully formed answers. We have many patents filed on inference [in terms of search technology] to know exactly what you mean.

We’re quickly going from a world of touching and tapping [devices] into the talking world with Google Glasses, interactive cars, refrigerators, and so forth. The ability to have a database of natural language questions and answers is a plus. With ChaCha, the answers are very “snackable.” People tend to ask short questions and want short accurate answers.

X: Do the questions tend to be on pop culture? Are these random facts? Where is this market going?

DH: A lot of the questions are about entertainment and pop culture. Some of them are personal relationship questions. A fair number of them, millions really, are questions that are deeper regarding health, legal matters, and getting jobs. It’s very diverse when you look at the spectrum of questions being asked and answered.

For example, I was preparing for a meeting [with Ford] and I wanted to see how many questions we get every day regarding the carmaker. I was surprised to see it’s around 300. They are very specific questions that deal with buying, people interested in knowing the difference between a Ford and a Chevy, or interested in engine size or fuel economy.

People might ask ChaCha what’s the trigger point for insurance rates to go up? People want to know “What is the grade of beef that Arby’s is using in their roast beef sandwiches?”

X: How do you control the quality of the answers?

DH: Someone asked a question about the Kia Soul but they wanted a full narrative on it. There was a short answer and also link to a longer answer, which was a detailed profile of the car as if from the manufacturer.

We have 129 million question-and-answer pairs in our database. When a question comes in, our dashboard has all the answers that relate to that question. A great many of the questions have been answered before. We sort it through the database and try to answer them in the most expeditious way.

X: How do you make sure answers are even-handed, especially with questions that can elicit strong opinions?

DH: We track everything that is fact-based but we do get questions such as “Do you think Justin Bieber would like me?” We try to say something polite with a neutral response because there is no way to answer that question.

X: What makes ChaCha different from using a general search engine to find answers?

DH: If you ask a question regarding a car on Google, maybe you’ll get 150 links. You have to sort through all of that. When you ask specific questions [on ChaCha] we’ll get the answer more quickly with less work. We also have the ability to ask the question in voice. We’re working toward the capability to ask questions in voice and get the answer via voice.

X: What are some of your plans as president?

DH: We’re going through a new design of the ChaCha website, which you’ll see in the next couple of months. We want to increase our social presence in terms of being that uniform go-to-market, work with advertisers, and listen to consumers.

X: What can we expect to see from the New York office?

DH: Our marketing function will be based here; our business development may be based here but also San Francisco. We also have an L.A. office. We’re opening a Seattle office, a Chicago office.

X: Where do you see growth coming from for ChaCha?

DH: We have a platform called Social Reactor where we reach more than 7,000 influencers. It’s a closed network; everyone has to go through us. We work with advertisers, whether a media company or a brand. If advertisers have a message they want to share, such as watch videos or download content, the influencers can reach about 100 million followers. They’ll tweet something to alert people to the brands’ ads. That spreads virally and helps drive viewership and engagement. It helps the brand connect on a mass level to groups that identify with entertainment, sports, fashion, or any of those types of verticals.

Trending on Xconomy