Gaurav Oberoi knows a thing or two about startups, and even more about making them successful, having started and sold BillMonk and Precision Polling in the US in the last four years.

BillMonk, which helped roommates split bills, was sold to Obopay and SurveyMonkey bought Precision Polling, which creates automated telephone polls. It wasn’t all a walk in the park, as Gaurav will explain when he discusses his experiences at Accelerate 2010.

We caught up with him before the event to get his thoughts on entrepreneurship and startup life.

Hear Gaurav Oberoi live at Accelerate 2010, Asia’s largest innovation conference on Sept 22-23. Book by 31 Aug for more than 30% off tickets. Use promo code ‘E27EARLY’.

On taking the risk of leaving employment for a startup:

The decision to leave Amazon [for first startup BillMonk] wasn’t easy, I went back and forth deciding. I was in my early twenties, freshly graduated (Computer Science from Rice University) and working at Amazon but I made the jump after careful thought.

I would say, looking back on my experience, that founding a startup isn’t the risk that many perceive it to be.

I was young and in a period of my life where I was as free from responsibility as I ever could be. If I was ever going to pursue my dream and start my own company, this was it.

Though I would leave my job, my employer was supportive – leaving the door open for a return – while finding another industry job would not be difficult given the experience I had.

Then there was development potential. Fail or succeed, the new experience would develop me beyond anything I could experience at Amazon.

On the preparation that went into founding Precision Polling, his second startup:

Before I left my job I talked to 25 or so contacts having already modelled the price and estimated roughly costs. When you have an idea, you need customer feedback to know your product target better. It is an old school technique that not enough people do these days.

After congealing the information from market research I concluded that I wasn’t crazy and people would actually pay for the service. Once I knew that I took the leap.

On what aspiring startup founders can learn:

There are a number of questions you need to answer.

If you take the leap what is the average revenue per user? How much will it cost to acquire customers?

If you have a product, build a model, then take a top, middle and low guess, you’ll be within 50% which is enough to get an idea of the potential of your product.

I estimated the average revenue of customer very conservatively, it turn out to be a lot higher, but even using my original estimates we had budget to go out and acquire customers through marketing and PR.

On the importance of an exit strategy:

Anytime you start a company you need to clearly identify your exits, even if it isn’t the immediate plan. If you’re going to go and speak to investors, who care about raising money, they will need to see an exit strategy.

If your story doesn’t jive into that, you’re going to have problems raising money, and if you don’t plan on raising money then you’re planning on never building growth.

On the value of PR, marketing and advertising for startups:

When we were ready to go public I spoke to Jeff (CEO of Twilio, the platform on which Precision Polling was built) and we agreed a Twilio post announcement which was put on Hacker News, to gauge interest, before we reached out to GigaOm, TechCrunch and others.

After reading the TechCrunch article (which heralded Precision Polling as “a Survey Monkey for the phone) a Survey Monkey guy biz dev guy signed up. I made sure to get hold of him over the phone and thank him for his interest, the conversation went from there and we were subsequently acquired by them.

TechCrunch gets the credit for broadcasting, we generated some new customers and other interest but advertising in our target verticals – such as political journals – really got us a return on investment. We also won a number of political awards, and other PR plays, which brought in news leads and carried us into our target market.

The TechCrunch article served more of a stake in the ground than PR, we knew this was a good idea and wanted to be the first to claim the concept and create mind share and VC buzz.