That said, some basics apply. You need to do some homework so you can project some quality numbers. To start with, visit some of your competitors and get whatever info you can about their company history, visitor stats and ad pricing. That'll give you some insight on actions they pursued and a timeline to achieve realistic numbers of users.
If you don't have any major brands committed to an ad buy (which of course would help sell smaller entities), you have several options. One way is to launch with minimal ads that are mostly self-promotional, along with three-month basic hyperlinked tile ads at a ridiculously low cost to a select group of business peers/contacts that your site (and your projected site users) would be interested in.
The size of your site and each page would dictate whether you have three ads per page or a total of 50 spots, etc. Your research should help set up an intriguing intro rate.
A second option is to offer ground-floor opportunities in tile, listings or leaderboard categories to a larger group of advertisers, good for at least six months. It'll help you get a sale if you can outline what you'll be doing to generate users, along with listing accredited stats from similar sites and your best user projections given your planned marketing initiatives. You can sweeten the deal by offering to send out weekly or biweekly e-mail blasts about your advertisers, too, as an additional "give" to launching ad buyers.
Know that whatever you choose to do, you'll need to invest in a bunch of diverse, strategic marketing initiatives (online ads, interaction and ads in social communities and blogs, and direct mail and PR) to draw users and advertisers. The larger your user base becomes, the more you can charge for ad opps after the Intro phase is over.
Best of luck!