Until the technology is developed to a stage that it is likely to draw the attention of an established player in the market, the best strategy is to get the product on the market, and start bringing in revenue.  Competitors will see your product on the market and very possibly express an interest in licensing and/or purchasing the rights.  If it is too expensive to get the product on the market, the next best strategy is to develop a prototype, and take that around to companies that are interested in manufacturing or buying the product.  Either way, the goal is to interest an established player in the market, and then put them in touch with your patent attorney to hammer out a license.

There are some very good books on licensing of patents and technologies, including:

Patent Licensing: Strategy Negotiation Forms(Pli Press's Intellectual Property Law Library) (PLI Press' Intellectual Property Law Library) by Mark S. Holmes

Technology Patent Licensing: An International Reference on 21st Century Patent Licensing, Patent Pools and Patent Platforms by Larry Goldstein, Brian Kearsey.

Although less comprehesive, some of the best guidance is given in Strategic Patenting.

No.  Although there may be exceptions, every patent marketing company that I know of does a horrible job of both securing a patent and marketing the technology.

No, but it may well be a good idea to do so.  Aside from incurring damages and legal fees, it is very frustrating to spend a year or more developing and getting a product or service on the market, and then have to shut down operations in the face of a competitor's patent.  Depending on market practices, it might also be necessary to buy back infringing product from one's customers.  Note that as a general rule there is no compulsory licensing in the United States.  A competitor holding a patent has not obligation whatsoever to license his patent to you.