Most salespeople strongly prefer warm leads to cold ones, and they have a point. Warm leads always have some sort of built-in advantage; that's what makes them warm leads. Unfortunately, it's a very rare salesperson who has enough warm leads to entirely drive his prospecting. While cold leads may take a bit more work on your part, it's quite possible to convert a respectable percentage of them to customers.
First, it's important to confirm that the cold leads you're working with are the right leads. Ideally, your leads should contain a high percentage of actual prospects. After all, the lead who isn't a prospect is a waste of everyone's time. A little pre-call research can often help you eliminate the most obvious nonprospects. For example, if your product is intended for parents of young children, you can probably identify leads who aren't parents just by spending a few minutes doing some Googling. This level of research won't eliminate every non-prospect, but it should help you remove the worst ones.
Next, once you've actually connected with a lead but before you meet with him in person, do a bit more qualifying. That will cut out even more nonprospects before you've spent a lot of time with them, and it doesn't need to be a big deal – asking maybe three or four quick questions can eliminate 90% of nonprospects, if they're the right questions.
At this point, having eliminated nearly all the nonprospects from your lead list, your task is essentially to convert those cold leads not to customers but to warm leads.
It's extremely rare for a cold lead to be willing to buy immediately. Most prospects will want to build a certain level of trust and rapport with you before they're willing to buy. The more expensive the product or service you sell, the higher the level of trust you'll need to create before a prospect is willing to hand over that kind of money. So at this point in the sales process, the name of the game is proving yourself to the prospect. One way to think of it is that rather than selling a product at this point you are selling yourself.
One great way to start proving yourself is to give the prospect something of value without asking for anything in return. Value does not necessarily mean cash value – in fact it's much better to give a prospect something that has little or no monetary value but is valuable to the prospect in other ways. For example, if your Google research turned up the fact that a particular prospect is trying to get into compliance with a new government regulation, and you happen to have a customer who has just gone through the same process successfully, you might arrange an introduction so that your customer can advise the prospect. This could be extremely valuable to your prospect, and will definitely make him feel like you are trying to help him rather than trying to help yourself.
At the same time you are starting to establish rapport, you should also be running through the final and most specific level of qualification. A prospect might meet all the obvious qualifications for becoming a customer, yet have no need at the moment for your product. Before you can proceed along the sales process, you need to identify the prospect's need for your product. Note that the prospect himself might not be aware of such a need. After all, he probably knows far less about your product than you do. He may not realize that the product is exactly what he needs to solve a pressing problem. That's where your expertise can really come in handy: once you've uncovered whatever needs are most important to the prospect, you can determine if any of those needs are something that your product can meet.
If you successfully build rapport with a prospect, uncover a need, and show the prospect how your product will meet it, you have an extremely good chance at closing that sale. Building rapport means that your prospect now trusts your expertise; pointing out a need and explaining how the product will fix it shows the prospect that your product is well worth what you're asking for it because of how much it will help them. Why would he say no?