Which business model is better for the new web startup?

I would like to ask you for your opinion on pros and cons of 2 different business models for a social networking+user advertising start up.

For a very vague description and comparison, imagine this project is somewhat similar to linkedin. Project target market is business sector.

Which of the two do you think its better business model:

a) Have the website completely free for everybody and base the revenue on advertising?

or

b) Offer some of the features free, however limited and base the revenue on paid subscriptions which user needs to pay in order to get access to the premium content and "full usability".

Which of the two models do you think offers bigger opportunity for growth? Or do you perhaps have any other suggestion that might work?

Thank you very much for you opinion on the subject. It is much appreciated.

I think it depends on the size of your potential market.

If its 1000s then a freemium model is better, with a lot of functionliaty being free ... Basically if your adding a high amount of value like a tool to run someones specific business (we do a range of these), then your likely to get a lot less overall customers but you can charge an ongoing fee for it ... These sites "feel cheaper" if they have advertising all over them. Listen to this week in startups interview with Phil Libin, Evernote CEO for the best description of running a freemium model I have heard.

If its millions then an advertising model is probably more suitable because you can estimate a rougly 1.5 to 3% click through rate so millions of visitors means more clicks and no barrier to entry means more people.

As a "rule of thumb" low volume sites will struggle to make revenue from advertising alone because of the 1% to a maximum of 5% rule (unless you do a stackexchange for adverting and people points are based on clicking on the ad) ... not a bad idea that :)

Of course every business has to make money but what I'd say is try to focus on the product first and get it off the ground in terms of its users. Without having the critical mass of users a site like this isn't going to command any advertising revenue potential. Think about what your users need and love - make them rave about your site and then monetization ideas will come naturally.

With that in mind I'd go for your first model because it is the simplest model that allows you to focus on building a user base without putting any barriers in place. If you cannot get a user base built up with an essentially free service, then you'll have trouble with a subscription-based model. Demonstrate you can get the users first. If you need a model for your business plan - go with the simplest one that gives you the most flexibility to adapt as you grow the business. The chances that a startup company will still be growing on the same business model it first came up with are low. You'll have lots of crazy new ideas as you start the business and you'll be happy that you started off simple to give you the freedom to try out different things along the way.

Clearly, if you want to get significant following to your website, you need to start with an everything-free model, and generate income from advertising. You should also think if you can become an affiliate of other sites selling goods or services related to your website's content. This could generate additional income.

However, you should plan ahead and build your website in such a way, that later on you'll be able to charge users for premium services. Once you get significant traction, you'll want to start generating income from your users. In order to do be able to do that later, you should design your website's software to support this capability.

This model have worked for large websites like LinkedIn and smaller ones like Conceptispuzzles.

It makes sense to define what is a business model in first place. What I prefer, is a concept of nine building blocks of the model (by A.Osterwalder). You may want to check out this article about how to define and compare different biz models.