For a two-sided marketplace I am figuring out our marketing strategy. After launch I want to track retention rate and when that's good, push harder on marketing. How would you approach the marketing of our product?

It really depends on your vertical. I think you need to sit down and ask a yourself a couple of questions: The launch of a two-sided marketplace really requires both a demand and a supply side. Is one easier to access then other? Is there more of a supply side (are sellers looking to find a marketplace?) or is there more demand? Can you create supply or demand? Uber initially paid drivers to taxi people around, creating easier supply. Is it easily accessible? Do you have a good on-boarding process for both? Can you compartmentalize your vertical? If you're a bicycle marketplace, it might be easier to start with only electric bikes in college campuses. Don't bite off more than you can choose. Try to consider exceptional measures to cultivate earlier customers that may not be scalable.  For example, AirBnB sent professional photographers to some apartments in the beginning. Are there existing resources out there that your target buyers are already using? Can you mine Craigslist/LinkedIn resources? I believe that the consensus is don't rely on shows for big launches. And when it comes to press, sometimes less-sexy trade/niche press is far more effective for sales (not visibility) than something larger. Good luck!

We are also working on a 2-sided multilingual marketplace for freelancing services at http://galilea3.com. Our plan was to collect emails via a landing page, which we did for some time. We managed to get a few hundred people registered so that we could get some feedback from them while we continued working on our product in the background. One problem with using a landing page form, is that you collect the minimum possible data, usually email and name. Big forms kill conversion. Later there is no way to know if the person who registered is a buyer or seller. As a result there is no way to customise email campaigns to target buyers and sellers separately. Imagine what will happen when you send messages meant for buyers to sellers and vice-versa. Unsubscribes will hit the celeing and you will start losing users before you even launch. I am hoping that when the users start completing their profiles, we will be able to collect the rest of the information. As for the demand and supply side, we plan to first get the sellers to create their services in stealth mode, so that when we launch the beta version there will be something for the buyers to check out. Paul Graham said that startups need to do things that don't scale. I have noticed that leaving comments in the right places, such as forums and blogs, really works. Quite a number of people have registered that way. This is just one of many ways we get people to register without spending money on marketing. Of course, we will spend when the right time comes. I have also heard from a lot of startup owners that depending on PR is not a good idea as it turns to have a flash-in-the-pan effect. Search marketing, SEO, blogging, etc. are considered to be marketing methods that continue to bring users over a longer time frame.