how would one come up with an early valuation for a content/community based startup (like twitter, pinterest)

I'm trying to generate a reasonable valuation for my content/community based startup. It has not launched yet. How do content/community based startups come up with their first valuation. For example, any ideas on how twitter convinced investors of its worth early on, before it exploded? Or pinterest? I understand a common practice is to project user growth and assign a dollar value to each user. That just seems completely arbitrary for a pre-launch community based startup. How does one determine the $ amount per user before the startup is making any money or even really has users?

Valuation is part science and part art. As with most science, as you peel back the layers on current processes, there's generally less certainty than hard and absolute measures.

With that said, valuation for b2c and social media models usually ties back to many financial models. In general, it's how much revenue each unique user is likely to generate and the influence over others' buying decisions. So, if Kim Kardashian tweeted for her followers to buy NoNameAv, what's the likely conversion rate. From the conversion rate, you can derive projected revenue. There's more variables that go into valuation of a tweet but it ultimately ties back to revenue.

For a community-based startup, start by specifying your target market. For example, 20-30 year old females, likes music, grad school education, lives in NYC, buys name-brand fashion apparel, etc. From a model of a typical user, you can forecast the amount of time they might spend on your site. Multiply that time with the number of users at launch, user acquisition, "stickiness," and year-over-year growth. Figure out how you'll introduce some level of monetization strategy (freemium, ad, subscription, combination, etc). For example, banner ads, product placement, etc. Such an approach will provide some measure to start an initial valuation calculation with justifications. With that said, there's no single model and each model will have to be changed especially if you receive some sort of external funding.

Since modeling is fundamentally accounting, be ware of the dangers of getting "too creative" as Groupon realized.

Generally, it seems that valuations are relative to what similar companies are valued at.

Let's say your direct competitor is company X. Company X is officially valued at 1 million dollars. Company X has 100,000 users. So each user is worth $10 towards the company valuation.

If you have 10,000 users, you could easily argue that you are worth ($10 x 10,000 users) $100,000.

The more similar you are to company X (content, services, monetization strategies) the more justified your valutaion will be.

You just need to do a little bit of research (perhaps look for valuations that were made when a company was in a similar stage as yours).

*I'm not an expert, but I think this logic makes sense.

If you have no users or income your company valuation is pretty easy. ;)
Most serious investors get involved once you have some traction and growth.

Where you are currently at is putting together some information which is typically in your business plan. It's not really evaluating what your company is worth today as much as your projected worth down the road. You would identify your target audience, how big that audience is, how you will reach them, how many you think you could capture, what is the cost of acquiring those customers. You would make projected traffic estimates, projected ad revenue estimates, site any examples of successes of other web sites in different business verticals.

In your plan you would want to outline revenue strategies. Yes some companies create revolutionary products and don't really worry about income other than VC but I feel you need to be in an environment that supports that type of model. Which means you need to move to California :] But if that is your exit / revenue strategy to sell it and you think it's based on your users count versus the revenue your website is making then it is based on the type of audience, longevity with the website, how often they visit it, and many many factors. Plus, probably plenty of arbitrary ones. But if you have 1,000,000 members on your site and 50% of them use it 3 hours a day that's different then everyone checking it once a week for recent News.

I read that Instagram's acqusition by Facebook ~$1 billion broke down to about $30/user, Youtube sold for $1.65 billion to Google and probably had less users but lots of eye balls... I imagine if you broke it down per 'user account' it would have been much higher per user.

So, you have to make some assumptions - and more importantly be able to convince investors your assumptions are accurate.
But Youtube was acquired for