Here's my two pennies..
This is a question that has pretty much kept me up at night for over ten years. In that time I've come to the conclusion that it's the right question, asked the wrong way. The question we're all asking is actually; How do you measure the value of intangible assets?
Sit with CFOs and ask this question (instead of the ROI of social media) and you'll get a decent conversation out of them, because they've been looking for the (right) answer for centuries.
To drive this point home consider the value of a public company when it's being sold. A significant component of its value is based on what accountants call Good Will. For example (from Wikipedia), 'a software company may have net assets (consisting primarily of miscellaneous equipment, and assuming no debt) valued at $1 million, but the company's overall value (including brand, customers, intellectual capital) is valued at $10 million'.
The point here is that calculating Good Will is entirely arbitrary. Yes there are some accounting rules but the number agreed upon has no rigour to it in a mathematical sense and yet huge companies are acquired by other companies where the value of Good Will (or intangible assets) can be as high as 60%-70% of the purchase price. Remember Time Warner/AOL, PWC/IBM, Compaq/HP? The list is endless. Share prices also have a good will element.
Most studies show that over 60% of mergers and acquisitions don't return value to shareholders mostly because of the inaccurate calculation of good will. As an aside, the post mortem carried out by the US government following the Enron/Worldcom 'issue' and the dot com crash that brought about the Sarbanes-Oxley Act (that was supposed to prevent the latest 'credit crunch') fund managers around the world committed to allocating 1%-3% of their revenues on research into calculating the value of intangible assets. There has been some progress but I won't go into it here.
Ok, so enough theory. At a more practical level I use a handy guide as a starting point to construct a business case for investing in social media. There are both tangible and intangible benefits. Intangibles are recorded under a column of Brand Awareness, Reputation and Customer Loyalty. Tangibles are recorded under a column label of Direct Revenue & Lower Operational Expense. Rows are labelled;
Advocacy/pass through revenue
Offline events (for talent acquisition, for example)
Product development & feedback
You can download it here;
I've found that most of the benefits are intangible (hence my diatribe above) but I'd be interested in how others would 'tick the boxes'.
My experience of working with senior managers and the boards of companies when justifying investment is that the most compelling benefits are in two main areas;
1. Demand Sensing. If a statistically significant number of customers say (or behave in a way) that tells me what existing product/service they want to buy and when they will buy it then I can plan/prepare. Think of how immensely effective this has been for Tesco's Club Card loyalty scheme, propelling it to the UK's number one grocer - and keeping it there. Three years ago a global brand approached me and said they were about to invest billions in a new product. They'd already done all their home work but as a quick additional check they wanted to ask a million people from a particular demographic if they would buy this product. They needed the results in 48 hours. They were prepared to pay up to $1 for every valid response. That's how important and valuable demand sensing is.
2. Product & Service Development. Companies understand the value of risk mitigation. If a million customers say they want a product or service then when it's brought to market, it kind of reduces the risk of getting it wrong. Unilever has been using online (closed) communities, under the radar for product development for at least 8 years and claim they've taken a third of their costs out of global R&D (Cluetrain: You want me to pay? I want you to pay attention!). That's a very, very big number. Witness the explosion of the use by Pharma companies of innocentive.com
The last project I worked on in social media was for Britain's eighth oldest company (founded in the 1600's) that no one has ever heard of. We ran a video competition. We devised it in a way that forced a high level of quality from participants. We got 53 submissions. Five of these were of sufficient quality to air on TV (as commercials) with no editing - they really were that good. The brand saved millions on commissioning commercials from traditional ad agencies. The business case was constructed on this premise from the outset.
In my humble opinion, the ground breaking research conducted by Yochai Benkler on the why Open Source (because that's what we're really talking about) is *a lot* more efficient than 'the firm' is easily the most insightful body of work to reference when building an ROI case for social media. His subsequent book called The Wealth of Networks is dry, academic and a struggle to get through but pretty much all the answers to this question are there.