Saturday, September 26, 2009

What is Twitter worth?

Yesterday's big news story, at least in valuation circles, is that private equity investors have invested $ 100 million in Twitter for a roughly 10% stake, suggesting a billion-dollar valuation for the nascent company.
Twitter, for those who may be living in the middle ages, has about 30 million members who post short messages (less than 140 characters) that other members can read (if they choose to follow your tweeting). Every celebrity (sports, politics, media) seems to be tweeting now. There are three questions that came up after the story:

1. How did the equity investors in Twitter come up with the $ 1 billion value?
We assume degrees of sophistication to private equity investors and venture capitalists that they usually do not possess. In my experience, venture capital valuations often represent back-of the-envelope computations with hefty discount rates (target rates if 30-60%) taking care of the uncertainty. I was not privy to the valuation of Twitter but I can read the tea leaves and guess how they valued the company. A few months ago, Facebook (a company that Twitter aspires to be at least in the new term) raised equity from a group of Russian investors, who attached a value of $ 6.5 billion to the company. At the time. Facebook had approximately 200 million members, which works out to about $32.5/ member. As of last week, Twitter had about 30 million members. Applying the $32.5/member to this estimate, I get $975 million (suspiciously close to $ 1 billion). This may be pure coincidence but given the pull towards relative valuation on the Street, I think it may explain the valuation.

2. Could Twitter be worth $ 1 billion?
"Could" is a very weak word. Of course! What Twitter has going for it is the numbers. With 30 million members, all I need to be able to do is to generate a small cash flow from each one and the valuation will be justified. A billion dollar value for a firm requires that the firm be able to generate about $ 100 million in operating income in steady state. (I am applying a 10% cost of capital, typical of mature firms, and assuming zero growth). With 30 million members, that works out to $3.33/year from each member. If you are a Twitterer, the question I would have for you is this: Would you be willing to pay an annual membership fee of $ 5 or $ 10 to follow your favorite celebrities thoughts? If the answer is yes, the billion dollars is paid for... If not, I will keep looking...

3. Is Twitter worth $ 1 billion?
Interesting question. As an ongoing business, I don't think so and here is why:
a. You don't buy a business that does not have a business model yet. Twitter has a lot of members but it really does not know how to make money of these members (yet). Advertising alone will not do it. Any blatantly obvious way to earn money (such as charging per tweet) will very quickly decimate the membership. So, where will the additional profits come from?
b. You are buying a business that may be a fad, at the peak of its faddishness: Twitter is hot right now, because it is in the news. However, most of the tweets that I read are inane: it is tough to be profound 24 hours a day and to express that profundity (is that even a word?) in 140 characters.

However, I think that you can justify a $ 1 billion value for Twitter at least to some investors and that is to think of it as an option. What you are buying then, when you buy this firm, is access to 30 million potential customers, who may not know each other but are tied to one another. There are at least two types of investors who may find this investment appealing:
a. A firm with deep pockets and products/services that may be appealing to the membership of Twitter. The 30 million members of Twitter tend to be techno-savvy, older than Facebook members (on average) and well off. They also tend ot think well of themselves or at least their opinions. To illustrate, Microsoft did take a position in Facebook a few months ago and I can see other companies with products (especially in entertainment) do the same with Twitter.
b. Risk money: While no investor in his right mind should be investing the bulk of his portfolio in Twitter, it may be a good investment for risk money, i.e., money you want to invest in high risk, high reward investments and are willing to lose. Spreading your bets across multiple investments like Twitter may create a portfolio that has a good risk/return trade off, especially if you can bring some selection acumen to the process.

P.S: Facing the scorn of my teenage daughter, I created an account on Twitter about 6 months ago. I have never tweeted but I have 228 followers. Scary!!!!


Mahesh Sethuraman said...

"If you are a Twitterer, the question I would have for you is this: Would you be willing to pay an annual membership fee of $ 5 or $ 10 to follow your favourite celebrities thoughts?"

No way.....
I can't imagine an internet user paying for any of the services offered on internet like FB, Orkut or Twitter!

The argument that the valuation may come from the investors' specific advantage by gaining access to 30 million people, still looks like shooting in the dark. We have already seen enough services like Orkut, FB, Linkedin or Blogger building a huge membership population in a short span of time. If its that easy, the cost of reaching out to such a large population through one medium will only get cheaper (in fact much cheaper considering the pace at which the internet economics keep changing).

Just imagine, if one day Wikipedia starts allowing advertisements in its page and decides to use it for some charity purposes or something, wouldn't it completely change the dynamics of the industry?

Aswath Damodaran said...

I am skeptical too, but being skeptical does not wipe out the probability of a breakthrough pricing model delivering huge payoffs. That is why I characterized it as an option rather than a conventional investment.

shikhil said...

Well if you look at it i am amazed that a search engine can have a valuation of USD 150b matching apple, ibm and even microsoft.

my question would be if that valuation is translated into per user annual fee than no one would use google.

you can compare apples with oranges on basis of their price per kg or taste but what if you have some thing that is not edible and doesn't have any weight!

Dennis Chiuten said...

One of the cofounders of Insight confirms your hunch about how they came up with the valuation:

"When we modeled it, we were looking at revenue somewhere between Google and Facebook. Google monetizes at $30 a user and Facebook is about $2 a user."

Immortal said...

its purely relative to Facebook valuation....


Mark said...

A piece we did a while ago on valuation of early-stage companies:

And applying a DCF methodology to early-stage companies:

Andre Vitorio said...

Great article. That being said, Twitter users are nowhere near as valuable as Facebook users. Facebook holds much more information about their member base than Twitter. It's much easier to monetize from Facebook's audience compared to Twitter's.

In this market what matters is how much info you know about your user in order to translate that info into needs and send targeted advertisement to them. This is the reason why Google's value is up in the sky. Even though Google doesn't hold half of the information per user a website like Facebook does everyone tells Google exactly what they want.

Anonymous said...

I don't really think that face book, twitter were made with the idea to get users to pay $5 - $10 annual membership, that would be like sitting on the earth and hoping for the moon. What i personally feel is that the large collection of members, their database is of considerable value if used in an appropriate manner. Though it makes sense to view this as an option rather than a conventional investment, however the option here would be if they are able to use the large membership base to their advantage which i think is possible if not easy. Thing of joint ventures between Facebook and other companies making a new products (example a travel company arranging exotic travel around the world) which has more likely hood to generate annual revenues through the members and using this database to advertize their new product to the right people in the data base. The joint venture could then have a contact valuing the database that Facebook has. This would remove legality issuse regarding sale of the database as Facebook would be using the database itself.

Anders Sundelin said...

In any multi-sided business model you have the chicken-and-egg problem, and the importance in this case I believe is to get scale of users generating data to be mined. To charge users for creating data or follow celebrities would be crazy.

There are many speculating about future premium versions using a Freemium model, but with or without paid versions I believe there are so many great ways to monetize the Twitter data. Twitter is increasingly becoming a standardized platform for real-time data, now used by actors such as Microsoft and Google, and their search algorithms will probably be updated with algorithms ranking pages on links from Twitter users, RTs by other Twitter users, by the number of followers, by webpages linking to Twitter profiles etc.

Content in Twitter messages will probably soon trigger other services. Say that I Twitter about a book that I have read, it would be in the interest of Amazon to ask me to comment the book on Amazon and recommend a new one, in the interest of the writer to get my feedback, in the interest of my local cinema to tell me about when the movie is coming out, and in the interest of mine to know what friends also read the book and who liked it and who didn’t. By cross-referencing twitter messages with past messages, messages by others or from friends in social networks such as Facebook or LinkedIn, there will be great opportunities for interesting services and business models.

The key question is how much of the value created that will be captured and controlled by the company, and how much of the value that will be extracted by others mining their data.

The business model database