beyond 140 for @lewmoorman

Twitter really is the next wave of search. I also find myself constantly searching twitter to get opinions about current topics and products.

While Yahoo/Google have built incredibly complex algorithms for search, Twitter is simply "what are people saying right now" and there's a lot of value in that.

I can definitely see people moving to twitter search when they need info about something, and search traffic means revenue.

I remember a couple years ago I was telling my dad that I thought Facebook would pose a serious threat to Google, using all their social graph info to target ads better than Google's algorithms could. But FB hasn't done it yet.

Can twitter? Can they target ads because they know exactly what people are saying right now? They know what's hot, and what people are most likely to want to click on/read?

The amazing thing to me is how valuable the search is and they have not even done anything to make it so. It is just basic returns for word mentions. You are right about the data. Super rich. Organized, it could be super valuable.

Twitter is a combination of search, chat, community, photo sharing, link-sharing and more. The fact I can do all of these things, in one place, in near-real time is pretty amazing. And is what makes Twitter now just interesting to me, but valuable. Take a photo, share it. Find a link, share it. And then find it all again later.

But unlike other search engines I find the results I want "within the confines of my own space" - Google doesn't offer that - at least not as the FIRST option. I have to dig deeper - and often what I am really interested in is "lost in the chaff".

Nice post. If I could buy Twitter shares now, I would.

Rob

Not to rain on the twitter love fest (menage a Twitter?) but here's a thought.

One of the appealing things about Twitter is it's genuineness and authenticity. You are searching real people's comments, not the results of a machine's complicated algorithms. That creates two problems with the comparison to Google:
1. How do you squeeze sponsored links on the search page results? I'm sure this can be done but bear in mind that the average Google search has about 8-10 paid links. They'd have to find room for that but leave plenty of room for the 'real people's' thoughts as well. And more importantly,
2. Does the presence of sponsored links somehow ruin the authenticity? While I've never attended, I've read that Burning Man was 'cool' and 'authentic' before advertisers sent people there to study it and make their products back home more 'burning man'. There is profit to be made once the mass market adopts, but does that push people to find the next Twitter?

I'm not saying it's not valuable - just that using the same valuation multiple less your 20% discount might be aggressive.

Until Twitter starts to execute on the search strategy we won't know.

Great points. When twitter goes commercial it will lose something. However, I think there are a couple was to limit this. One, only put ads on the search results. In some ways, these ads should add some value to the experience. When you are searching you are trying to get information and having vendors help you is not necessarily a bad thing. Plus, the search service is viewed as a separate part of the experience and I think ads will be tolerated. Next, you could create a whole new search service just based on the data from twitter. The whole point of the service would be search, where ads are expected.

All this said, you hit a critical point which is that making this transition will be delicate for twitter.

I have six general issues with your approach

1) The google and yahoo valuations already include growth, so you are
in essence double counting future growth when you apply the 3x
multiple. If either google or yahoo was not expected to grow (at least
nominally) their market value would collapse.

2) Given the nature of the beast, you can make a case that the value
of a portion of a social system is not linearly related to the market
share. In other words, owning 50% of the market is not exactly 5x as
valuable as owning 10% of the market. When talking about ~1% vs ~80%
this would need to be accounted for.

3) Why value a point of market share? If twice as many people use the
internet 3 years from now and the market shares stay the same, has
google not increased in value? I think it makes more sense to value by
eyeballs, then you can independently control for the growth in
eyeballs and the growth in market share. Also looking at it this way
helps you normalize to determine what the intrinsic value of market
share is in a connected market system.

4) Why assume 5% of internet users are like you? A year ago there were
1 million twitter users, some estimates for today are around 5
million. The internet (the search eyebal market) currently has 1.6
billion users . I think that a more likely ratio to use here is 5
million to 1.6 billion or .3%. Yes I realize that means I'm classing
you in the top three tenths of a percent of internet users (I'm saying
3 out of every 1000 internet users is a savvy as you), but when
talking about global advertising dollars on the web, including rural
china and grandma tildy that still uses AOL dialup, I'd say that's
more accurate than 5%.

5) A stated assumption is that twitter's monetization, when it
happens, would have the same characteristic value as google/yahoo.
This is actually a big assumption. A lot of the value of a twitter
search comes from direct individual opinions, you aren't always led to
a site. While arguably being more valuable to you as a consumer/
reader, it is equally arguable that this is less valuable to a firm
spending ad dollars.

6) The growth of internet usage is most heavily happening in Asia. 1/3
of the world lives in India and China. How much are they twittering?
Is the growth you perceive in the US a global phenomena or a western
phenomena. If you are valuing based on market share this is a key point.

This analyses was somewhat stream of conscientious, so these are not
ranked in any particular order but if I took #1 and #4 alone that
would require a reduction of 50 times to your valuation. I may be
undercutting with some of these estimates and I do acknowledge there
is some good upside opportunities for revenue using twitter data in
novel ways, but I do think some of my rebuttal should give you cause
for re-evaluating your napkin. Let's just say I wouldn't let you spend
my 3 billion on it without some further justification.

jason,few comments and rebuttals to you points. as always, i love your insights and skepticism.

1. i would say google and yahoo have little growth built in. yahoo is not growing and google is expected to flatline in the near term. yes, there is some built in potential for each user base and profit maximization, but it is hard to see a lot of growth in either company.

2. agree there are some network effects. some. when it comes to search the more important value determinant is ad placement engine. google gets more per click not because they have more people but because they have better technology. they get more clicks per view. after all, each search is a single transaction. they should be independently valuable.

3. i would say any of the growth in number one is captured by using this metric. it actually support my case. if you assume share is static but absolute growth is baked in then market share is the perfect way to value.

4, this is my weakest point, i admit. estimates this week put at 15mm. i also bet these are heavy users. so, it might not be 5% of users, but it could be close to the people that do 5% of the searches since they spend their lives on the internet. honestly, even if you cut this by 10x you can easily adjust some other estimates to get to my value.

5. is google's natural search valuable or not? if it is super valuable then it is hard to make this point. people do care who has money to spend on keywords. that is why they click. i do agree the profile might not be the same but it is hard to see that it would be highly different.

6. this is just upside. google is irrelevant in china. same with yahoo. so, i think us share still is a good proxy.

your turn...

1. I would agree that goog and yhoo don't have astronomical growth ahead of them, but I would still argue that their pricing doesn't fully currently reflect that reality. They have an aggregate analyst estimated 18% and 15% annual projected growth over the next 5 years (not too shabby for how large) and are still commanding earnings multiples of 30 and 50. To be fair in your calculation if you are going to apply a later multiple you should normalize the current market cap back to a reasonable level for a company growing at 15-20% a year. If you don't you are essentially pricing twitter as if it would trade at 90-150x earnings on the open market (you applied a 3x multiple above your derived value). Again this may just be me, but I don't like paying 150x earnings.

2. Google *got* more clicks because they had better technology. If someone else had better technology now, how would you find out? Network effect is real. BTW, I didn't really specify which way this would adjust the price, as I'm not sure myself. I merely said I think needs to be accounted for if you are deriving value from a starting point of market share. Which leads right up to #3

3. I acknowledge the interplay between this and #1 and #2, I'm just not as comfortable as you taking a gut estimate and assuming things even out. I'd prefer to separate the variables to control the estimate. I'd like to know what the true driving factor of revenue is (I'm positing its eyeballs), then separate take a stab at the growth of the market and of market share. Smooshing it all together amplifies the error and I'd like to know what my error bars are. I don't want a baked number, I want to measure the ingredients. Much easier then to run scenarios based on wrong predictions, especially when looking at a multivariate system where simultaneous corrections can move the end result in the opposite directions.

4. "honestly, even if you cut this by 10x you can easily adjust some other estimates to get to my value." That actually supports my point in the weakness of your valuation. That statement is an implicit acknowledgement that your numbers and approach have at least 10x wiggle room. Horseshoes and handgrenades indeed! I can use the same logic to hand-wave a 10x cut. Remember I'm not proposing a value, I'm just suggesting I don't agree with yours. Maybe I should come to the table with my own, but I'd have to spend some time ironing the above and below issues to have any confidence.

5. My point on this item is that when you search using twitter, what is the outcome. You get an opinion from someone and thats all well and good, but how often does your search drive you directly to a website? It does 100% of the time when you use google. Even if twitter search drove you to an end website as high as 75% of the time, I would expect that to be a different enough profile that its not valued the same way. Remember there is a ton of dollars at stake for knowing exactly how many seconds you have from customers attention span before they bounce off your site. It's even easier to not buy when you never go to the site to begin with.

6. I wouldn't be so immediately dismissive. I've seen quotes that google has anywhere between 15% and 30% of the search market in china. 30% of China is basically the US. Google China revenue only accounted for 1% of GOOG's total revenue for the first three quarters of 2008. Maybe a lot of the growth we are debating about in #1 and #3 is actually in GOOG's estimates based on expectations of that 1% rising. If that is the case and twitter has essentially zero penetration in China then I think its another reason to try and adjust out assumptions of growth based on your valuation approach. I haven't looked into India at all.

Also, what happens when the early adopters move on to something that's not being used by Oprah, P Diddy, Ashton Kutcher or Jimmy Falon. Twitter is not google just yet, it still has the opportunity to jump the shark before its really worth anything. I'm not saying that will happen, but clearly the uncertainty of it is worth some discount to me.