Thousands of years ago, warlords, kings, and prophets changed the world. Today, those dreams belongs to scientists, innovators and technologists. I shall follow mine . . . RIP Steve Jobs.

And only Uncle Sam can stop them . . .

Here is the deal on the long tail of display advertising - its all about data - “behavioral”, contextual, search, location, and social. In the premium inventory world you can sell ads simply because you are WSJ (online) or American Idol (TV) - when it comes to 99% of inventory out there, especially online, you either got to sell “audience” (a polite way of saying, “mr advertiser, dont worry about the 100 sites you’ve never heard of, focus on the fact that the people on those sites are exactly the people that you ask for”) or you sell “performance” (a polite way of saying “mr. advertiser, dont worry about the 10,000 sites you’ve never heard of, focus on the fact that you only pay for a visit to your site”). For either cases - data either allows you to create the right audience segments or help optimize click yield for a given page view inventory.

The problem for Google is that while it owns the biggest treasure trove of data of anyone out there - the portion of that data that is proprietary to Google (search intent) is lumped in the same bucket by the government and consumer advocates as “evil behavioral.” The second problem for Google is that due to strict consumer protection laws, it is not allowed to use PII (personally identifiable information) for targeting or attempt to tie the disparate data it has for a particular user back to a single identity - be it email address, real name, or social security number. (common restriction for ALL ad networks).

Here is the crazy part, Facebook’s advertising model is entirely based on “personally identifiable information” - not just of the person seeing the ad but their friends as well. Facebook social graph is not some anonymous “those who bought x also bought y” graph popularized by Amazon, but it is literally “Joe is friend with Mary and if Joe clicked/liked on an ad/advertiser, Marry might also click/like that ad/advertiser.” To add fuel to the fire, Facebook will put Joe’s picture next to an ad for Marry. Added in the “behavioral” signals based on sharing (rather than searching), you have an entirely new and deeper dimension of data for targeting.

It gets even scarier. Currently, Facebook’s advertising inventory is captive on Facebook.com and perhaps it doesnt have a good business reasons for building a network (its reach is already unrivaled AND its effective fill rates are probably under 50%). However, if Facebook aims to be a platform and expand its ecosystem - building out a monetization system is the only step it has left to tackle.

PLUS FacebookConnect has already build a network of publishers effectively rivaling in size and scale to that of Google Adsense. What does FacebookConnect has to do with advertising you ask? Unless the government stops Facebook, I can see the argument that by using FacebookConnect, the user has essentially opted into allowing the publisher and Facebook to use its PII augmented social graph AND behavioral information to not just personalize the site experience but also to serve/target ads.

Imagine if Facebook knows exactly who you are, who your friends are - your and your friend’s profile (read demo) information, location history (geo tagged photos, status updates), likes, shares, content consumed (comments) etc etc. And not just on Facebook.com but on any website where its FacebookConnect enabled (and you have logged in/cookied). It will be able to give Google a run for its money.

Coming back to the first paragraph of the post - Facebook can use its data to create audience segments (from Justin Beiber fans to PRIZM Young Digerati) with virtually 100% accuracy rate. Facebook can also optimizing its performance based ad products using the various signals that it has on the user and his/her social graph not accessible by Google and others.

PII is the holy grail (or the spawn of satan, depending on which side you are on ) for advertising - Facebook is by definition PII (social is “personally identifiable”) . Any ad network that knows “who you are” - is going to be a force to be reckoned with. The only thing left, is for FTC to figure out whether the use of FacebookConnect by the consumer constitutes an opt in for the user to receive targeted advertising from Facebook.

The line is blurred - the usual 25 years olds probably could careless. Maybe that’s the answer.

pardon the meaningless post . . . just playing around with a new toy.

Interesting and thoughtful post from Bill Gurley. Even more enlightening comments, just showing how smart people are in the industry we live in.

For Google, the economic castle is clearly the search business, augmented by its amazing AdWords monetization framework. Because of its clear network effect, and amazing price optimization (though the customer bidding process), this machine is a monster. Also, because of its far-reaching usage both on and off of Google,AdWords has a volume advantage as well. . . . So here is the kicker. Android, as well as Chrome and Chrome OS for that matter, are not “products” in the classic business sense. They have no plan to become their own “economic castles.” Rather they are very expensive and very aggressive “moats,” funded by the height and magnitude of Google’s castle. Google’s aim is defensive not offensive. They are not trying to make a profit on Android or Chrome. They want to take any layer that lives between themselves and the consumer and make it free (or even less than free).

A simple value chain or five forces analysis will help someone come to the same conclusion. That competitors is not just someone selling the same product as I am. Instead, the security of my strategic position in the industry is dependent on my suppliers as well as my distribution channel. No one wants to have Walmart being the one and only distribution channel, and in Google’s case, Google doesn’t want to rely on iOS or Blackberry to be its main method of reaching the mobile user. Forward integrating to the OS layer, taking out a potential source of dis-intermediation or competition is a good & common strategy. If nothing else, it gives Google better negotiation leverage with Apple.

The problem worth exploring though, is at what point, does Google’s position in search puts them in the classification as a “monopolist” and when that happens - using profits from one industry to subsidize the lack of profits in another for marketshare is very much considered illegal. (”if it hurts the consumers” being the pivotal legal distinction). We’ve seen Microsoft do this in the browser space. Google has one-upped the “micrsoft way” by doing the same playbook in browsers (chrome), OS (android), application (docs), and more . . . Microsoft was once the great destroyer of industries - instantaneously gaining huge market shares while eliminating competitor business models and thus imploding addressable marketing opportunities.

Google’s vigor, persistence, and focus is refreshing and unseen since Bill Gates was at the helms of MSFT dismantling the desktop productivity application industry. Welcome the new juggernaut, nothing lasts forever, this only makes the game more interesting.

(and yes, its refreshing, I dont think its up to Google to self-censor, but it is up to our legal systems to determine limits and freedom of allowable actions)

Universal acceptance is one of the flanks that payment companies have been leaving open for over 10 years. Startups like Paypal 10 years ago, and Square most recently, took the open invitation to build successful businesses which are starting to threaten the marketshare of Visa, Mastercard, and American Express (through either owning the merchant or payer relationship and eventually creating their own network).

(Payment companies did this to protect the 2-4% interchange/merchant fee cash cow - and to be honest, its about time that gravy chain ended - BTW they really need to redefine their fee structure depending on whether its a credit or a debit transaction to stay competitive)

Anyways, it looks like Visa has woken up. It announced the ability for any visa branded card (debit, prepaid, credit) to receive money where previously the closest thing it can to is to receive a charge back/reversed transaction. After a decade of being dis-intermediated be wallet companies like paypal (and soon Apple and Google?), the card itself is finally becoming the wallet.

Ironically, if anyone remember, Paypal didnt want to be a wallet in the first place. It was directly crediting the money a person receives on his or her credit card (without a corresponding transaction) and got in trouble with the credit card companies. So instead, it decided to start hold funds themselves - which again created an incentive for the account holder to link up their Paypal account to their bank account. Which in the end, created PayPal’s business model (i.e. arbitraging between ACH & credit transactins) as well as help Paypal create the leverage position it has in the marketplace today vis-a-vis the credit card companies. Like in the telecommution space, the value is moving from the network/transport layer to the application layer for the payment industry - now that visa has updated its network rules for the universal acceptance future - it needs to figure out its application strategy.

I wouldnt call it luck, but life takes you in interesting places.

Just copying Danny’s post on Venturebeat. (Notice how much more friendly it is, when you take off just a single word. I like to be friendly )

Maybe I’m piling on. . . but oh well. I too worked at eBay at the same time as Danny. And yes, it was a magical place. . . I have yet to work at a company that was as customer focused as eBay. At eBay, we called them our “community” of buyers and sellers. Every meeting ended with “what would the community think?” Every decision considered not just how the company could make more money, but how the community could benefit. eBay “crowdsourced” its business before 2.0 was hot. It was the original “social” startup . . . but is it now?

I’m going to focus on macro issues with the business rather than specific problems since I dont have a competing start up

The auction is dying - The community intricately tied to the gaming mechanics of the “auction” is/was what eBay was built around. The network effects of an auction based business model is much higher than that of a fixed price online retailer. And as a result, as long as, auctions are the preferred purchased method for consumers to purchase & sell online, the barrier to entry was extremely high. Unfortunately, that is not the case anymore. As more buyers move toward fixed price purchasing behavior, eBay began to lose its competitive advantage in the marketplace. From the end user perspective, eBay.com could properly support an optimal fixed price experience no matter how hard it tried. Adding to the problem is that eBay could not foresake the auction oriented buyers either given its still the majority of its revenue base - and as a result it got stuck in no man’s land.

Amazon broke the network effects - Blocking and tackling matters. Amazon made a concerted effort to court eBay sellers on a category by category basis. They made it easy for seller to migrate to Amazon. They made it financially attractive to sell on Amazon. Amazon had permanent pages that SEO’d so much better than eBay that it did not rely on SEM for traffic as much as eBay did. Amazon simply out executed eBay on catering to sellers.

Google became the superset - As successful sellers graduated beyond eBay, a seller’s website became his or her definite source of content and inventory. Regardless of economic outcomes, no one seller like to rely on a single channel for their revenue. And even though in practical terms, relying on Google is just as bad; building a website, building traffic on Google Adwords, submitting products to Google Base atleast gave sellers a visceral sense of control that they dearly lacked on eBay. Once they understood that they could build a business off eBay, the sellers are gone. On the buyers side, once they realize they could either bid all week to get a best price or do some comparison shopping off Google for the same outcome - auctions (for those that were only attracted to it for financial results) became an after thought.

eBay is not going to die though. I still spend more money on eBay than any other etail destination. Its just that its heyday is gone. Nothing lasts forever, we all become just a cash cow one day - and thats really not that bad.

Part II of my previous post. . .

While the whole world continues to obsess over NFC chip’s inevitable arrival on Android and iPhone, I would like to argue that they/we are missing the point. If there is anything we learned from Pay-By-Touch, its that there is really very little difference between all the different proposed way of consummating a transaction. Swiping a card, waving a phone, bumping phones, scanning a finger print - all pretty much the same . . . perhaps milliseconds improvement in speed, utility, and usability. So how is it again that we intend to change ingrained consumer behavior towards so-called mobile payments?

A quick definition to start perhaps. I hate the word mobile payments. While many startups/press seems to want to make you believe Dog = NFC, Tail = Payment; I believe Dog=Payments, Tail =Mobile Applications. (i.e.Mobile Payments is the proverbial tail wagging the dog).

I rather call it “payments enhanced through mobile applications” There I just gave it away. I believe mobile applications can 10x the volume of electronic transactions offline. To the extend that NFC chips can create a better application experience it is certainly an important innovation (e.g. wave phone once vs. slide card, pull smart phone) - but by itself, it certainly doesn’t do much to make things interesting.

While I do not have the ultimate crystal ball on what are the killer apps for mobile payments (ya I know, kinda anti climatic) I believe there are three rough application categories where the catalyst will appear.

The convergence of prepaid-debit-stored value-checking - Credit will be scarce going forward. This is the new normal. Traditional retail branch banking model is under significant margin pressure in the U.S. AND probably will never develop in emerging markets. Mobile applications have an opportunity to create a new user experience (e.g. customer service) much the same way ATM did 20 years ago, and allow new entrants as well as incumbents to re-create their entire value chain. The winner will recombine existing network assets and card based features such as reload networks, direct deposit, CPG retail channels, mobile applications to create a new non-credit based banking product for the mass market.

Universal acceptance - We need to blur the line between peer to peer payment and merchant acceptance - and thus significantly reduce the friction needed to TAKE electronic payments and thus enable magnitude increase in the amount of money stored electronically rather than in hard currency. Square has done a great job thus far, but have not taking this far enough. Paypal is starting to “get it” and move offline from its e-commerce roots. In the end, the payment networks will have to get involved and re-define many of its archaic rules. (and also face up to the potential for cannibalizing merchant interchange fees in exchange of survival and relevance in the new world).

Moving up the payment stack - Here is where knowing how to create a great user experience will help win this lottery. Search, discovery, offers, reviews, reservations, sales, ticketing, rewards, navigation are all major vectors of the local commercial experience into the eventual payment function. The killer app will be able to embed and integrate the payment function as a seamless part of the overall local commerce experience. Whoever owns this app, will take pole position in exacting tolls from the rest of the payment value chain. The contenders are wide and varying - daily deal companies? payment networks? banks? local mobile app developers? online payment companies? prepaid card vendors? search engine? social network?

In the end the winner of the mobile payment “revolution” will need to win in all three inter-related areas. Anyone with a peripheral knowledge of the space will understand how big of a task this is (like the iphone, mostly through coordination, blocking & tackling, and selective innovation). . . but boy I cant wait. . .:)

I’ve convinced that we’ve finally ready hit the hockey stick. I’ve been waiting for this for over 7 years. I’ve spent various parts of my career bouncing around some seemingly unrelated markets: advertising, local, payments, e-commerce, and most recently mobile. Through bouncing around quite a bit, I had always insisted that there is a single thread of commonality in all these companies and industries. That a series of cascading events is slowly converging (all at different pace but generally towards the same direction) and that if we are lucky enough for them to not miss each other, these industries, as we know it, will forever be changed and re-constituted.

Mobile – Mobile application proliferation to be exact. Mobile is the great catalyst. Always on. One to One. Merging the digital (online) and analog (offline) by providing location context to digital applications. For the first time, “where we are” is constantly fed as an input to the giant distributed processing system known as the Internet. Mobile changes everything.

Local Digital Advertising – Not many internet and technology companies have cracked the local / SMB nut (Intuit, MSFT to name a few, but only through brute force). Selling anything to SMB’s had always been a long hard road. It was no different in the online advertising space (I lump online marketing into the space for the sake of simplicity), local lags the general industry by atleat 5 years. In 2006 the world changed. Dollars started to flow out of newspapers and directory publishers into the digital world. Only a few hundred millions at a time in the beginning… now over a billion of dollars are gushing toward digital every year. (Groupons, ATT Interactive, ReachLocal). But the story is only half written.

Payments – In the online/web world, the ability for advertising exposure to be directly attributed to an transaction e-commerce and payment through click stream (the digital version of our “footprints”) forever revolutionalized online advertising. We can now calculate ROI. (see Wanamaker). Which enabled display and search advertising to grow unapologetically in scale with consumer media consumption. And thus giving birth to internet payment giants like PayPal and E-commerce monsters like Amazon & eBay. The story is about to be re-played.

The Big Bang.

The convergence of local advertising dollar coupled with adoption of mobile applications is about to forever change the payments industry. One thing I have been beating people over the head since 5 years ago is that “Mobile Payment” is not mobile at all. Mobile payments IS payments. Its offline. Its us swiping our card at Best Buy and Starbucks. Its us paying for bridge tolls using an NFC wallet. (or in Asia, paying for subway pass waving our mobile phone). Whatever that is about to take place is not about an incremental “new industry/opportunity” but the complete dis-integration and re-constitution of one of the largest industries in the U.S. These convergent events is about to impact an industry 5-10X larger than “online payments.”

Mobile Payments has been the holy grail of many many failed company for the last 10 years. Their timing was off.

Here is the thesis. Locational context on a mobile device, fed into mobile applications enables the tracking of our “physical footprints.” (think offline clickstream). Local advertising dollars are quickly flowing to digital advertising. There is finally a critical mass of merchants engaged in advertising to enable a relevant user experience (yes, in local, consumers view advertising as valuable content). Advertisers will begin demanding proof of ROI for their dollar spent. Exposure to mobile “ad/offer” for an offline merchant can be tracked along with subsequent visit to that merchant through location aware smart devices and applications. If payment is also conducted by the user through the mobile device, the entire process becomes a closed loop transaction. We can completely track a consumer from impression, movement/location, to payment. Are we ready to blow up the way advertising is sold, bought, and charged? (has Groupons stumbled onto something?)

AND mobile advertising will become better than online advertising. One of the problems with online advertising is its inability to capture brand advertising budgets because its impossible to quantified “probability of purchase” (awareness/recall measurements is really a proxy for probability of purchase) given that click stream is really only 1/20 of a users overall behavior. (did you buy Pampers instead of Huggies at Target after seeing a Pampers banner ad on ESPN.com?). How does recency of exposure tracked PERPETUALLY change the way awareness advertising is planned and bought? Done right, mobile payments will open up the flood gates of brand advertising dollar onto the mobile device.

Don’t get too excited. Just a few problems. Privacy for one. And really the bane of all mobile payment startups – what is the value proposition and catalyst to users that will convince someone to change their ingrained behavior of “swiping” a credit card to “waving” their phone? . . . The consumer killer app(s) for payment is almost here.

Ruminating on a few esoteric thoughts this afternoon. . . nothing to preach . . . just a bunch of questions for myself.

Say you have a set of feature you plan to implement with the goal of achieve X from metrics perspective. X could be X seconds improvement on time on site, or it could be X % improvement in visit / MUV (loyalty), or it could be X incremental transactions / month.

The set of features could be labeled F1,F2,F3,F4 and F5. Outcome X = F(F1,F2,F3,F4,Ff5,).

In the software world, since we burn CD’s (or send over the internet updates); typically, product managers would bundle F1 - F5 into a single release and pray that after release, objective X would be reached.

In the web world, we are taught to release as often as possible, so over the length of, say three month, we would sequentially release F1 through F5. This gives us a opportunity to isolate the impact of F1 individually from the impact of F2 and so forth. Couple implications. First, we could abandon the remaining set of features if F1 is not impacting X as predicted. Second, we can actually model f’(X) over F1 through F5 by observing and measuring X as these features are released.

1) After thousands or hundreds of observations what if we are able to create a few functions (linear, quadratic, exponential,log, power etc) that can generally describe F(F1-F5) for different type of features (e.g. social, worflow improvement, viral, merchandising, etc) ?

2) If we are able to determine which function a certain set of feature would follow, can we predict impact of F5 based on incremental information we gathered from F1 through F4?

3) If we can be fairly certain of (1), can we set a confidence level and decide when we should abandon the roadmap because F2 or F3 (depending on confidence level) did not achieve its impact ?

4) If we are even smarter, can we sequence roadmap F1 through F5 so that we release/build feature that gives us the most amount of information with the lease amount of investment ?

hmm…

10% raise. 23,000 employees. Or ~$400M in incremental annual expenses.

$2.55 in quarterly income. About $10B in net income annually. ~5% hit to net income annually. . . . and the stock goes up by almost $4 (from $619 to $623).

Listen up corporate America. Pay your employees more money and your stock will go up! Your employees are your most important assets!