Saturday, February 09, 2013

Will users' awareness of Facebook inability to protect privacy be the Facebook killer?


Facebook monthly active users (MAUs) were 1.06 billion as of December 31 2012. The company noted that now stores over 100 petabytes of media (photos and videos). Hence, each Facebook user needs on average almost 100 Mbyte for photos and videos. In addiction, there's data necessary for personal user profile information, but as it's text based it's likely to be some size level lower. 

A great deal of data. But why does Facebook business model need so much data? "Our goal is to help every person stay connected and every product they use be a great social experience," Mark Zuckerberg says. In fact, Facebook business model is focused On Line Adverstising, payments and other fees (aka virtual goods; e.g. fees from games). Sandberg also views local businesses as a growth opportunity for the company. "Local is the holy grail of the Internet, but local businesses are not very tech savvy. More than 40 percent have no Web presence at all. Facebook has a huge competitive advantage because they are using Facebook personally, and seeing messages from other businesses," she said. "It's a smaller leap (to use Facebook), and the numbers bear that out -- 7 million small businesses are using Facebook Pages on a monthly basis, and hundreds of thousands are upsold to become advertisers."  

Average Revenue per User (ARPU) Q2'12 was $1.28 with some differences:
  • US & Canada: $3.20 
  • Europe: $1.43
  • Asia: $0.55
  • Rest of the world: $0.44
Q2'12 revenue totaled $1.18 billion, an increase of 32%, compared with $895 million of Q2'11:
  1. Revenue from advertising was $992 million, representing 84% of total revenue and a 28% increase from the same quarter last year.
  2. Payments and other fees revenue for the second quarter was $192 million
This revenue structure is rather consilidated as full year 2012 results shows:
  • 84% revenue from adv
  • 16% revenue from payments and other fees.
Interesting that mobile revenue represented approximately 23% of advertising revenue for Q4'12 up from 14% Q3'12. Globally, 2012 revenue increased 27% up to $5.089 billions from $3.711 billions of 2011.

Different music on operating margin & net income side.   2012 (non-GAAP) income from operations decreased to $538 millions from  $1.756 billions of 2011. Looking at the main costs and expenses:
  • Research and development (+261% vs 2011): $1.399 mln 
  • Marketing and sales (+128% vs 2011): $896 mln   
  • General and administrative  (+184% vs 2011): $892 mln
Facebook is obviously investing heavily on R&D , MKTG & Sales and it's completing its expansion. So, the 2012 net income shouldn't be interesting (!!!). Actually, 2012 net income is $53 mln vs $1 billion of 2011 (before IPO). That's a clear sign that Facebook is a long term investment and not a speculative bet for extempore investors, but that's not the point here. The point is that Facebook went public to get the necessary resources do do these technology/product/corporate investments and developing a unique competitive avdantage (Facebook MAUs are actually unique!) and how the return could be affected by the untechnological factor of users' awareness of Facebook inability to protect privacy.

Where is Facebook investing and where it can get a unique advantage?
  • Big Data (for final users: search, social graph search, recommendation engine, etc.; for clients: Sentiment Analysis, Marketing Campaign Analysis, Customer Churn Analysis, Customer Experience Analytics, etc.)
  • Mobile support & integration (iOS, Andrioid, etc.)
  • Product development (Messenger for Android and iOS, Facebook Camera available in 18 languages, new advertising products such as Custom Audiences, Facebook Exchange, Offers, and mobile app install ads, Created Facebook Stories, global App Center, etc. ) 
  • Corporate development (first international engineering office in London, etc.)
Talking about product, Google+ is considered actually a better product by someone, but unluckily Google+ hasn't the Facebook MAUs. I think MAUs and its exploitation is the core asset Facebook can use to get a unique competitive advantage over competitors and get satisfactory return to investors.   

Talking about MAUs explotation, technologies behind Big Data are
  • Hadoop & related fameworks (Hadoop Distributed File System, MapReduce, Hive, Pig, HBase, Flume, Oozie, Flume, Ambari, Avro, Mahout, Sqoop, HCatalog, BigTop, etc.),  
  • NoSQL databases (HBase, Cassandra, Aerospike, MongoDB, Accumulo, Riak, CouchDB, DynamoDB, etc.). 
Effectively, some of these technologies were initially developed in Google and Facebook. Now, they most live as open source projects and applied by Facebook in order to make applications / services interesting for Facebook users/clients, such as:
  • Recommendation Engine: Web properties and online retailers use Hadoop to match and recommend users to one another or to products and services based on analysis of user profile and behavioral data. LinkedIn uses this approach to power its “People You May Know” feature, while Amazon uses it to suggest related products for purchase to online consumers. Initial question turns here into the following one: how users' awareness of Facebook inability to protect privacy can prevent users to recommend some kinds products or users? Is there particular products or relationships between users more sensitive than others? 
  • Sentiment Analysis: Used in conjunction with Hadoop, advanced text analytics tools analyze the unstructured text of social media and social networking posts, including Tweets and Facebook posts, to determine the user sentiment related to particular companies, brands, products or politic parties. Analysis can focus on macro-level sentiment down to individual user sentiment. Initial question turns here into the following one: how users' awareness of Facebook inability to protect privacy can prevent users to express their opinion about companies, brands or products? Is there particular companies, brands, products or politic parties more sensitive than others? 
  • Social Graph Analysis: In conjunction with Hadoop and NoSQL databases, social networking data is mined to determine which customers pose the most influence over others inside social networks. This helps enterprises determine which are their “most important” customers, who are not always those that buy the most products or spend the most but those that tend to influence the buying behavior of others the most. how users' awareness of Facebook inability to protect privacy can prevent users to influence others users? Is here particular companies, brands, products, political issues or arguments more sensitive than others?
A Final metaphor 
In information theory, the Shannon–Hartley theorem tells the maximum rate at which information can be transmitted over a communications channel of a specified bandwidth in the presence of noise.



        C =  B \log_2 \left( 1+\frac{S}{N} \right)

where C is the channel capacity (bits per second), B is the bandwidth of the channel (hertz), S is the average received signal power over the bandwidth, N is the average noise or interference power over the bandwidth (watts), S/N is the signal-to-noise ratio (SNR). Hence, the more the noise power, the more the signal power necessary to get a given capacity. Or, fixed the signal power, the more the  noise power, the less the channel capacity.   

Perhaps, we can think the channel capacity (C) as the value delivered to Facebook clients by services such as Recommendation Engine, Sentiment Analysis or Social Graph Analysis; the signal power (S) as the effectiveness of Big Data technology in order to discover such information; the noise power (N) as the interference because of users' awareness of Facebook inability to protect privacy in letting users having a full social experience; the bandwidth (B) as the Facebook MAUs. 

Hence, if that metaphor works, given Facebook MAUs, the more the users' awareness of Facebook inability to protect privacy, the more the effectiveness of Big Data technology to get the same value to clients, the more R&D investments, the higher the price for Facebook clients ... the less the likelihood that Facebook business model works... 

Sunday, February 03, 2013

WSJ: Big Data Pay Premium ‘Highest It Will Ever Be’

According to a January report from the IT research firm Foote Partners LLC, which gathered compensation data from 2,435 employers. For example, expertise in Hadoop and Cassandra, platforms capable of processing massive amounts of unstructured data like feeds from social media, commanded pay premiums of up to 16% and 14% respectively.