Hardware Lister: lshw

RPMs: http://dag.wieers.com/packages/lshw/

Sample Output: 
root# lshw |more
x.domain.com
    description: Desktop Computer
    product: HP dx5150 SFF(PE680AV)
    vendor: Hewlett-Packard
    serial: 2UA5450KR5
    width: 32 bits
    capabilities: smbios-2.4 dmi-2.4
    configuration: boot=normal chassis=desktop uuid=80C9A306-F6BE-1010-B4A0-AC7D0633B1F2

Question: How do I access my Packet8 voicemail?

Answer: From your Packet8 phone: 012-0555 From outside the U.S.: 1-650-353-4400 Toll free from the U.S.: 1-888-898-7887

Book - Leadership Challenge by Kouzes and Posner

Summary some of the lessons for me: Part 2: Model the Way - Chap 3: Find your voice by clarifying your personal values

• Look in the mirror.
• Take time for contemplation.
• Write a tribute to yourself.
• Record the lessons from the leaders you admire.
• Write your credo.
• Engage in a credo dialogue and assessment.
• Collect stories that teach values.
• Audit your ability to succeed.

- Chap 4: Set the example by aligning actions with shared values.

• Create alignment around key values.
• Speak about shared values with enthusiasm and confidence - even drama.
• Teach & reinforce through symbols & artifacts.
• Lead by storytelling.
• Put storytelling on your meeting agendas.
• Ask questions.
• Keep score.
• Do a personal audit.

Part 3: Inspire a Shared Vision - Chap 5: Envision the future by imagining exicting and ennobling possibilities

• Read a biography of a visionary leader.
• Think about your past.
• Determine the "something" you want to do.
• Write an article about how you've made a difference.
• Write your vision statement.
• Become a futurist.
• Test your assumptions.
• Rehearse with visualizations and affirmations.

- Chap 6: Enlist others in common vision by appealing to shared aspirations

• Get to know your constituents.
• Find the common ground.
• Draft a collective vision statement.
• Expand your communication skills.
• Breathe life into your vision.
• Speak from the heart.
• Listen first - and often.
• Hang out.

Part 4: Challenege the Process - Chap 7: Search for opportunities by seeking innovative ways to change, grow and improve.

• Treat every job as an adventure.
• Seek meaningful challenges for yourself.
• Find and create meaningful challenges for others.
• Add fun to everyone's work.
• Question the status quo.
• Renew your teams.
• Create an open-source approach to searching for opportunities.
• Send everyone shopping for ideas.

- Chap 8: Experiment and take risks by constantly generating small wins and learning from mistakes

• Set up little experiments and develop models.
• Make it safe for others to experiment.
• Break mindsets.
• Break it up and break it down.
• Give people choices.
• Accumulate yeses.
• Admit your mistakes.
• Conduct pre/postmortems for every project.

Part 5: Enable Others to Act - Chap 9: Foster collaboration by promoting cooperative goals and building trust

• Conduct a collaboration audit.
• Be the first to trust.
• Ask questions, listen, and take advice.
• Always say we.
• Create jigsaw groups.
• Focus on gains, not losses.
• Make a list of alternative currencies.
• Take a lot of human moments.
• Create places and opportunities for informal interactions.

- Chap 10: Strengthen others by sharing power and discretion

• Offer visible support.
• Assign critical tasks.
• Enrich people's jobs.
• Use modeling to develop competencies.
• Stop talking and start building at staff meetings.
• Enlarge people's sphere of influence.
• Educate, educate, educate.
• Create a learning climate.

Part 6: Encourage the Heart - Chap 11: Recognize contributions by showing appreciation for individual excellence

• Be creative about rewards.
• Make recognition public.
• Provide feedback en route.
• Be a Pygmalion.
• Foster positive expectations.
• Make the recognition presentation meaningful.
• Find people who are doing things right.
• Don't be stingy about saying thank you.

- Chap 12: Celebrate the values/victories by creating a spirit of community

• Schedule celebrations.
• Install a public "Bragging Board."
• Create a commemorative award honoring exemplary actions.
• Demonstrate caring by walking around.
• Show passion and compassion.
• Be a cheerleader - your way.
• Have fun.
• Set the example - plan a celebration right now.

Book website: http://www.leadershipchallenge.com/

Web 1.0 vs. 2.0 from Tim O'Reilly

Summary of the lessons from Tim's article: 1. The Web As Platform - The value of the software is proportional to the scale and dynamism of the data it helps to manage. - Leverage customer-self service and algorithmic data management to reach out to the entire web, to the edges and not just the center, to the long tail and not just the head. - The service automatically gets better the more people use it. 2. Harnessing Collective Intelligence - Network effects from user contributions are the key to market dominance in the Web 2.0 era. 3. Data is the Next Intel Inside - The race is on to own certain classes of core data: location, identity, calendaring of public events, product identifiers and namespaces. 4. End of the Software Release Cycle - Operations must become a core competency. The software will cease to perform unless it is maintained on a daily basis. - Users must be treated as co-developers 5. Lightweight Programming Models - Support lightweight programming models that allow for loosely coupled systems. - Design for "hackability" and remixability. 6. Software Above the Level of a Single Device 7. Rich User Experiences

Scope of ERP

- Analytics (Strategic Enterprise Management, Financial Analytics, Operations Analytics, Workforce Analytics) - Financials (Financial Supply Chain Management, Financial Accounting, Management Accounting, Corporate Governance) - Human Capital Management (Talent Management, Workforce Process Management, Workfoce Deployment) - Procurement and Logistics Execution (Procurement, Supplier Collaboration, Inventory and Warehouse Management, Inbound and Outbound Logistics, Transportation Management) - Product Development and Manufacturing (Production Planning, Manufacturing Execution, Enterprise Asset Management, Product Development, Life-Cycle Data Management) - Sales and Services (Sales Order Management, Aftermarket Sales and Service, Professional Service Delivery, Global Trade Services, Incentive and Commission Management) - Corporate Services (Real Estate Management, Project Portfolio Management, Travel Management, Environment, Health and Safety, Quality Management)

Information Security

Eight fundamental concepts in security: - Identification (user ID, application ID, or system ID. IDs are used to indicate to a system who or what is trying to gain access) - Authentication (validate the claimed identity of a user or resource) - Authorization (determine if an entity is permitted access to a particular resource) - Confidentiality (ensure that only authorized parties have access to sensitive data. Privacy is a concept related to personal information, whereas confidentiality is a mechanism or a goal) - Integrity (verification that data has not been garbled, modified, or lost inadvertently, system has not been tampered with) - Availability (DDOS, performance) - Non-Repudiation (legally admissible proof that a transaction occurred, such that neither participant in a transaction can later deny having participated) - Accountability (processes and technologies necessary to track system usage, identify inappropriate actions, and address the problem)

CRM Process - Customer Life Cycle

1) Acquire (Direct Marketing) 2) Enhance (Cross-sell, Up-sell) 3) Retain (Proactive Service)

launchmany-console 4.3.2beta Usage

Usage: launchmany-console [OPTIONS] [TORRENTDIRECTORY] If a non-option argument is present it's taken as the value of the torrent_dir option. arguments are - --ip ip to report to the tracker (has no effect unless you are on the same local network as the tracker) (defaults to '') --forwarded_port world-visible port number if it's different from the one the client listens on locally (defaults to 0) --minport minimum port to listen on, counts up if unavailable (defaults to 6881) --maxport maximum port to listen on (defaults to 6999) --bind ip to bind to locally (defaults to '') --display_interval seconds between updates of displayed information (defaults to 0.5) --rerequest_interval minutes to wait between requesting more peers (defaults to 300) --min_peers minimum number of peers to not do rerequesting (defaults to 20) --max_initiate number of peers at which to stop initiating new connections (defaults to 40) --max_allow_in maximum number of connections to allow, after this new incoming connections will be immediately closed (defaults to 80) --check_hashes, --no_check_hashes whether to check hashes on disk (defaults to True) --max_upload_rate maximum kB/s to upload at, 0 means no limit (defaults to 20) --min_uploads the number of uploads to fill out to with extra optimistic unchokes (defaults to 2) --max_files_open the maximum number of files in a multifile torrent to keep open at a time, 0 means no limit. Used to avoid running out of file descriptors. (defaults to 50) --start_trackerless_client, --no_start_trackerless_client Initialize a trackerless client. This must be enabled in order to download trackerless torrents. (defaults to True) --max_uploads the maximum number of uploads to allow at once. -1 means a (hopefully) reasonable number based on --max_upload_rate. The automatic values are only sensible when running one torrent at a time. (defaults to 6) --save_in local directory where the torrents will be saved, using a name determined by --saveas_style. If this is left empty each torrent will be saved under the directory of the corresponding .torrent file (defaults to '') --parse_dir_interval how often to rescan the torrent directory, in seconds (defaults to 60) --launch_delay wait this many seconds after noticing a torrent before starting it, to avoid race with tracker (defaults to 0) --saveas_style How to name torrent downloads: 1: use name OF torrent file (minus .torrent); 2: use name encoded IN torrent file; 3: create a directory with name OF torrent file (minus .torrent) and save in that directory using name encoded IN torrent file; 4: if name OF torrent file (minus .torrent) and name encoded IN torrent file are identical, use that name (style 1/2), otherwise create an intermediate directory as in style 3; CAUTION: options 1 and 2 have the ability to overwrite files without warning and may present security issues. (defaults to 4) --display_path, --no_display_path whether to display the full path or the torrent contents for each torrent (defaults to True) --torrent_dir directory to look for .torrent files (semi-recursive) (defaults to '') --data_dir directory under which variable data such as fastresume information and GUI state is saved. Defaults to subdirectory 'data' of the bittorrent config directory. (defaults to '') --filesystem_encoding character encoding used on the local filesystem. If left empty, autodetected. Autodetection doesn't work under python versions older than 2.3 (defaults to '') --language ISO Language code to use: af, bg, ca, cs, da, de, el, en, es, es_MX, fr, hu, it, ja, ko, nb_NO, nl, pl, pt, pt_BR, ro, ru, sk, sl, sv, tr, vi, zh_CN, zh_TW (defaults to '') --keepalive_interval number of seconds to pause between sending keepalives (defaults to 120.0) --download_slice_size how many bytes to query for per request. (defaults to 16384) --max_message_length maximum length prefix encoding you'll accept over the wire - larger values get the connection dropped. (defaults to 8388608) --socket_timeout seconds to wait between closing sockets which nothing has been received on (defaults to 300.0) --timeout_check_interval seconds to wait between checking if any connections have timed out (defaults to 60.0) --max_slice_length maximum length slice to send to peers, close connection if a larger request is received (defaults to 16384) --max_rate_period maximum time interval over which to estimate the current upload and download rates (defaults to 20.0) --max_rate_period_seedtime maximum time interval over which to estimate the current seed rate (defaults to 100.0) --max_announce_retry_interval maximum time to wait between retrying announces if they keep failing (defaults to 1800) --snub_time seconds to wait for data to come in over a connection before assuming it's semi-permanently choked (defaults to 30.0) --rarest_first_cutoff number of downloads at which to switch from random to rarest first (defaults to 4) --upload_unit_size how many bytes to write into network buffers at once. (defaults to 1380) --retaliate_to_garbled_data, --no_retaliate_to_garbled_data refuse further connections from addresses with broken or intentionally hostile peers that send incorrect data (defaults to True) --one_connection_per_ip, --no_one_connection_per_ip do not connect to several peers that have the same IP address (defaults to True) --peer_socket_tos if nonzero, set the TOS option for peer connections to this value (defaults to 8) --bad_libc_workaround, --no_bad_libc_workaround enable workaround for a bug in BSD libc that makes file reads very slow. (defaults to False) --tracker_proxy address of HTTP proxy to use for tracker connections (defaults to '') --close_with_rst close connections with RST and avoid the TCP TIME_WAIT state (defaults to 0) --twisted Use Twisted network libraries for network connections. 1 means use twisted, 0 means do not use twisted, -1 means autodetect, and prefer twisted (defaults to -1)

BitTorrent RPM Problem

Problem: You downloaded and installed the RPM from the official site and got this error.

$ launchmany-console
Traceback (most recent call last):
  File "/usr/bin/launchmany-console", line 16, in ?
    from BitTorrent.platform import install_translation
ImportError: No module named BitTorrent.platform
$ python -V
Python 2.3.4

Solutions: Some people suggested upgrade Python to 2.4, or you could get the source code, read the INSTALL.unix.txt How to build an .rpm format package and install it: ----------------------------------------

Generate an .rpm format package:
$ python setup.py bdist_rpm
Then install the .rpm:
$ rpm -Uvh dist/BitTorrent-4.x.y-1.noarch.rpm

BitTorrent - Files renamed since 4.1.4

4.1.4 released for Windows & *nix Renamed most BitTorrent command line scripts: * btdownloadgui.py ⇒ bittorrent.py * btdownloadheadless.py ⇒ bittorrent-console.py * btdownloadcurses.py ⇒ bittorrent-curses.py * btmaketorrentgui.py ⇒ maketorrent.py * btmaketorrent.py ⇒ maketorrent-console.py * btlaunchmany.py ⇒ launchmany-console.py * btlaunchmanycurses.py ⇒ launchmany-curses.py * bttrack.py ⇒ bittorrent-tracker.py * btreannounce.py ⇒ changetracker-console.py * btrename.py =� REMOVED, use maketorrent*.py * btshowmetainfo.py ⇒ torrentinfo-console.py"

Supply Chain Management

Supply Chain Management - Forecasting and Planning - Sourcing & Procurement - Manufacturing - Warehousing & Distribution - Transportation Forecasting/Planning - Demand Planning - Supply Planning Sourcing - Supplier Development - Proposal Management - Negotiation/ Contracting - Supplier Administration Procurement - Planning - Purchasing - Purchasing Administration - Settlement/Payment Manufacturing - Planning - Configuration - Production Control - Manufacturing - Inventory Management Warehousing - Planning - Receipt Processing - Value Added Services - Shipment Processing - Warehouse Management Transportation - Planning - Sourcing - Operations - Transportation Admin

The Best Web 2.0 Software of 2005

A list of Web 2.0 apps

read more | digg story

nmap 3.95

Sample Output: nmap -v -A xxx.xxx.xxx.xxx Interesting ports on xxx.xxx.xxx.xxx: (The 1645 ports scanned but not shown below are in state: closed) PORT STATE SERVICE VERSION 25/tcp open tcpwrapped 42/tcp filtered nameserver 80/tcp open http Microsoft IIS webserver 5.0 111/tcp filtered rpcbind 135/tcp filtered msrpc 136/tcp filtered profile 137/tcp filtered netbios-ns 138/tcp filtered netbios-dgm 139/tcp filtered netbios-ssn 161/tcp filtered snmp 199/tcp filtered smux 445/tcp filtered microsoft-ds 513/tcp filtered login 1059/tcp open msrpc Microsoft Windows RPC 1234/tcp filtered hotline 1433/tcp filtered ms-sql-s 1434/tcp filtered ms-sql-m 1524/tcp filtered ingreslock 1666/tcp filtered netview-aix-6 2301/tcp open http Compaq Diagnostis httpd (CompaqHTTPServer 5.94) 3389/tcp open microsoft-rdp Microsoft Terminal Service 6667/tcp filtered irc 6668/tcp filtered irc 7000/tcp filtered afs3-fileserver 49400/tcp open http Compaq Diagnostis httpd (CompaqHTTPServer 5.94) Device type: general purpose Running: Microsoft Windows NT/2K/XP OS details: Microsoft Windows 2000 SP4 or XP SP1 TCP Sequence Prediction: Class=truly random Difficulty=9999999 (Good luck!) IPID Sequence Generation: Busy server or unknown class Service Info: OS: Windows

Parse phpinfo() for module settings

This function parses the phpinfo output to get details about a PHP module. /** parse php modules from phpinfo */ function parsePHPModules() { ob_start(); phpinfo(INFO_MODULES); $s = ob_get_contents(); ob_end_clean(); $s = strip_tags($s,'<h2><th><td>'); $s = preg_replace('/<th[^>]*>([^<]+)<\/th>/',"<info>\\1</info>",$s); $s = preg_replace('/<td[^>]*>([^<]+)<\/td>/',"<info>\\1</info>",$s); $vTmp = preg_split('/(<h2>[^<]+<\/h2>)/',$s,-1,PREG_SPLIT_DELIM_CAPTURE); $vModules = array(); for ($i=1;$i<count($vTmp);$i++) { if (preg_match('/<h2>([^<]+)<\/h2>/',$vTmp[$i],$vMat)) { $vName = trim($vMat[1]); $vTmp2 = explode("\n",$vTmp[$i+1]); foreach ($vTmp2 AS $vOne) { $vPat = '<info>([^<]+)<\/info>'; $vPat3 = "/$vPat\s*$vPat\s*$vPat/"; $vPat2 = "/$vPat\s*$vPat/"; if (preg_match($vPat3,$vOne,$vMat)) { // 3cols $vModules[$vName][trim($vMat[1])] = array(trim($vMat[2]),trim($vMat[3])); } elseif (preg_match($vPat2,$vOne,$vMat)) { // 2cols $vModules[$vName][trim($vMat[1])] = trim($vMat[2]); } } } } return $vModules; } /** get a module setting */ function getModuleSetting($pModuleName,$pSetting) { $vModules = $this->parsePHPModules(); return $vModules[$pModuleName][$pSetting]; } // Sample Usage getModuleSetting('gd','GD Version'); // returns "bundled (2.0.28 compatible)" Sample Output

[gd] => Array
(
  [GD Support] => enabled
  [GD Version] => bundled (2.0.28 compatible)
  [FreeType Support] => enabled
  [FreeType Linkage] => with freetype
  [FreeType Version] => 2.1.9
  [T1Lib Support] => enabled
  [GIF Read Support] => enabled
  [GIF Create Support] => enabled
  [JPG Support] => enabled
  [PNG Support] => enabled
  [WBMP Support] => enabled
  [XBM Support] => enabled
)

[date] => Array (
  [date/time support] => enabled
  [Timezone Database Version] => 2005.14
  [Timezone Database] => internal
  [Default timezone] => America/Los_Angeles
  [Directive] => Array (
     [0] => Local Value
     [1] => Master Value
  )
  [date.timezone] => Array (
     [0] => no value
     [1] => no value
  )
 )

Google Analysis

Industry Analysis

Google started when online information is expanding and the need for a better search engine is growing. At the time, searching is perceived by many businesses as a peripheral activity and consumers would not care if they have easy access to information. The concept of mass marketing was being used to advertise on the existing search engines with little targeting mechanism.

Mission Statement

“Organizing the world’s information and making it universally accessible and useful”

This statement is a very ambitious mission for Google but it also leaves many doors of opportunity opened for Google. Their recent acquisitions and projects cover many areas. Information comes from many sources and online information is only a very limited part of it. Google started out with online information and is growing into offline channels. Information from multiple sources has many different formats. For example: different file types, images, audio files, paper books, and so on. Thus, the big picture is chaotic and unorganized. Making information useful and universally accessible also include online searches as well as from other sources such as wireless and voice searches. The presentation of the search result is also important. The most relevant results should be at the top of the list for ease of access.

Business Goals

• To organize the world’s information, Google started with searching information from online websites and is expanding into other media types. • Universally accessible ability means whatever, whenever and wherever the users need the information. • Create the network effect, try to deliver an excellent product to attract more users and profitability will be much easier to obtain.

Value Proposition

1. For Consumers: most other engines at the period of 1998 were focusing on the repetition of certain keywords within a page to determine its importance for those keywords. Google took a different approach and considers an important page as a page with references from other important pages. With this concept, PageRank was invented to rank which pages should appear first on the long list of results. The idea of more relevant pages come first make searches for information quicker and consumers enjoy the accuracy of Google. 2. For Publishers: the ability to earn money for display targeted text advertisements on their websites. Ads are targeted to the content of the site and are considered as added benefits to online visitors rather than annoying distractions. 3. For Advertisers: advertising dollars spent on the right prospects, conversion rate is much higher than the traditional mass advertising.

Research Paper - Click Fraud & Future of Online Advertising

Executive Summary

Online advertising spending plummeted after the “dotcom bubble” is recovering with Google leading the industry in search marketing. However, this rising star is faced with a serious problem, click fraud. That is clicking on ads without any intention to buy the product but with the intention to generate revenue or to waste competitors’ advertising budget. This problem has implications on all parties involved in the process including the advertisers, publishers, search companies and the fraudsters. Additionally, it initiated an array of businesses providing service to prevent, protect or to assist these frauds. It has both business and technical issues that needed to address quickly before the confidence in online advertising begin to erode.

Introduction

Internet advertising is not a recent innovation; it has been around since 1994 with HotWired launching the first banner ads on the Internet (Riedman 106). However, Google recently lead the improvement to suit the new media era. Online advertising services offered by Google, Yahoo and Microsoft enable smaller advertisers to compete with big-budget media buyers. During and before the dot-com period, online visitors were exposed mainly with mass branding campaigns by large advertisers. Effectiveness was justified with potential higher brand exposure (Briggs 44). Advertisers were paying upward of $70 per 1,000 impressions. With both click and sale conversion rates of less than 1%, the result is an expensive acquisition cost of $700 per customer (Hoffman 180). Ads were displayed everywhere yet yielded low returns, both in term of visitors coming to the site and the end sale revenue. Few or no company focused into delivering relevant messages to the right consumers or offered what consumers really wanted and when they wanted it. Click-through rates (CTRs) from advertisement banners, defined as the number of clicks divided by the number of ad impressions, were declining sharply (Dreze 9) as users learned to ignore and considered advertising as a distraction from their web browsing activities (Cho 94). Therefore, marketers explored different ad delivery methods to get back visitors’ attention. One of which was the infamous pop-up advertisement format. Pop-ups were too intrusive for many users and they widened the relationship gap between marketers and consumers. On the road to uncover the ideal online advertising model, that is not only highly effective for advertisers but also welcome by consumers, marketers will continue to experiment with different formats. One of the current successful ad formats is text-link, which is currently used as the main ad type in Google and Overture. The focus switched from exposures, views, and impressions to clicks when measuring the effectiveness of an advertising campaign.

Google is a good example to examine the online advertising model and its biggest issue, click fraud. Google currently implements an advertising model similar to that of Overture and other search-advertising firms (Exhibit 1). It has several types of entity including advertisers, publishers, online visitors and the search engine firm. Google started out as an excellent search engine with quick and accurate access to information. Google observed there is a high correlation of user interest in information and commercial offers from advertisers about the search keywords. Therefore, search results from Google contain paid ads with short text descriptions and a link to the advertiser’s website. Advertisers pay Google to broadcast these messages to targeted online consumers. To keep the ads relevant, Google only serves ads that meet a pre-defined threshold click-through rate. Both relevance to the search terms and bid price determine the position in search results. To expand its reach, Google also pays publishers, who operate websites with enough traffic, for displaying ads related to their site contents. As more advertisers and publishers participate to this advertising program at Google, Yahoo or Microsoft, this model reveals a major problem.

Case Study

“It did not make any sense. Kevin Steele, co-owner of Karaoke Star, a Phoenix retailer of karaoke equipment, noticed that the number of people clicking on his paid search-engine ads had shot from 200 to 800 a day. But despite the apparent jump in traffic, sales had not budged. Steele and his partner, Diane Frerick, had built their business on Internet advertising, and more clicks almost always meant more revenue which the pair had invested in a new office, more inventory, and a call center to field technical questions. Steele thought he had pay-per-click advertising down to a science. Karaoke Star spent about $2,000 a day on search-engine ads at Google and Overture, a subsidiary of Yahoo focusing on keywords like ‘karaoke’, ‘karaoke player’, ‘karaoke song’ to generate about $6,000 a day in sales. Suddenly, it had to budget the same amount just to get $3,000. With each keyword costing anywhere between 40 cents and $3 a click, Karaoke Star found itself being nickel-and-dimed to death. ‘One day we were doing great,’ says Steele, ‘and the next it was as if someone had turned off the lights.’ ” (Penenberg 29)

Karaoke Star experienced click fraud, a rising problem with online advertising. Click fraud “occurs when people click on paid search ads with no intention of buying anything” from the advertisers (Penenberg 29). This problem has different effects on all the parties involve in the advertising process. Publishers, advertisers, some alleged search firms and certainly the professional fraudsters are all possible sources of click fraud.

Advertisers

Click fraud lowers the Return On Investment (ROI) because artificial inflated number of clicks gives no projection to the revenue generated from these visitors. For Karaoke Star, Steele says the fraudulent clicking has cost him nearly $400,000 over the past two and a half years (Penenberg 29). With little or no protection from click fraud, confidence in the promising future of online advertising will decline and advertisers will spent less on online advertising. Comparing with the low ROI rate from online advertising, other advertising opportunities could become more effective. That was the case with CharterAuction, an advertiser with both Google and Overture. After realizing monthly spending of up to $20,000 on paid clicks did not yield enough sales and losing confidence in the investigation of search firms, CharterAuction now “spends the minimum to keep their accounts open while exploring other avenue such as print ads” (Quiton 14). The majority of Google revenue comes from paid search ads. For the third quarter of 2005, 56% of total revenue was from ads on its own websites and 43% of total revenue was from partner sites through its AdSense program (Mills). With 99% of total revenue from paid search advertising, Google must prove to advertisers that their investments generate reasonable returns. Because click fraud directly affects advertisers, they need to be on the front line in the fight against these illegal activities. One of the detection methods is to implement a click-tracking program on their own system. AdSpeed is a business service provider that uses Google to advertise its products. Because current Google statistics reports (Exhibit 2) do not have enough details, AdSpeed employed a tracking program (Exhibit 3) to understand more about the clicks and visitors from these clicks. The tracker separates clicks into several main sources including “Affiliate”, “Google” and “Overture”. For each source, individual clicks were logged with information about the search keywords, Internet Protocol (IP) address, and the clicker’s country. The “Using” column shows information about the browser and the operating system (OS). Besides these technical details, this tracker also displays the referring pages, from which visitors clicked on the ads. Details from click tracker could help identify potential click fraud and provide the source of fraud like certain IP addresses or countries. With this information, the advertiser could log into Google to block serving ads to these countries or send a complaint to Google for further investigation. Furthermore, this could serve to match records from Google’s reports and dispute any discrepancy between the two reports.

The biggest problem to click tracking programs is that not all data sources are reliable. IP addresses and originating countries can be altered through a web proxy that is in a different country than the fraudster’s location. Browser and OS information could also be manipulated to any text string. The referring page can be blocked or modified easily and is not a requirement in the web request. Even though fraudsters can easily forge individual fields, these fields are useful in a large aggregated report. More and more companies are looking to implement similar tracking systems. This spawned an entirely new niche market for applications and services to monitor click fraud. Upon detection of fraud, advertisers could choose to arrange it with search companies to dispute the charge or sue the other party, whether it is a publisher, an advertiser or a search firm.

On the negative side, there are incentives for advertisers to commit click fraud. When a visitor browses through search results, there are paid ads from an advertiser as well as ads from its competitors. With many advertisers bidding on any keyword, click fraud could take out a competitor by raising advertising expenses beyond their budget. From the case study, Karaoke Star received “an anonymous email tip from someone claiming to be a former employee of Ace Karaoke”, a direct competitor. “Attached to the email, according to Steele, was a video that showed an automated click fraud program employed by Ace Karaoke” (Penenberg 30).

Publishers

The publisher earns more revenue if his or her website generates more clicks. With this incentive, some publishers encourage visitors or hire people to click on paid ads on their sites. Google recently won a $75,000 court judgment against Auctions Expert International LLC for abuse of Google’s AdSense advertising program (“Click Fraud Lawsuit“ A6). “Auction Expert allegedly recruited as many as 50 people to click on online advertising, generating about $50,000 in ad revenue” (Vise, F01). Publishers with the intention to make money using this method will try to find the right balance, safe enough so they do not get caught yet generate the maximum amount of revenue. On the other hand, if search companies were too aggressive in scrutinizing clicks, good publishers would be afraid that some valid clicks are ignored and they could join other ad networks that can generate higher revenue.

Search Engine Firms

All major search companies realize click fraud is a serious problem that needs effective and immediate solutions. Google CFO George Reyes said at an investor conference in December 2004, “I think something has to be done about this really, really quickly, because I think, potentially, it threatens our business model.” (Crawford)

To address this problem, major search firms like Google and Yahoo should utilize both technical and business strategies. The strategies could be defensive, offensive or a combination of both. Defensive strategies look to detect and prevent click frauds while offensive strategies use lawsuits to punish the fraudsters and raise public awareness.

To prevent click frauds with technology, IP addresses should be logged for every click and any concentration of clicks from one single IP address should raise a warning. Information stored in cookies could help identify the individual users from multiple computers accessing the Internet through a proxy or internet gateway. Google Adwords program currently allows advertisers to target their ads to certain countries, languages and even down to certain cities. However, city-level targeting is not available to all countries. This geographic targeting feature has at least two benefits for advertisers. First, they can choose a more relevant market to target. For example, a national insurance company in the U.S. would not want to waste its advertising budget on visitors from other countries than the U.S. Second, this gives advertisers some control to minimize the potential fraud by blocking suspicious locations, those that generate many clicks with only few sales.

Another valuable tool Google and other search firms should add is the detailed record, which contain detailed information about the clicks. This log should be similar to the click trackers with information about the referring pages, IP addresses and other technical data about the visitors. Even though Google privacy policy explicitly states that information can be shared with third parties to prevent frauds (“Google Privacy Policy”), all statistics reports in Google Adwords only contain aggregated non-personal information. At the time of this research, Google and Overture, the two major paid search companies, only have simple counters for clicks, views, positions and costs for the keywords (Exhibit 2). To identify frauds, both search firms and advertisers need information. Search firms most likely have these data internally to analyze the effect and extend of the problem. However, to show advertisers they are putting the effort to fight frauds, Google and other firms should provide additional information to advertisers who have to pay for all the click fees. Even though advertisers could install a click-tracking system on their web servers, it would be much simpler and more effective if the search firms internally support it. One of the reasons against this practice is that Google might afraid to show the referring web pages if these pages contain sensitive or private information, such as email messages in Google’s Gmail.

These techniques are not bulletproof and fraudsters will always try to outsmart the search companies. This problem represents the overall situation of Internet security. Attackers do the damage then the authority develops techniques or regulations to detect and prevent these activities. However, the circle repeats itself with the attackers figuring out a new method to bypass the protection or a loophole with the new regulation.

To calm and raise the confidence of their stakeholders, search firms has to invest more resources into the research of effective methods to identify and prevent click frauds, actively investigate reported cases by advertisers and publishers. Without the commitment and real actions against this problem from search firms, advertisers would lower their spending and consider other marketing channels. Publishers would choose another advertising network where they could earn higher ad revenue. There is an interesting argument by Vise that search engines have little incentive on their own to prevent click frauds. If search companies were too aggressive in blocking clicks, it would ultimately hurt their bottom lines because they charge advertisers for these clicks (Vise F01). As for Google, failing to meet the ever-rising Wall Street expectations might crash Google’s stock. With rising pressure from advertisers, Google, Yahoo and other ad networks need to be more proactive in stopping fraudulent activities. The goal is to protect the confidence in online advertising.

Click Fraudsters

With the intention to defraud and to diversify the risks, some publisher and advertiser hire external contractors to perform these illegitimate activities. There are programs, “hit bots”, designed to imitate visitors surfing from one web page to another and clicking on advertisements. IP address can provide useful information during investigation because it resolves to the Internet Service Provider (ISP) and the location of the user. Dial-up accounts normally have dynamic IP addresses, a different number for every session. In contrast, broadband connections generally have static IP addresses. Dynamic IP address makes it difficult for search engines to detect the fraud. For example, multiple clicks from one single IP address should raise alerts for potential click fraud. However, with IP address changing frequently, there is no concentration of clicks from one IP address. The clicks are masked as from many users instead of from a single one. Cookies are also a useful tool in detecting and preventing click frauds as history of clicks can be stored in a session and traced back to one user. However, with “58% of online visitors deleting cookies manually or using software to delete them” (Marshall 1E), the fraudsters could clear these valuable traces. The low cost of labor in countries like China, Vietnam, creates profitable businesses, “pay to click” programs, with clicks made by real people. As soon as search companies detect and block a method, it is only a matter of time before the fraudsters develop a new technique to bypass the detections.

Conclusions

To diversify and limit the damage of click fraud to the industry, there should be multiple performance metrics and enough information for all parties. There are evolving online advertising approaches that are not vulnerable to click fraud. One of them is Cost-per-Action (CPA), in which the focus shifts from counting clicks to evaluating actions. Actions are flexible and can be defined as registering a new user account, signing up for newsletters, or ultimately, placing an order (Liedtke). Companies like Google, Yahoo and Microsoft will continue to refine their models and offer advertisers and publishers with more diverse and effective options. The future of online advertising contains many uncertain yet exciting opportunities for those in search for a consumer-friendly and financially sound advertising model.

Bibliography

Briggs, Rex; Hollis, Nigel S. “Advertising on the web: Is There Response before Click-through?” Journal Advertising Research 37 (1997): 33-45

Cho, Chang-Hoan; Cheon, Hongsik John. “Why do people avoid advertising on the Internet?” Journal of Advertising 33 (2004): 89-99

“Click fraud lawsuit brings a $75,000 court judgment.” Wall Street Journal (Eastern Edition). 6 Jul. 2005: A6

Crawford, Krysten. CNNMoney. “Google CFO: Fraud a big threat.” 2 Dec. 2004. CNN. 24 Oct. 2005

Drèze, Xavier; Hussherr, François-Xavier. “Internet Advertising: Is anybody watching?” Journal of Interactive Marketing 17 (2003): 8-23

“Google Privacy Policy Highlights.” Google Privacy Center. 14 Oct. 2005. Google Inc. 24 Oct. 2005

Hoffman, Donna L.; Novak, Thomas P. “How to acquire customers on the web.” Harvard Business Review 78 (2000): 179-188

Liedtke, Michael. USAToday.com “Snap.com, ‘cost-per-action,’ coming after Google.” 19 Jul. 2005. The Associated Press. 25 Oct. 2005

Marshall, Matt. “Which way will the cookies crumble?” Mercury News. 18 Apr. 2005: 1E

Mills, Elinor. News.com. “Google shares soar on healthy revenue report.” 20 Oct. 2005. CNET Networks, Inc. 24 Oct. 2005

Penenberg, Adam L. “So many clicks, so few sales.” Inc. Magazine. Aug 2005, Vol. 27 Issue 8: 29-30

Quiton, Brian. “CharterAuction alleges click fraud.” Direct Magazine. June 2005: 14

Riedman, Patricia. “In the beginning.”Advertising Age 71 (2000): 106

Vise, David A. “Clicking to steal.” Washington Post. 7 Apr. 2005: F01

Appendix

Exhibit 1: Current Google’s Online Advertising Model

Exhibit 2: Google AdWords’ Statistics Report – Source: Google Inc.

Exhibit 3: AdSpeed’s Click Tracker – Source: AdSpeed.com