Difference between revisions of "Projects:Common Answers"
| Line 14: | Line 14: | ||
* [[Common Answers, Sept 1-15]] | * [[Common Answers, Sept 1-15]] | ||
| − | == General Strategy == | + | == Description == |
| − | * Track frequent questions (searches) on [[CommonAnswersQueue]] | + | |
| − | + | Common Answers is designed to organize our content effort under one uniform directive. The idea is to refine our content efforts to both 1) meet demand for information (as it is perceived on the web and on our site individually), and 2) leverage the domain directory resource we've already built (mostly). | |
| + | |||
| + | The project centers on these two areas in the follow respect: | ||
| + | |||
| + | |||
| + | |||
| + | === General Strategy === | ||
| + | * Track frequent questions (searches) on [[CommonAnswersQueue]]; compile a separate list from categories with 10-100 articles | ||
* Provide good wiki pages to answer common searches. Also show a page that tracks which are the most common searches, and how good the answer page for the search is. | * Provide good wiki pages to answer common searches. Also show a page that tracks which are the most common searches, and how good the answer page for the search is. | ||
| + | * Track how our content efforts effect the rankings of individual pages in google | ||
==Tools/questions that will help in doing this project== | ==Tools/questions that will help in doing this project== | ||
| Line 39: | Line 47: | ||
*[[User:Blake Hinckley|Blake Hinckley]] | *[[User:Blake Hinckley|Blake Hinckley]] | ||
| − | ==Concerns== | + | == Concerns / Discussion == |
How to eliminate redundancy and work already done. We will need a bot for this later that will check to see if an article is present (# of bites combined with # of edits?). | How to eliminate redundancy and work already done. We will need a bot for this later that will check to see if an article is present (# of bites combined with # of edits?). | ||
: ''Why should pages not be edited if they have been in the past? It seems like if it's a hot topic there's something to be written about it. '' | : ''Why should pages not be edited if they have been in the past? It seems like if it's a hot topic there's something to be written about it. '' | ||
| Line 61: | Line 69: | ||
* Delve further into the list: go more long-tail. I don't have the next 1000 search engine keywords so I can't say whether the list gets more diverse as we go down and the competition less fierce. Does anyone know where to find this information? | * Delve further into the list: go more long-tail. I don't have the next 1000 search engine keywords so I can't say whether the list gets more diverse as we go down and the competition less fierce. Does anyone know where to find this information? | ||
| − | ==Results of Meeting With Tak and Scott== | + | ===Results of Meeting With Tak and Scott=== |
====The Plan==== | ====The Plan==== | ||
First step is to onboard people in Pakistan and train them to build out basic wiki pages. We will populate the list of basic wiki pages to build by examining the list of common searches (dev card). Pop culture and game portals will be built after we have a larger trained staff. We also brought up the role of page-rank for Google's search-results and questioned if Google looks at how many people have edited a page. | First step is to onboard people in Pakistan and train them to build out basic wiki pages. We will populate the list of basic wiki pages to build by examining the list of common searches (dev card). Pop culture and game portals will be built after we have a larger trained staff. We also brought up the role of page-rank for Google's search-results and questioned if Google looks at how many people have edited a page. | ||
Revision as of 21:56, 29 August 2007
(Ray, Ward, Chris, Scott, Julia, Blake, Asad Butt, Obed Suhail, Tanwir Shah, Asma Khan, Salihawaqar)
Contents
Why are we doing this?
We want to be a visible search result for all search terms.
Done When
There are pages for the top AboutUs searches and the top Google searches
Related Tasks
Description
Common Answers is designed to organize our content effort under one uniform directive. The idea is to refine our content efforts to both 1) meet demand for information (as it is perceived on the web and on our site individually), and 2) leverage the domain directory resource we've already built (mostly).
The project centers on these two areas in the follow respect:
General Strategy
- Track frequent questions (searches) on CommonAnswersQueue; compile a separate list from categories with 10-100 articles
- Provide good wiki pages to answer common searches. Also show a page that tracks which are the most common searches, and how good the answer page for the search is.
- Track how our content efforts effect the rankings of individual pages in google
Tools/questions that will help in doing this project
- Images
- List of categories
- find out whether a redirected page gets into the SE Index as well as a "normally named" page
- subcategorization tool
- category keywordifier tool (Ward to explain)
Generating the Common Answers List By Hand
Common Answers will derive directly from metrics on web queries in the following areas:
- Top queries on AboutUs - Derived from an internal list
- Top Web Queries - Derived from SearchEngineGuide.com
These lists will be updated weekly at CommonAnswersQueue and culled to eliminate redundancy (pages already in existence on the site) and inappropriate content (i.e. adult or illegal). This will be done by hand for the time being. Ideally we will be up to date with all time sensitive searches but also provide full content on "historically" (past three months or so) popular material. Over time our information will only improve. We talked in today's strategy meeting about building out subdomains and reaching out to the blogosphere. We could start with the Technorati Top 100.
Concerns / Discussion
How to eliminate redundancy and work already done. We will need a bot for this later that will check to see if an article is present (# of bites combined with # of edits?).
- Why should pages not be edited if they have been in the past? It seems like if it's a hot topic there's something to be written about it.
We should rethink the name common answers and come up with something more obvious and catchy.
Overview and Issues
After looking through the top 500 search engine keywords of the week, the results are mainly
- Popular Websites: myspace, youtube, yahoo, and ebay in all their spelling variations
- Popular Culture: names of models, actresses, and artists with the occasional song
- Service-related searches such as maps, e-cards, game cheats, etc.
It would be very difficult to compete with service-related searches because those pages are very specialized: people are not going to come to AboutUs to send an ecard or look up song-lyrics. We could provide a directory of good ecard and song-lyrics pages, but these don't seem to be at the top of Google's search-results. The portals are also hard to crack because the first 5 pages of search results for Yahoo are all Yahoo pages, same for MSN.
So where is an opportunity for AboutUs?
In the big companies and the websites, we could create amazing pages about the company. Wikipedia is a top entry for Ebay, YouTube, Mapquest, and Craigslist (4/12 top searches). Part of working our way up google's results is continually updating the page so it appears fresh and adding additional content. It would take a decent amount of work. The other possibility is less noble, but could be lucrative. We could create pop culture portals dedicated to the likes of Britney Spears (27th most popular search), Emma Watson (30th), Lindsay Lohan (9th), Paris Hilton (19th), Jessica Biel (35th). It would certainly be a departure from our more family-friendly portals like Health and Fitness, but it could steer some traffic our way. The advantage of pop-culture portals is that our competition isn't as tough as cracking the front page for Yahoo or MSN. The disadvantage is that pop culture is dynamic by nature so we can assume Paris and Lohan will not be sitting on top of web-traffic once the scandals die down. However, Paris is never far from headlines and even without a scandal these celebs generate a decent amount of traffic.
What Do You Think
I think we have three options for answering the requests posed by the top 500:
- Extremely in-depth company and website pages a la Wikipedia (we would need intense paid content or community work to achieve this, but it is on-message and reflects well on our business as a whole).
- Pop-Culture Portals: dedicated to Artists, Celebs, etc. We would have to move fast to stick with the traffic, and it could feel slimy.
- Delve further into the list: go more long-tail. I don't have the next 1000 search engine keywords so I can't say whether the list gets more diverse as we go down and the competition less fierce. Does anyone know where to find this information?
Results of Meeting With Tak and Scott
The Plan
First step is to onboard people in Pakistan and train them to build out basic wiki pages. We will populate the list of basic wiki pages to build by examining the list of common searches (dev card). Pop culture and game portals will be built after we have a larger trained staff. We also brought up the role of page-rank for Google's search-results and questioned if Google looks at how many people have edited a page. Tak points out that we could be a decent portal for service related searches (eg. e-cards, online games, etc.), by linking to valuable, popular websites.
Little Discussion
Its a brilliant idea to generate traffic and after filtering the searches we can focus on the query requirement, compelling the users to a path where they can find their exact information. We can also create portals related to:
- 'top rated movies' of Hollywood
- Current News, by relating to popular news sites
- Business Updates
- News about Latest Electronic Devices, as now a days people are interested in inquiring about getting new models of mobiles or other electronic stuff.
We may put some funny stuff, some glowing colorful stuff, which in turns interests the user, on answering page which is being created for the query.User:Asma Khan
