Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
SlideShare a Scribd company logo

Configuring share point 2010 just do it

•
•
Marianne Sweeny
Marianne Sweeny

Finding, or not finding, information is consistently the most called out issue in the enterprise. Technology companies spend millions developing features that remain idle because, while everyone is concerned about optimizing enterprise search, no one is doing anything about it. The PM cuts the budget because "the devs will do it." The IA/UX architects do not have the specific expertise. The developers want to do it but do not have appropriate guidance. This is a call-to-action for developers and ITpros to make sure that they get what they need to make search in the enterprise work. Because, after the interactive marketing agency has left the building, they are the ones that will be hearing "search sucks" directed at them.

1 of 27
The inabiliy to find information is consistently the most called out issue within the Enterprise. This is because, most of the time, a very good enterprise search engine has not been configured to live up to its potential. This is a triple pity because: 1. Enterprise search affords the most opportunity for direct intervention to impact performance 2. The user group is readily available and eager to offer feedback and fix the problem 3. Enterprise search engines provide rich user data and other reports to help with refinement and revision 1
2
Xerox UK study: July 2009 http://www.pitchengine.com/xeroxcorporation/xerox- survey-finds-information-overload-a-hinder-to-electronic-health-records-/15485/ People are more frustrated with searching for information than being stuck in traffic – that is what we’re up against in the enterprise. And, search is more painful in the enterprise because the searcher takes longer to give up. They know “it” is out there and will keep looking far longer than they would in a Web search. And, search in the enterprise is more costly because of this additional time and what the searcher does when they cannot find what they are looking for: • Randomize a co-worker for help ($) • Use an older version of the document ($$) • Recreate what they looking for ($$$) • Give up and go without the information that they need ($$) 3
Using the Internet: Skill Related Problems in User Online Behavior; van Deursen & van Dijk; 2009 Adding insult to injury, the Pew Internet Trust did a study where Web searchers believe that they are very good at finding information and trust the search engine implicitly. So, if they cannot find it…it must not be there. They bring these expectations into the workplace and to the enterprise search engine. 4
There are 2 kinds of searches, navigational and informational How we look for information is different between people and between people and machines. Humans are limited by their ignorance. We don’t know what we’re looking for much of the time and so do not know how to find it. We often rely on technology to provide parameters to narrow our scope and put us on the right track. Unfortunately, technology is “face value” and so does not know how to interpret our queries. Does not understand that we can have a single word mean multiple things (order a meal, put things in order) or multiple terms mean the same thing (star: celestial entity, celebrity) 5
This was recently put to the test in the US with an item that caused an uproar. A woman wants to buy designer eyeglasses and save money. She chooses the #3 result on Google. The frames that are delivered are obviously fake. When she returns them for refund, the owner of the business responds with harassment and threats. To the customer, relevant means honest and high quality. To Google, relevant means many links and many, many social media mentions. What the search engine did not understand is that most of the mentions were warnings of bad quality and service. When the story came to light, Google’s response was that they would “tune” their sentiment algorithm. 6
While many of the Web searches are of little consequence to anyone but the searcher, enterprise search failure can have far reaching consequences such as: work stoppage, wasted money or sub-optimal outcomes due to bad information. 7
Microsoft and others have spent millions developing cool features that are evangelized at conferences all over the world. These features are designed to alleviate many of the issues with finding information in the enterprise. They go underutilized because of: skewed priorities, bad estimating or development difficulties (too hard). This is because search optimization is not a priority in the enterprise and it should be. Enterprise SEO should be part of development and not a nice-to-have that is all too often cut from the budget before launch. 8
They work OK but not up to the client needs or expectations. These products are meant to be tailored to the client environment, beyond the collection definitions and beyond the indexing rate. 9
Governance is about who and how, not about why. It is meant to influence people behavior, not system behavior and not how people engage with the system. 10
On its own, not successful because it is hierarchical, based on external factors not related to how your colleagues search for information. 11
A rigid hierarchy is good for stocking shelves, not for organizing digital bits that can (and maybe should) reside in many places or be available from many places Not everyone shares the same meanings as the guy who put it together Useful for facets and filters Search works best when the searcher is able to use the search engine to get a sense of how they should be looking for what they need 12
Tag! You’re it! Even if developers/itpros are not told explicitly that they are responsible for the success of search, they are held accountable after that interactive agency has been paid and left the building. The developers and itpros are the ones who get ambushed in the halls with “search sucks” and who get the calls from those on Mahogany Row. IA/UX: they have visions of dropdown menus dancing in their heads Content Strategy: is busy dreaming of the complex auto-classification software that they will get to magically work PM: likes saving some $$ by having search configuration as SEO…they just didn’t tell you that yet…or ever 13
Tools User research, user research, user research Discovery session to set stakeholder expectations Search-specific user experience design/ia 14
15
Users look to search engines for guidance. We can provide similar guidance with user controls Search as you type: Jquery customization for SP 2007 16
Jared Spool did a site search study some time ago that found users successful 37% of the time when using site search and 50+% of the time when navigating Users don’t like navigation at the outset but will use it if contextual and in a form that they can influence 17
18
Guided Tours: built on analysis of other user pathways and knowledge of corpus Produced Views: page of assembled content items focused on a single subject Task List Drop Downs: “I Want To…” links to pages of assembled content focused on single common task Related Links: related as in “next steps” not what Marketing wants to be a next step Best Bets: editorially assigned result that may not be chosen by the search engine SharePoint Specific: Top Answers, Did You Mean that uses only terms found in the indexed content for query suggestions 19
Humans are the best tool to configure search for other humans because search engines are factoring in human behavior more and more SP2010 has behavior relevance in the form of click-through analysis for relevance ranking and query suggestions that are mined from search logs. There is also an Outlook connector that will mine SENT email for expertise and interest information 20
Many new social features in SP 2010 21
Don’t do it alone.: • Insist on a dedicated IA • Make the time part of the deployment/development budget • Enlist power users as beta testers, evangelists and to report back on performance • Have an executive sponsor • Create measurements for success 22
23
24
25
Zero results are the road to Perdition Tips •Verify indexing by using crawl logs to see what and how sites are being indexed •Verify site collections and lists have proper config (“allow to appear in search results”) •Verify crawler has appropriate perms •Tell content owner to create a BB 26
27

More Related Content

Configuring share point 2010 just do it

  • 1. The inabiliy to find information is consistently the most called out issue within the Enterprise. This is because, most of the time, a very good enterprise search engine has not been configured to live up to its potential. This is a triple pity because: 1. Enterprise search affords the most opportunity for direct intervention to impact performance 2. The user group is readily available and eager to offer feedback and fix the problem 3. Enterprise search engines provide rich user data and other reports to help with refinement and revision 1
  • 2. 2
  • 3. Xerox UK study: July 2009 http://www.pitchengine.com/xeroxcorporation/xerox- survey-finds-information-overload-a-hinder-to-electronic-health-records-/15485/ People are more frustrated with searching for information than being stuck in traffic – that is what we’re up against in the enterprise. And, search is more painful in the enterprise because the searcher takes longer to give up. They know “it” is out there and will keep looking far longer than they would in a Web search. And, search in the enterprise is more costly because of this additional time and what the searcher does when they cannot find what they are looking for: • Randomize a co-worker for help ($) • Use an older version of the document ($$) • Recreate what they looking for ($$$) • Give up and go without the information that they need ($$) 3
  • 4. Using the Internet: Skill Related Problems in User Online Behavior; van Deursen & van Dijk; 2009 Adding insult to injury, the Pew Internet Trust did a study where Web searchers believe that they are very good at finding information and trust the search engine implicitly. So, if they cannot find it…it must not be there. They bring these expectations into the workplace and to the enterprise search engine. 4
  • 5. There are 2 kinds of searches, navigational and informational How we look for information is different between people and between people and machines. Humans are limited by their ignorance. We don’t know what we’re looking for much of the time and so do not know how to find it. We often rely on technology to provide parameters to narrow our scope and put us on the right track. Unfortunately, technology is “face value” and so does not know how to interpret our queries. Does not understand that we can have a single word mean multiple things (order a meal, put things in order) or multiple terms mean the same thing (star: celestial entity, celebrity) 5
  • 6. This was recently put to the test in the US with an item that caused an uproar. A woman wants to buy designer eyeglasses and save money. She chooses the #3 result on Google. The frames that are delivered are obviously fake. When she returns them for refund, the owner of the business responds with harassment and threats. To the customer, relevant means honest and high quality. To Google, relevant means many links and many, many social media mentions. What the search engine did not understand is that most of the mentions were warnings of bad quality and service. When the story came to light, Google’s response was that they would “tune” their sentiment algorithm. 6
  • 7. While many of the Web searches are of little consequence to anyone but the searcher, enterprise search failure can have far reaching consequences such as: work stoppage, wasted money or sub-optimal outcomes due to bad information. 7
  • 8. Microsoft and others have spent millions developing cool features that are evangelized at conferences all over the world. These features are designed to alleviate many of the issues with finding information in the enterprise. They go underutilized because of: skewed priorities, bad estimating or development difficulties (too hard). This is because search optimization is not a priority in the enterprise and it should be. Enterprise SEO should be part of development and not a nice-to-have that is all too often cut from the budget before launch. 8
  • 9. They work OK but not up to the client needs or expectations. These products are meant to be tailored to the client environment, beyond the collection definitions and beyond the indexing rate. 9
  • 10. Governance is about who and how, not about why. It is meant to influence people behavior, not system behavior and not how people engage with the system. 10
  • 11. On its own, not successful because it is hierarchical, based on external factors not related to how your colleagues search for information. 11
  • 12. A rigid hierarchy is good for stocking shelves, not for organizing digital bits that can (and maybe should) reside in many places or be available from many places Not everyone shares the same meanings as the guy who put it together Useful for facets and filters Search works best when the searcher is able to use the search engine to get a sense of how they should be looking for what they need 12
  • 13. Tag! You’re it! Even if developers/itpros are not told explicitly that they are responsible for the success of search, they are held accountable after that interactive agency has been paid and left the building. The developers and itpros are the ones who get ambushed in the halls with “search sucks” and who get the calls from those on Mahogany Row. IA/UX: they have visions of dropdown menus dancing in their heads Content Strategy: is busy dreaming of the complex auto-classification software that they will get to magically work PM: likes saving some $$ by having search configuration as SEO…they just didn’t tell you that yet…or ever 13
  • 14. Tools User research, user research, user research Discovery session to set stakeholder expectations Search-specific user experience design/ia 14
  • 15. 15
  • 16. Users look to search engines for guidance. We can provide similar guidance with user controls Search as you type: Jquery customization for SP 2007 16
  • 17. Jared Spool did a site search study some time ago that found users successful 37% of the time when using site search and 50+% of the time when navigating Users don’t like navigation at the outset but will use it if contextual and in a form that they can influence 17
  • 18. 18
  • 19. Guided Tours: built on analysis of other user pathways and knowledge of corpus Produced Views: page of assembled content items focused on a single subject Task List Drop Downs: “I Want To…” links to pages of assembled content focused on single common task Related Links: related as in “next steps” not what Marketing wants to be a next step Best Bets: editorially assigned result that may not be chosen by the search engine SharePoint Specific: Top Answers, Did You Mean that uses only terms found in the indexed content for query suggestions 19
  • 20. Humans are the best tool to configure search for other humans because search engines are factoring in human behavior more and more SP2010 has behavior relevance in the form of click-through analysis for relevance ranking and query suggestions that are mined from search logs. There is also an Outlook connector that will mine SENT email for expertise and interest information 20
  • 21. Many new social features in SP 2010 21
  • 22. Don’t do it alone.: • Insist on a dedicated IA • Make the time part of the deployment/development budget • Enlist power users as beta testers, evangelists and to report back on performance • Have an executive sponsor • Create measurements for success 22
  • 23. 23
  • 24. 24
  • 25. 25
  • 26. Zero results are the road to Perdition Tips •Verify indexing by using crawl logs to see what and how sites are being indexed •Verify site collections and lists have proper config (“allow to appear in search results”) •Verify crawler has appropriate perms •Tell content owner to create a BB 26
  • 27. 27