A Matter Of Degree

How Reliable is Your SharePoint Implementation? Eight ideas to make a SharePoint-based intranet or KM system more reliable.

Reliable, originally uploaded by Phill Price.

Last night I attended a presentation by Simon Trussler and Anna Gilpatric from Health Dialog on the subject "Getting the most KM bang-for-your-(Information Technology) Bucks. They discussed and demo-ed their use of out-of-box SharePoint as a knowledge management system ("Health Dialog University"), and true to their presentation's title they have really leveraged MOSS 2007 for their needs without investment in third-party tools or custom code.

From their talk and screenshots, I saw quite a few reasons for their KM system's success, mostly based on human time and effort:

• a great-looking interface with graphical elements and usability emphasis
• a dedicated knowledge broker resource to populate the system and review metadata
• active promotion and training
• use of metrics, tracking, and KPIs
• development of workflows and content types to automate processes and make file types more consistent

Trussler summed up the presentation with his three keys to success: promotion, reliability, and usability. I wanted to focus on the "reliability" factor because I feel it's not typically emphasized for the end-user experience of SharePoint implementations as often as the other two concepts.

Some of these points will overlap with usability, but I think it's useful to look at them in the light of reliability - can your end-users depend on the system, and know they will get the same result each time they use it?

1. Are your files and metadata consistent and standardized?  Health Dialog University's dedicated knowledge broker reviews the metadata of every document subitted to the knowledge library.  Pick lists help standardize the choices, and content types are used, in Trussler's words,"to simplify data entry and hide complexity from the user."  There is continuous quality control and quality review of the material.

2. Is your material up-to-date?  When files are uploaded to the HDU system, they are automatically assigned an expiration date which lets the knowledge broker know when they need to be reviewed.  Different kinds of material have different expiration dates - e.g. quarterly vs. yearly.

3. Can users find what they need?   Although SharePoint's out-of-box search capability can be used to find material in the system, HDU also offers a browse-based navigation based on document metadata, in addition to featuring different views of the knowledge library.  Some of these topics and views are a direct response to user requests.

4. Do users receive timely feedback?  When end-users submit a document to the HDU system's pending area, they receive an immediate feedback email (generated by a SharePoint Designer workflow), and are also notified when their document has been approved and moved to the knowledge library, with a path to the new location.

5. How is the response time for approvals?  For HDU, the knowledge manager relies on a system of workflows and alerts, and can act quickly to review and approve pending documents.

6. Do links take the user where they expect to go?  Beyond the issue of broken links, do links behave in a consistent manner (e.g. either all links open in a new window, or none do), and do users know what they're going to get when they click on a file (i.e. will they know they're opening a Word document, a PDF, or a web page)?  (This one's firmly in the "usability" realm but a bad experience could cause the user to feel the system is less reliable.)

7. Where can the end-user turn for help?  On many of the HDU pages, Gilpatric has included a graphical link to a brief training screencast on how to use the features on that page.  This seems to be a really effective way to provide more specific on-the-spot training than the generic "help" link that SharePoint provides out-of-box.

8. How is the overall performance of the system?  Entire books have been written on this subject, and there are many factors to consider which are out of the scope of this post, but Trussler and Gilpatric mentioned that their KM system has experienced no crashes and very little, if any, downtime, and this certainly contributes to a feeling of trust for the end-user.

Some additional reading on SharePoint reliability (There are thousands of pages of material out there on this subject - please drop me a comment if you know a particularly good resource!):

10 suggestions for a reliable, low-latency Search deployment

Microsoft Best Practices Analyzer for Windows SharePoint Services 3.0 and the 2007 Microsoft Office System

 

Six (additional) Worst Practices for SharePoint

This is broken, originally uploaded by Lodigs.

I recently saw that Ben Curry (of Mindsharp and the SharePoint Best Practices Conference) has a list of the Top 10 Administrative Bad Practices for SharePoint. I definitely agree with his choices.

The SharePoint Practice team at my company has been keeping our own list of Worst Practices for SharePoint, and I'd like to take this opportunity to expand on Ben's list and add a few that we have seen over the years.

Please note - this appears to be a list of recommendations; IT IS NOT.

1. Give all your power users and content owners Full Control.
When you can't decide which level of rights to give to your content owners (for example, those individuals who will need to create new metadata categories and build new lists and libraries as well as contribute content to a site), you can be confident that Full Control will give them all the rights they need, but there is a better way. Manage Hierarchy is often an acceptable level of security, which will let those power users do more than contribute, but not do irreversible damage to the site.

2. Don't use metadata to categorize documents or databases; just create a new folder, list or library every time you need a category.
While there are often reasons to create new lists and libraries to segregate content, and there can even be good reasons to use folders within a list or library, this shouldn't be your default content management model. This Worst Practice is related to Number Three:

3. Don't use metadata to categorize documents; just put all the information in the filename field.
Companies with document naming conventions are often used to long filenames that give a lot of information about the document. By using metadata instead, SharePoint (and similar content management systems) can free you from the need to keep all the information in the filename, and provide greater power for filtering and finding documents.

4. Use your old legacy servers for your new SharePoint implementation.
You're not sure if SharePoint is the right product for you, so you'd like to give it a test-drive without making a big investment in hardware.  Unfortunately, what we've seen many times is that running SharePoint on underpowered servers and forcing it to share space with other legacy applications may not give you an accurate test-drive of the product, and in fact could cost a significant amount in terms of time spent troubleshooting.

5. Don't set up any test user accounts.
When you're building out SharePoint, you're most likely logged in with the powerful permissions of a farm administrator, site collection administrator, or site collection owner with Full Control.  As you prepare for launch, moving around the site and testing functionality, you can see and do everything you need to.  Your end-users with Read and/or Contribute rights may have a very different experience - and you don't want to hear this from them on the day of launch.  Create, at the very least, one end-user test account, so that you can log in and out as needed to test your site.  Ideally you would create one for every location, department, or high-level user role, especially if you're using features like Audience Targeting, so you can really see what your different user groups see.

6. Use a folder or document library called “Old.”
Like a lighter held aloft at a concert, this practice keeps the spirit of the file share alive in your new SharePoint implementation.  Folders with names like "Old," "DO NOT USE," "Don't Need," and "Blurry Photos" can be a comforting structure for the end-user who's missing that big old P drive on the network.  In a system where content can easily be surfaced by Search rather than by browsing, however, it may not be apparent to the searcher that the likely-looking filename they've turned up is sitting in the "_delete" folder. 

For your long-weekend enjoyment, check out some more Worst Practices in the physical world at the "There I Fixed It" site.

How the Apollo 11 moonwalk became a tagging and records retention tale of woe

Stepping Onto the Moon's Surface, originally uploaded by NASAImageoftheday.

Today is the 40th anniversary of the Apollo 11 moonwalk, and this morning's NPR story "Houston, We Erased The Apollo 11 Tapes" is a classic tale of tagging, search, and records retention policies (or lack thereof). The article and broadcast are really worth a read/listen, but to sum up:

The original recording of the moonwalk was not in a format that was compatible with television, so the recording was converted for the TV broadcast in 1969 - and degraded in the process. Engineers knew there was a better-quality recording from the camera on the moon, but for years they couldn't find it. They had to search a vast repository of content using much-used terms like "Apollo" and "tape." It turns out that the recording was most likely recorded-over due to NASA's increased need for magnetic tapes in the 1980s.

This scenario makes me think of my clients - many of whom have expressed to me, in the process of converting to a SharePoint implementation, the notion that "any content over five (or three, or six, etc.) years old is worthless." I understand the concept - policy and regulation changes, business and market trends, and fresh research and information all change the relevance of older content. But as the NASA story shows, you may not realize now what's going to be gold in 40 years.

If you're managing content (and who isn't, whether on an organizational or individual level?), do you have record retention policies in place to expire content after a certain time, and if so, are you notified to check the content first, or is it automatically deleted? Are you at risk of having files erased due to exigencies like lack of storage space? Are files backed up on a regular basis, and if they are, could you restore those backups successfully? Have "delete" permissions been given to those who may not fully understand the importance of the content? In your search results, are the "gold" files getting buried under layers of more recent sediment?

NASA invested time and resources to restore and improve the quality of the recordings they did have, since they couldn't access the original. While you're pondering these questions, you can see the video here. (Personally, I'm a fan of the grainier original look; by removing noise they've flattened out some of the detail, a metaphor to explore another time...)

And a nod to you conspiracy theorists out there - pretty convenient that they "lost" those tapes, eh?

Issues with Manage Content and Structure Move / Copy functions in MOSS 2007

I like the convenience of using Manage Content and Structure to move batches of files from one document library to another, while maintaining their metadata (custom fields as well as metadata like "modified by" are brought over from the source library to the target library).

However, I have been seeing some strange and concerning results, mostly having to do with content types.  In a recent example, I had a library of documents with a yes/no checkbox to allow for displaying certain of the documents in a view on the home page.  Another library needed to be merged in with this one, which had the same yes/no checkbox (same column name).  I recommended that my client move the files from one library to another using Manage Content and Structure.  After she did this, she reported that the documents from the merged library were not apperaing on the home page view even though she had checked "Yes."

When I looked more closely at the document library, I saw that there were now two identical fields for the yes/no choice in the same document library.  SharePoint had copied one over and kept the original - and the values were different in each field.  (In the field properties, I saw that the system had appended "0" to one of the column names on the back-end.)

In this case, it was a simple matter of making sure all the data matched from one column to another, and then deleting the duplicate column.  But in some cases (e.g., when the Move function added the "Publishing Start Date" and "Publishing End Date" fields to a document library) I've had to switch on "Allow management of content types" and hide the extra fields that were added during the move process so that my end-users wouldn't be confused by the newly-added fields.

Consider this a word of warning - Manage Content and Structure may beat Explorer View in terms of convenience and speed when moving batches of files, but it can cause unintended outcomes that could take you more time to troubleshoot and fix in the long run.

You know you're living in an interesting society when...

...you type "modifying" into the Google search field and the third most popular result that appears on the drop-down of query suggestions is "Modifying nerf guns."

XML Mapping of Document Properties with Excel and SharePoint - Not Simple or Intuitive

Last May I blogged about XML mapping with Word and SharePoint - i.e., inserting document properties into your Word Document using content types.  Today one of my clients asked about doing this with an Excel template, and after much investigating the Excel ribbon and Googling the issue, we discovered that you can't do it in the simple, out-of-box way that Word provides.  Apparently it can only be done with VBA (Thanks, Douglas M. Smith at BrainBench!).  I'm a little surprised that the MS Excel team's blog does not discuss this, and that the MS Word team's blog on the XML mapping subject does not mention that this is not the same for Excel.

As usual, what turns up when researching this issue is lots of people asking the question, and no answers.

Luckily, my client is a developer.  She found this solution (thanks, Chip Pearson at Pearson Software Consulting, LLC!) with downloadable code which let her access a property and set it in the spreadsheet.

The code needed modifications to get it to work and suit my client's needs.  I won't detail the entire solution here, but a few notes:

Chip's code handles the Office document properties, not the Server properties (i.e. SharePoint).  The property you need to use to access the server properties is WB.ContentTypeProperties.

The Excel template must be saved as .xltm (a macro-enabled Excel template) rather than xltx.  We stored it in the Forms folder of the SharePoint document library, and changed the default template URL in the Advanced settings for the document library.

The upshot is:  it can be done, but you'll need a developer if you're not one already.

xobni - five minutes and I was hooked

I learned about xobni from a colleague last week, and downloaded it today.  Xobni ('inbox" backwards) bills itself as "the Outlook plug-in that saves you time finding email conversations, contacts and attachments" and / or "a more socially aware email environment."  I would prefer to describe it as "the free Outlook tool with addictive metrics."  Once it started analyzing, showing me, for example, who among my contacts is quickest to respond to me, I was fascinated.  Then when it gave me the option to send said contacts their stats, I was hooked. 

The design is bright and appealing (maybe a little too appealing - compared to the bland light-blue-and-white Outlook color scheme its orange, black and purple tones make it difficult to look away or ignore the pane) and I found it user-friendly and intuitive.  After eight hours I'm thinking of my correspondents in terms of their xobni rank ("Yeah, he's my number three") and trying to figure out imbalances ("Why am I sending her two emails for every one she sends me?").  It interfaces with LinkedIn.  It bubbles-up attachments.  It threads conversations.  It hasn't broken anything on my pc, although the comments on this post about xobni's business model indicate that it may cause problems with FireFox.  For my setup, there is no downside and a tremendous upside.

I would rave more, but I've got to go reply to an email from my #1.

Tag as you search?

Today I attended a seminar hosted by the Boston KM Forum on categorization and tagging.  One of the speakers, David Hobbie, brought up a great point - that tagging / categorizing in the enterprise happens when a document is saved into the content management system, when a better way would be to have people tagging documents as they find them in search results.  The first scenario – tag on upload – involves questions such as Which version do you tag?  How can other users tag and rate?  The second scenario – tag on search – accounts for these issues and will lead to richer search results in the future.  I’m not advocating doing away with tag-on-upload, but I’d like to see Hobbie’s idea implemented in the enterprise applications used by my clients.

 

Last night I attended a talk on Enterprise Search from Microsoft by Tara Seppa at the New England SharePoint User Group.   Microsoft emphasizes the value of SharePoint’s “actionable search results” (by which they mean “preview helpful summary information and clear graphical representations of files; move, delete, copy, and drag & drop files; send, forward and reply to messages directly from search results, and open and run applications from the results”). The tag-as-you-search concept would be great functionality to add to future releases of SharePoint Search.

David’s blog:

http://caselines.blogspot.com/

New England SharePoint User Group:

http://www.clearwaypartners.com/SUGHome.htm

KM forum topic:

http://kmforum.org/blog/?p=70

Actionable search results definition from the whitepaper Microsoft’s Approach to Enterprise Search:  Bridging the Gap between Information Management and Enterprise Search

How to insert SharePoint metadata into your Word document using Quick Parts and Content Types

I just learned this from a colleague who learned it at a SharePoint Developer class: When you create a custom Content Type for a document template, and add custom metadata fields to it, you can insert the metadata values into your Word document as Quick Parts.

Several elements need to be in place to make this work smoothly. In my example:

1. I created a Content Type for a NonDisclosure Agreement (NDA)
2. I created a Site Column for Opportunity (i.e., the client name)
3. I added the site column to the NDA content type
4. I added the content type to one of the document libraries on my SharePoint site
5. I created a new NDA from the "New" menu on the document library and
6. IMPORTANT! I saved the document back into that document library. (The Properties Quick Part will appear grayed-out until the document is saved.)

Then, with the document still open in Word, I was able to choose Insert -> Quick Parts from the ribbon, and my "Opportunity" field is available.

If the metadata field is already filled in for the document, you will see the actual data appear in your document (just as when you add the a Date field and today's date appears).

If the metadata field is empty, you can type the information into the document field, and you will see it will appear in the Properties (Server) window in the ribbon..

Lots of potential for bringing your paper forms online and ensuring better metadata compliance!

Here's a nice post from the Microsoft Word team on the subject – they refer to it as XML Mapping.

Update 9/8/08:I am not aware of a way to insert these document properties into an Excel spreadsheet in a similar manner.

Your metadata on Library of Congress photos

Parris Island, S.C., barrage balloon (LOC), originally uploaded by The Library of Congress.

Flickr has partnered with the Library of Congress to allow everyone to tag, comment, and add notes on a portion of photos from the LOC's collections.

Read about The Commons project here.

I first heard about this from my colleague Mike, who read about it on Everything is Miscellaneous.