Welcome to MSDN Blogs Sign in | Join | Help

The Visual Studio Content Survey Results Are In (Part 2)!

In this posting, I’ll show you results from Visual Studio Content Survey for questions related to how customers find Help, whether they give feedback on documentation (and if not, why not!), and general feedback given on the documentation. Note that the survey is still active, and we continue to get additional feedback. I will try and update the results in a few months.

How Help Is Found

Most customers find help through Internet searches, IntelliSense, the index, and local Help search. The table of contents and links to other topics is used the least.

F1  Index

IntelliSense  InternetSearch

LocalSearch TOC

 Links

I don't use the table of contents, because if I need information, I'm already needing to search. Scanning through a table of contents is not going to help me. I don't use F1 for two reasons. Firstly, and most importantly, often I am looking for documentation about a particular method of a particular class. The F1 implementation seems to simply search for the first class it finds with a method of that particular name. I find it faster to go straight to the help index. Secondly, it doesn't always seem to pay attention to my preference - as set in Visual Studio options - to search local help first, online second.

It's much faster to search in the index of the help than to find the topic you need in the table of contents. The TOC is nested really deep at some points and the topic titles aren't always clear.

I am never using the Internet Search because it is simply too slow to search something from the MSDN Document reader.

F1 in general seems to take too long and doesn't go to a relevant page often enough.

Reporting Errors/Omissions in the Documentation

50% of the responses indicated that developers don’t report errors and omissions found in the documentation. For those that do, it’s typically through the feedback link on offline content, or the rating system on online content.

2-Reporting Errors

The main reason for not reporting errors is the time it takes. In Visual Studio 2003, the feedback link in the documentation sent an e-mail to the doc teams so that we could track errors. In Visual Studio 2005, the link took folks to the Connect site where it was time consuming to report an issue. With Visual Studio 2008, we’ve gone back to the e-mail feedback to make the process easier.

2 - Dont Report Errors

 

Regarding the perception that there’s no point in reporting errors because it never gets fixed: We hear you and we actually are (and have been) listening. With the number of writers we have assigned to a documentation set, we tend to focus on writing for the new technologies, but we also want to improve our existing documentation. We recently started a continuous publishing process where we can update our online documentation on a regular basis. We do focus more on high-priority issues, and tend to pay more attention to the issues with the highest volume; but we do look at every one of the comments you send us. And we appreciate your feedback, so please don’t stop submitting issues and suggestions!

Sample Accessibility

We asked whether having Visual Studio samples installed separately from the documentation would make the samples more easily accessible, and soon realized that we didn’t ask this question clearly. By samples we mean the sample applications that you can install and run, not the code examples or snippets that you’d typically find within a topic. Those examples/snippets were not (and will not be) moved outside of the topics. The purpose of moving the samples was to make them accessible to developers who might not have installed the documentation and to enable you to download all of them at once, rather than going to each individual topic to download them.

Areas for Improvements in the Documentation

We asked which content types customers wanted to see improvements in. (After getting feedback that this survey question only allowed users to select one content type, I changed the question to ask for the top content type instead of top 3). Overall, the How Tos and Samples (or did you mean examples?) were the top areas, followed by Reference documentation.

3-Areas A

3-Areas B 

3-Areas C

General Feedback – Top Issues

We received a lot of general feedback about the documentation and Help experience. The top issues were around difficulty in finding the information needed, such as issues with search, F1 and filtering.

4 - General Feedback 

The following quotes describe the top issues reported:

“search is still the most common access. I often find myself being led away from the answer rather than toward it because links tend to diverge rather than converge on some topics. Although this is hard to fight I would like to see some smarter search ability”

“I get lost more than finding what I need. Maybe I just don't know enough to ask the right question. I've been programming in C and Clarion for 20 years and I feel like an idiot with Visual Studio. Takes me forever to do what I can do in 20 minutes with my 'bread & butter' tools.”

“Honestly, I find myself using it less and less for the following 2 main reasons. 1. If I press F1 for context-level help I wait for a long time for the help program to display, while that is loading I can open an IE instance, go to the Google home-page, search the entire internet for the subject I'm interested in and find a discussion on the subject BEFORE the help program has even loaded the results. When the help program is displayed, it is typically very generic and out of context. 2. Generally t he help topic does not include any meat. There's a lot of information above the fold on how to access it, what it returns, but not what it does.”

“It is waaaaaay too slow to open. If it can't be speed up, then at least start opening it (from VS) on a background thread so we can get on with something else while it opens. Also, an option that prompts you to confirm you want to open the help might be nice, since I and several of my colleagues often hit F1 accidentally... then go make a coffee.”

“*********FILTERING DOESN'T WORK!!!!!!!!!!!!!***************It is nearly impossible to track down specific reference documentation even though I know it is there. You try the index or try searching for a simple language keyword or library function and see how many of the listed options apply to the language or the standard libraries. And how many hits do you get for stuff like Active Scripting and ADO.NET etc.? When I want to find a CLR function, I will tell you. When I want to find a standard C++ element, please don't choke me with CLR.”

“I would like to see the documentation only for the language I am using. If I am using one language, seeing the documenation for the same topic for several other languages is distracting, confusing, and wasteful of my time.”

The top issues reported were more about the functionality of the Help system, rather than the content itself. These are problems that we really want to solve and we are looking into ways that we can improve the overall Help experience in future versions . . . stay tuned.

 

--Kathleen

Published Saturday, January 05, 2008 8:57 PM by kmcgrath

Comments

# Geek Lectures - Things geeks should know about » Blog Archive » The Visual Studio Content Survey Results Are In (Part 2)!

# re: The Visual Studio Content Survey Results Are In (Part 2)!

I missed this survey, but I would certainly have had some fairly pithy things to say about both the performance and the thin 'auto-built-from-xml-comments' nature of a lot of the content.

I'd suggest a quick look at the MSDN library from around 1995 - that had notably fast search, and it's been going down-hill ever since.

And the 'F1-blocks-the-editor' bug couldn't possibly take more than 10 minutes to fix - it should be a particular source of shame that it's STILL there in VS2008.  Where's RicoM when we need him?

Saturday, January 05, 2008 5:36 PM by willdean

# re: The Visual Studio Content Survey Results Are In (Part 2)!

sounds like it could be an interesting read, but the pie charts are two small to see - the key is so small i cant work out which colour is what.  can you make them bigger?

Wednesday, January 23, 2008 11:10 AM by Paul Robinson

# re: The Visual Studio Content Survey Results Are In (Part 2)!

Thanks Paul.  I tried to enlarge the graphics as-is but they became distorted. I'll have to take some new screenshots and will update this soon...

--Kathleen

Wednesday, January 23, 2008 2:16 PM by kmcgrath
Anonymous comments are disabled
 
Page view tracker