Thoughts, links, and essays from Peter Merholz
petermescellany   petermemes


Archives before June 13, 2001

RSS Feed

Adaptive Path (my company!)

About peterme

Most of the Time
Oakland, CA

American history around the time of the Revolution, figuring out how to marry top-down task-based information architecture processes with bottom-up document-based ones, finding a good dentist in San Francisco Oakland
Designing the user experience (interaction design, information architecture, user research, etc.), cognitive science, ice cream, films and film theory, girls, commuter bicycling, coffee, travel, theoretical physics for laypeople, single malt scotch, fresh salmon nigiri, hanging out, comics formalism, applied complexity theory, Krispy Kreme donuts.

Click to see where I wander.

Wish list
Show me you love me by
buying me things.

Track updates of this page with Spyonit. Clickee here.

[Editor's note: began as a site of self-published essays, a la Stating The Obvious. This evolved (or devolved) towards link lists and shorter thoughtpieces. These essays are getting a tad old, but have some good ideas.]
Reader Favorites
Interface Design Recommended Reading List
Whose "My" Is It Anyway?
Frames: Information Vs. Application

Interface Design
Web Development
Movie Reviews

2, 4, 6, 8, Let's All Collaborate Posted on 12/14/2002.

Anyone who has worked with others knows that the best collaborative tools are the simplest. It is likely that, in the last 10 years, the most creative thinking and innovation has emerged from sessions gathered 'round whiteboards.

Among the topics discussed at last week's Supernova conference was "collaborative business." In a panel moderated by David Weinberger, a bunch of smart consultant types talked about the use of software tools to support people working together. And the picture was bleak. John Parkinson, from Cap Gemini Ernst and Young, said that after helping spend $1 billion in collaborative software, and developing instruments to study its impact, there was only one tool that had a noticeably positive effect on collaborative productivity: e-mail.

What I'm guessing that means is that all the knowledge management, web conferencing, content management, and lordknowswhatelse simply provides no real return on investment (and is likely a loss, since companies pay millions for these tools, and then ignore them and use email).

This made me consider which collaborative digital tools seem to work? What gets people to coordinate, to work together to a common goal? And my answer: dumb simple ones. Email. Instant messaging. Simple bulletin boards like bugzilla. Voice telephone calls. Weblogs. And, when stepping out of the world of business, SMS.

This definitely fit in with Supernova's theme of "decentralization.". Utility and agency is being pushed to the endpoints, where, with little explicit coordination, people use a suite of simple tools to get things done. One of the things that this discussion made abundantly clear is that the solution to enable collaboration is not really a technical one (much beyond the simple tools). It's a management one.

This will likely frustrate the hell of out big software vendors, who want to develop over-engineered software solutions that require many servers and for which they can charge hundreds of thousands of dollars. Because, frankly, those things don't work. And these same vendors simply aren't interested in developing what does work, email applications and IM applications and web bulletin boards. And I suspect we'll see a lot of enterprise software companies go out of business, not because of a hurting economy, but simply because people realize that you can't automate unstructured collaboration, and that it's a foolish way to spend money.

One of the panelists put it best when he said that in supporting collaboration, what matters most is a "governance architecture." In other words, a system for using the simple tools. With a little bit of training and management, companies could see a huge return on the use of simple collaboration tools, far more than trying to buy some type of monolithic solution that requires everyone to work in exactly the same way, and in a way that runs contrary to how they operate now.

Now, all that goodness said about decentralization, it's important to acknowledge that you ought not go overboard. Which the recognition of the need for governance architecture suggests. As does the important role of some kind of knowledge sharing and management within a company. As a Web guy, one thing I've been exposed to is decentralization in the form of intranets, and let me tell you: it's not really a good thing. Different departments create their own intranets, utterly ignoring what else is out there, using their own terminology, navigation schemes, search tools, etc.

If folks simply lived within their own departments and really never ventured outside of them, this would be fine. But the reality is that in order for employees to accomplish many tasks, they need information or documentation or forms from across the company, and the inconsistency in presentation from department to department creates such hurdles that many folks either a) waste a lot of time hunting for stuff; b) don't bother looking for it in the first place; or c) if they *have* to have it, go to the nearby Nerd and ask them where to find it, never figuring out how to do it on their own.

The obvious summary of all this is that it's important to figure out what to decentralize (tools and opportunistic processes?) and what to centralize (company-wide processes and information classification?), and not to get carried away going too far in any one direction.

Steven Johnson wrote about a related matter with how Apple and Microsoft are diverging in their approach to software development, with Apple pursuing a swiss-army knife approach based on the tasks that people want to perform on various kinds of documents, and Microsoft attempting a one-size-fits-all interface.

(I referred to this approach when I wrote about iPhoto's task-based interface.)

Thinking about collaboration has made me realize I should poke around the literature of Computer-Supported Cooperative Work, a sub-field of HCI with quite a history and pedigree.

These Swiss Army Knife notions makes me think back to what *was* going to be The Next Big Thing before the Web became The Next Big Thing -- document-centered computing. Apple developed OpenDoc, Microsoft built OLE (Object Linking and Embedding), but these never took off.

But I wonder, with email, IM, SMS, and even to some extent, weblogs, we'll see a "communication-centered computing" platform emerge.

12 comments so far. Add a comment.

Previous entry: "Is There A UI Generation Gap?"
Next entry: "Bid on Some Things."


Bugzilla wouldn't be half as valuable without its email notifications.
Posted by Andy @ 12/14/2002 05:32 PM PST [link to this comment]

I once heard Natalie Jerimijenko say "people do dumb things with smart objects, and smart things with dumb ones." She was talking about stuff like smart refrigerators, but same idea.

I wonder what the characteristics of "dumb simple ones" like email, SMS, simple boards, and blogs (maybe) are? An extremely simple mental model is certainly one.

Being mostly text seems to be another characteristic, and specifically the "direct manipulation" of text (email is "type and send", SMS is "type like you talk"). As soon as collaboration (or using the intranet) becomes filling out forms, deciding on publishing privledges, you aren't really working with your text anymore.
Posted by Andrew @ 12/14/2002 09:08 PM PST [link to this comment]

This speaks to a frustration that I have had with the whole field Knowledge Management of late -- it has veered away from using sociological and anthropological methods to look at how knowledge is shared and used in organizations, and become all about hardware and software instead.

The whole field has a "if we build it, they will come" blind spot. Even the simplest technology acceptance models recognize that both perceived ease of use AND perceived usefulness are required for technology adoption. As far as I can tell, most KM solutions have neither of those qualities.

Itís the same as the old example of seeing where people walk, and then putting the sidewalks there (is that from the Design of Everyday Things? I canít remember.). If you are asking people to step off the path to use collaboration tools, it wonít happen. They have to be right on the path people use to do their jobs, or better yet, be part of the path.
Posted by juukii @ 12/16/2002 08:29 AM PST [link to this comment]

Did you see First Monday this month? May be interested in
Posted by Stuart Henshall @ 12/16/2002 03:49 PM PST [link to this comment]

Nice comment about CSCW. In fact, I'd highly recommend getting involved in the community. I went to the CSCW 2002 thang in NOLA last month. Chose it as a boondoggle (NOLA's one of the two best cities in all of North America ;) ) and was pleasantly surprised by the depth of presentations, papers, and posters. John Carroll (of scenarios fame) even said to me, "I think CSCW is where the real interesting HCI things are going" (or words to that effect).

Sometimes HCI _seems_ to focus on individuals working only on individual task completion, success, and affinity. Yet CSCW _seems_ to expand beyond that focus and instead zone in on how to better the interaction _among individuals._ Hmmmm....something to watch.
Posted by Joe @ 12/18/2002 06:38 AM PST [link to this comment]

A little more than two years ago, I started using "Groove" to support the accounting classes that I teach at CalState - San Bernardino. I now use Groove to teach online accounting classes, plus I use Groove to deliver an online CPA Review course that I developed in conjunction with Dr. Irvin Gleim (Gleim Publications).

I have used Blackboard to teach online courses and am familiar with other LMS system. Based on my experience with both programs, I prefer Groove to Blackboard or similar programs.

For online teaching purposes, I use a simple formula (i.e., Groove for collaboration and course management + a course website to store/deliver content). Groove and the website interact with each other. Simply stated, the approach works and works quite well.

With the development of an ASP service called "PopG" by Andy Swarbrick (Oxford, England), Groove works with any operating system. In many respects, PopG is to Groove what "Hotmail" is to email. It makes Groove readily accessible from just about any computer.

Throughout the past year I have followed the discussions about LMS and LCMS systems. I think your comments about people getting caught up in the "software play" are right on point.

In addition to university teaching, I have an extensive background in training and development. I understand the issues and pressures that drive training functions. It is easy to see how LMS/LCMS systems, as they have been developed, have become a means of empowerment.

I like to keep things simple. That is why I have opted to use a variety of programs to create and deliver online materials and to manage the distance-learning process.

Groove is an incredibly flexible, powerful collaboration tool that can be tailored to the needs of my students. It is inexpensive. My students use the "Preview Edition" for "free." That's a price that's hard to beat!

Groove enables me to deliver my form of "blended" distance learning. My accounting students like collaborating with each other and are carrying what they have learned to do with Groove into other classes. That is a most interesting outcome that I had hoped to see.

If you're not familiar with Groove, check it out at From what I have heard, v2.5 is supposed to be available by mid-January 2003.

If you would like to chat about Groove, I'll be happy to share my experiences with the program and the approaches that I'm taking to distance learning. Email me.

Best wishes for the holidays,

Rick Lillie, MAS, Ed.D., CPA
CSUSB - Palm Desert Campus
Posted by Rick Lillie @ 12/26/2002 04:25 PM PST [link to this comment]

The link to Groove in the previous message is correct, but comes up with a "Page Cannot be Found" error. Somehow the "period," at the end of the link got included in the link.

Try this hyperlink:

This works.


Rick Lillie
Posted by Rick Lillie @ 12/26/2002 04:32 PM PST [link to this comment]

I currently use a Wiki system to organise all my ideas and appointments. Before that I was using Entourage, a Mac OS X Outlook. Wakka is simpler, and therefore I use it more.

I host my site using PHP-Nuke, an enhanced weblogging program. But I spend most of my time kicking about on my mates "dumb forum".

I use lots more simple software these days. Unix is based on this - Windows has Outlook, Unix has sendmail, fetchmail and procmail. Each is a simple program, but does precisely what it needs to do - and it does it more reliably than Outlook does.
Posted by Tom Morris @ 12/29/2002 01:54 PM PST [link to this comment]

I second that WikiEmotion!

I hope that OSAF/Chandler simpler yet more flexible alternative to Outlook.
Posted by BillSeitz @ 01/03/2003 11:32 AM PST [link to this comment]

Peter points out that many of the X-info management monikers could be money wasted as these tools are not used and individuals return to using that which is familiar to them (e-mail, etc.). This seems to tie two problems together: 1) Individuals will often use what is familiar to them and not use new tools; 2) Information is created in various formats initially and stored in decentralized containers. The creators of information are individuals, each with their own preferences of applications for information creation (see There is no one or handful of successful solutions that can be tossed at the problem of knowledge or content caputre and information reuse.

The location of information creation is also a problem is X-info management tools. Information creation is a disperse task that can be done at an infinate stationary locations or even from mobile devices. The tools and location variations do cause problems when trying to access relevant information to increase one's own or a group's knowledge. This is the nut that is difficult to crack.

I do believe that a common information structure is one of the keys that will help sovle the data-to-information-to-knowledge transformation. The information creation and augmentation of existing information (comments in a blog or annotations to business report) aggregation is assisted by structuring the information itself. This is not the addition of metadata, but setting titles, subtitles, cites, links, and other information in a standard markup that can be easily parsed and accessed by any digital passer by. Making this easy to use or done by fiat rather than asking the user to set the information could be a route that would parse information and ask the user to verify what has been selected.

I have daily exposure to an enterprise that does not structure their information, but relies on the information reuse for many purposes. The lack of any structure is more helpful than incorrect structure or information stored in a proprietary tool that does not support structure. On the rare occations where inforamation is properly restructured the task of the reuse is very easy. Once information is structured crosswalks between terminology become more readily appearant.

This process seems to be best supported by dumb information stores with properly structured information and smart applications finding patterns and relevance, or tracking human patterns and devining relevance through this process. Proprietary tools wrapping information only inhibit the access and reuse of information. HTML and XHTML have been solid additions to providing ease of information reuse. The problem is the applications that many folks use to create their HTML/XHTML documents often does not assist with providing the proper structure to the information.

There are many good questions and we are just begining to find the path to the right answers.
Posted by vanderwal @ 01/05/2003 10:12 AM PST [link to this comment]


Obviouslly I agree with you. I hope you'll write more on this, I'm running into enough people starting to think the same thing and I really think 2003 is going to be the year we see some real interest in using weblogs within the corporate setting.
Posted by john @ 01/07/2003 07:31 PM PST [link to this comment]

I tried looking at Groove, but gave up very quickly.

I may have forgotten any early pain I had with wikis, but after a very short time, the wiki idea is obvious and easy to use. (First cut: Anyone can edit the (text) content on any web page.)

IMHO, TWiki is the best wiki (I reached this conclusion after testing many of the wikis I could find about two years ago) -- TWiki includes:

* Graphic content (IIRC, .jpg, .png, several others -- maybe anything understood by a web browser?)
* An online interactive drawing tool (TWikiDraw, a plugin)
* Content revision control via RCS (thus generally preventing TWiki from forgetting)
* The ability to limit editing to selected people via various protection schemes
* Simple markup, and none at all for simple things (these paragraphs would show up as bullets on TWiki) -- plain paragraphs are written just by typing the text and separating paragraphs with a blank line.
* Tools for organizing text, including: automatic ToCs based on headings, "structured data" using what they call "Forms" (IIRC), and multiple webs.
* Built in search engine (not the greatest) and able to use external and public search engines as well -- the content of "my" TWiki, WikiLearn, is indexed by many public search engines, including Google.
* GPL'd and written in Perl so others can improve it
* An active community of developers

(Second cut: Anyone, or a limited subset of people, can edit text, graphics, attachments, etc.)

Back to Groove -- I guess I see the signs of what I consider marketing here -- a lot of words about what Groove can do for you -- no simple explanation of what it is (a collaborative web authoring tool, a collaborative system for managing files, a ????, some combination of the above) -- with a wiki, WYSIWYG, in more ways than one.


Randy Kramer
Posted by Randy Kramer @ 01/09/2003 11:15 AM PST [link to this comment]

Add A New Comment:


E-Mail (optional)

Homepage (optional)

Comments Now with a bigger box for text entry! Whee!

All contents of are © 1998 - 2002 Peter Merholz.