Dave Pollard's papers on business innovation & knowledge management

trust your instincts

< £ Salon Bloggers & >

	February 18, 2004

PERSONAL CONTENT MANAGEMENT: AN EXPLORATION In a recent post I advocated almost a complete replacement of existing knowledge management systems and intranets with a three-tiered set of simple, intuitive tools consisting of: Personal content management tools -- to help people organize their personal information (and other information they've aggregated) their way, and identify who they will permit to access it under what circumstances ('permissioning') Metadata tools (invisible to the user) -- to automatically reorganize this personal content for effective, permitted use by others Social networking applications -- to help people identify other people (inside and outside their organization) with particular expertise or shared interests, connect and collaborate with these people and with people in the individual's self-defined networks, via Simple Virtual Presence, browse and subscribe to others' permissioned personal content, and publish their own permissioned content. In my early thinking about this, I proposed a new consulting discipline called Personal Productivity Improvement (PPI) to help individuals, starting with those in the front lines of organizations, make better use of the tools and content on their personal computers. When I spoke to people in several businesses in different industries, they were very enthusiastic about this idea. On giving it further thought, however, I wondered whether PPI was the solution to the wrong problem. If the tools and information on people's PCs and intranets are unduly complex, counter-intuitive, and inappropriate for the key business problems that front-line people need to solve, so that people use other processes (walking down the hall to speak to colleagues), other tools (the public Internet) and other sources of information (the people in their rolodex) instead of the ones supplied by their employer -- doesn't this suggest it's the tools that need 'improving', not the users and the processes they use? I believe personal content management tools are the place to start, because since the earliest days of business, the principal way of sharing information has been peer-to-peer, the most valued 'repositories' of business information have been personal filing cabinets, and the principal schema for organizing work has been the personal desktop. It makes sense, therefore, that tools that facilitate and reflect these well-established 'knowledge processes', information sources and networks should be much more successful than the complex, centralized, hierarchical knowledge management tools and repositories that have been foisted on users for the past decade. I wrote the other day about attempts to replace paper, and about Gladwell's study of why paper and documents have proven so durable and successful even in this electronic age (spatial flexibility, tailorability, browsability). And I believe any schema for personal content management needs to reflect and honour our most established 'information behaviour' -- the shuffling of paper. The founders of a company called Alias Research (now part of Silicon Graphics, but in the process of being spun off again) were powerful advocates of making technology adapt to human behaviour rather than the other way around, and I agree with them 100%. Lowest common denominator, across all job descriptions, levels and industries, are these fundamental 'knowledge worker' behaviours and needs: "Knowledge-work"-in-process management entails the dynamic, three-dimensional shuffling of paper and documents in a workspace, usually a physical desktop. The organization of the workspace is highly personal and varied, and often opaque to anyone else trying to figure out "how X organizes his stuff". People learn, and add value to others' work, through annotation, also a highly-personal and varied process Conversations, overwhelmingly one-on-one and face-to-face, are the principal means by which almost all knowledge work is done. Even research is more highly-valued if it is 'primary' (derived from personal conversations), rather than 'secondary' (derived from library or database searches). Context is critical to most knowledge work. In business conversations I have observed, three times as much time is spent understanding the context for an opinion or fact, as is spent actually understanding or debating the opinion or fact. Knowledge work's ultimate purpose is usually to enable informed decisions. Most meeting time is wasted because the decision has already been made, or because no decision depends on the matters being discussed in the meeting, or because people in the meeting cannot relate what is being discussed to a decision that they have a personal stake in. The process by which most business decisions are made should terrify most stakeholders -- this process is frequently emotional, biased, impulsive and uninformed. The executive's gut instinct, and opinions offered by his/her inner circle (usually arrived at by the same flawed process) both trump objective assessment. Much knowledge work is therefore used only to justify a decision already made subjectively, and contrary evidence presented is usually either discounted or ignored. That's not necessarily a bad thing -- we do expect decision makers to be able to make good judgements based on their experience, and not always have to rely on outside empirical knowledge. So, while we must be sanguine that it's not going to make much impact on how things are done in the corner offices anyway (which explains perhaps why execs I spoke to were not enthusiastic about investing in Personal Productivity Improvement), how would we design a personal content management suite of tools to improve the effectiveness of these knowledge worker behaviours and processes? I'd start by creating a machine-readable analogue of the physical workspace. We need a Workspace Tool that allows us to shuffle virtual documents the same three-dimensional way we shuffle physical ones. That tool should replace the 'arrow' cursor with a 'hand' cursor, like the Acrobat pdf cursor but a lot more flexible. The 'hand' needs to be able to pick up and move a document, and to pick up and read and browse a document, and to be able to clip a document or a piece of a document to another, either temporarily (so the documents could be separated again) or permanently (so they would become a new document), and to be able to place any document anywhere in a stack of documents. The 'hand' needs to be able to put two documents side by side and browse them simultaneously. The tool needs to allow the user to do this on multiple three-dimensional virtual workspaces, that the user can label as they see fit. It must allow the user to make multiple copies of the document, and move or change each copy in different ways. And it must allow the user to send any copy of the document to any number of other people (without opening another 'application') and to 'permission' the document to identify who else can 'subscribe' to it -- the set of people who they will allow entry to this virtual workspace to access it. Such a tool would allow us to capitalize on the economy of 'virtual' space by doing away with the 'filing cabinet' -- that horrible black hole invented by Dewey the librarian into which documents disappear never to be found again, which Windows has tragically copied. Instead, we would 'save' the entire workspace, with its three-dimensional array of documents intact. It would be neatly put away but, if we needed something in that workspace again, we would simply open the entire workspace again, arranged in the way that made sense to us, and instantly find what we were looking for by where it was in the space, not by having to remember what awkward name we gave it. And then on to the next project with a 'clean' new workspace. This tool would need to be indifferent to the document's format -- whether the suffix was .doc or .xls or .ppt or .html or .pdf would be irrelevant. More importantly, e-mail messages and other 'recorded conversations'  would need to be seamlessly accommodated just like any other document. There are some tools today that do limited parts of the above, but in awkward and unintuitive ways. This needs to be as simple as child's-play, and will probably require software designers to start from scratch and throw away all their familiar technological architecture constructs in favour of the human information constructs we have used at least since Gutenburg. The Workspace Tool could eliminate the need for Windows Explorer and similar 'file management' tools on most computers. OK, that's a start on the spatial flexibility and paper-shuffling spec for the tool. Let's go on to annotation. I've seen some limited annotation functionality in a program called FolioViews, that 'labels' each user's notes and/or changes in a publicly-accessible and centrally-controlled document. MS Word has some such functionality in its 'edit mode'. E-mail uses blacklining or indenting to create 'threads' of consecutive commentary. And wikis take it to the next step -- collaboration -- but at the cost of not distinguishing which individuals contributed and changed what, which requires enormous trust. All of these are forms of annotation. But you have to admit they're pretty clumsy. Again, let's look at how it happens in the physical world, and emulate that. For short additions we use the carat and write above the line. We cross out, without eliminating legibility, to indicate deletion. We use the margins, and, if that isn't enough, a separate page with a numbered reference for commentary and longer additions. We may use post-its for the same purpose, or for personal notes pertinent to the document. There are three reasons this is much easier with a pencil and paper than on a laptop. The first is flexibility -- by writing smaller or at an angle we can squeeze a lot of changes into a small area, and we can use graphics as well as text. And we can move stuff around within the document easily. The second is recognizability -- we can tell by the handwriting whose changes are whose. The third is comparability -- we can put two pieces of text side-by-side to compare them or see if they're compatible as we decide what edits or annotations to make. How could we do this in a simple, intuitive way on a laptop? This is much more challenging because of the different native formats of all the documents we annotate. I suspect any intuitive Annotation Tool would need to quietly convert each document to a bitmap in the background. It would also need to pre-set the user's annotation 'voice' -- using some distinctive font, typestyle, textstyle and/or font/background colour to set off the annotations from the rest of the document. It would use the pencil, rather than the hand or arrow, as the cursor symbol. It would need a simple 'insert or comment' functionality that would automatically expand the available space -- exactly at the point of insert -- to contain all that the user wanted to add. That functionality would include a simple freeform drawing tool for graphics. The tool would need a 'mark to delete' functionality that didn't obliterate what was proposed for deletion. It would need a 'replace' functionality that combined the 'insert' and 'mark to delete' functions. It would need a 'highlight' function. It would need a 'move' function. It would ideally need a 'cross-reference' function that would allow the annotator's inserts and comments to dynamically link to another place in the document, or a section of another document. The key again is simplicity and intuitiveness. When the user places the 'pencil' cursor in a space and starts drawing or typing, the tool would automatically interpret this as an 'insert or comment'. Click and drag would first 'highlight', and then if the user started drawing or typing it would be treated as a 'replace', whereas if the user hit the 'delete' key it would 'mark to delete' and if the user then moved the pencil cursor elsewhere in the document and hit the 'insert' key it would leave a numbered flag at the original point and move the highlighted content to the new location. The key sequence 'cf.' could activate the 'cross-reference' function. No menus, no special function keys to remember. In fact, this simple analogue to the pencil could even replace the word processor and html composing tool for all but the most sophisticated document preparation. For what is composition beyond starting with a blank page, and successively inserting, replacing, deleting, moving, annotating and cross-referencing? Even if this Annotation Tool isn't able to interpret and spruce up the hand-drawn graphics into more professional form, as long as it is able to compress the annotated document to a reasonable file size for storage and transmitting to others, its product could become the ubiquitous standard format in which virtually all documents are maintained on our computers. And most important, the Workspace Tool and the Annotation Tool together could obviate the need for most of us to ever print out anything in hard copy. So not only would we save a lot of paper, we'd no longer have to worry about page size, page cutoff or printer compatibility. As I've mentioned before, I think UXGA technology is also essential to getting us to this state, since it allows the user to review, without eyestrain or scrolling, two complete pages side-by-side on the screen. I also think significant productivity improvement will only come when the third 'layer' in the chart above -- social networking applications that allow us to identify relevant contacts, connect to them powerfully, simply and virtually, and share our permissioned content with them -- have been built on top of these newly-improved personal content management applications. Only the three 'layers' of tools working together can enable powerful, context-rich virtual conversations, so that Dr. Nonaka's famous 'virtuous cycle' of knowledge creation (pictured just above right) can finally become a reality. And then, decision-makers will no longer be able to blame awkward and inappropriate technology for being uninformed. 2:58:27 PM  trackback []  comment []

© Copyright 2004 Dave Pollard. Last update: 26/02/2004; 5:58:08 AM.