This is not a 'how to make money from blogging' post
You'll see I have experimented with some low-key ads at the top and bottom of each page, and have some thoughts about the technology -- I'm not qualified to suggest how to manipulate readers and Google ads in combination to achieve click through revenue (I'm not likely to become rich, or even moderately better off, through ads on this blog).
Instead I believe that the AdSense technology has some applicability to enterprise knowledge management applications, to facilitate the discovery of hidden information within and across enterprise knowledge stores, document libraries, intranet portals and other information resources. Let's call this capability KSense (K is for Knowledge...).
A use case
Imagine the following use case...
I'm sitting in Boston Logan airport, and have finally got my battered old T40 laptop's WiFi to connect. It's 6:30am (I'm not a morning person so I'm sitting close to Starbucks) and I'm doing a quick last minute piece of research around the use of enterprise portals in the federal government. I'll be flying down to see a certain potential customer with offices in a large, unusually shaped building in DC ('I could tell you, but I'd have to kill you' jokes get old fast).
I go to my corporate collaboration and knowledge management site (yes, Vignette my employer actually uses its own software, and its good), login and search for 'federal portal'. It returns me a list of workspaces and documents that meet the criteria: IRS, USPS, NASA...
As I click in to any of these items the system allows me to drill deeper and deeper into the information in that customer workspace. When a trail goes cold I come back up to the search results and traverse a different route.
What I really need is KSense.
Corporate knowledge AdSense
As I click through the corporate knowledge store I would like to see suggestions of related documents and workspaces, much like I see Google ads. These suggestions, KSense ads, would be based on the content of the current page or document I was viewing. Ads would be selected based on the content of other documents and workspaces, as well as their attributes or tags.
KSense ads related to my current view would allow me to follow my research trail by meandering across workspaces, libraries and information resources, gradually refining my context. WIth search I typically drill-down until I hit dead-ends then return to my search results to try again.
Some of the other ideas that drive AdSense could also be applicable. Document and workspace owners could rank the importance of their content for specific search terms, much like Google advertisers do, to promote ideas, or make common information easier to find from a single search.
For example, a product launch for new intranet portal capabilities could be tagged to appear in an ad banner at the top of search results for 'intranet portal', or through KSense relevant ads on a page containing portal information. For specific periods of time users' attention could be drawn to new information they may otherwise not be aware of. This enhances current search and subscription mechanisms significantly, and enables organizations to promote new information and products more effectively with their own employees.
Extending KSense
Although AdSense may present ads most relevant to the text on the page, it takes no account of the actual role of the reader. On this blog (here is an example page addressing BPM) the CIO of a major finserv organization may in fact be interested by the ad:
A free guide to Business Process Management (BPM) solutions.
Despite the fact that the ad is well matched to the context of the page it is unlikely that a more typical reader will be so inclined to click, unless he or she is doing some serious research of BPM vendors.
For example, if AdSense knew that the profile of the user was 'software developer; gamer; Boston, age 26', and that she had just landed at this blog having clicked through from the Scobleizer (yeah, I'm dreaming), the following advertisement may be more likely to earn some click-through revenue:
Render 128 million pixels / sec for incredible Red-Sox gaming realism.
Without using nasty click-tracking and spyware on users' PCs this is not likely to happen in the web-world. In the corporate world its in big-brother's interest to know its users a little better, so KSense could benefit from user profiles and viewing habits.
For example, within Vignette it could be that my profile says 'Solutions Architect; product experience: EDRMS, BPM;recently viewed: federal portal, DoD, intranet'. This should drive a different set of KSense ads for both search and browsing than my boss 'VP; product / solution marketing; management; recently viewed: vacation policy, competitive analysis'.
Summary
Enterprise search tools such as Endeca provide the ability to be presented with human classified categories to further refine search results. This is powerful, but it seems very tightly controlled through human categorization (for an example, see Forrester's view of search). Categories within search results can be used to drill down to the information you are hoping to find.
I would like to see the more freeform capabilities of KSense used in organizations. I think that the context aware and 'advertising' paradigm could help users find more useful information that would be otherwise lost in the cloud of traditional search results, and the typical route where users are forced to drill-down deeper at every click. The ability to meander across workspace and library boundaries through KSense ads seems to open up a lot of corporate knowledge and information.
Technorati tags: google knowledge+management search adsense
1 comment:
Cool Man! I was always interested in developing intelligent technologies. Your dream about Ksense is great. I still have to start working on this technology but I am familiar with google adsense. No wonder, just applying similar technologies in other areas of life will bring awesome results. Good work! Keep it up.
Post a Comment