In the Future, We Will Manage Data Using Conversational Synthesis Technology

In the early days, web pages primarily operated as static information providers, delivering a finite set of data to their visitors. Site administrators might update that data set on some sort of schedule, but the process would involve a manual change on their local machine and then transfer of it to the live version through a file transfer protocol (FTP). Other than this kind of scheduled update, the site remained the same. In terms of receiving user data, the initial scheme was just as primitive: hit counters, "mailto" links, basic contact forms, etc. would allow a site administrator to get basic feedback.


Today, however, database driven web sites with content management systems enable a much more flexible and sophisticated delivery of information. Additionally, the use of advanced tracking tools, CRM integration, and analytics tools has become standard procedure for receiving data; among those sources, most sites receive more data then they can reasonably sort through and process. Stop and consider for a moment how important this shift has been- websites are no longer only concerned with their own data, but also with receiving and synthesizing data from other sources. This is significant, and another reason why corporate websites will remain "command central" for marketing. But now that we have loaded ourselves up with so much data, we need a much better way to efficiently sort through all of it and extract the specific answers we need about how our site is performing.

Huge progress has of course been made in this area thanks to tools like Google Analytics, which, in my opinion, is the most valuable application that Google has created so far. Ever since we started using it, we've wanted to mix its data with the tracking data we collect through our CMS, which would allow us to have a central location to build and pull custom reports about our site. I'm certain that this is something that anyone working with analytics, tracking, and CRM data sources would want, too. Google recently made that possible by releasing a developer API for its analytics tool. In fact, we're using this API to build the newest version of our CMS with many new features, including real-time reporting that merges analytics data with our own tracking system. The result is a powerful dashboard that allows site admins to quickly toggle between a big-picture view of their site activity and a more granular view of individual lead activity or keyword performance. In the image below, I'm showing you a sneak preview of how our CMS will deliver a page-specific consolidated overview of tracking and analytics data in real-time. Incidentally, it's also a sneak preview of an upcoming redesign of our site... more on that soon. On the back-end, users will have a detailed dashboard providing the same kind of data site wide. Trust me, this is going to be incredible. (By the way, if you're interested in hearing more about our new CMS, let us know.)




We're not the only ones building a dashboard tool like the one I described—after all, it's not so much a novel idea as it is a need that we all are experiencing simultaneously. Nor am I likely the only one envisioning the next step, what I like to call "conversational synthesis." What if, rather than relying upon a pre-configured dashboard, you could easily query your site using natural language and then save that query as a routine operation? This would be an incredible mashup of intuitive Ubiquity-like functionality and the speed and scope of our own updated dashboard. But more than that, it would be a conversational, and thus more human, approach to reporting and measurement. I see this as an essential and inevitable development since the biggest barrier to good web analytics and data measurement right now is the fact that many people faced with that responsibility are simply squeamish about numbers. Conversational synthesis wouldn't remove the numbers, of course, but it would enable people to engage with their data in depth, regardless of whether they consider themselves "numbers people," by replacing numeric operations with verbal commands. I imagine what this might look like in the screenshot below:



Just like my expectation for browsers, I can envision merging a tool like this with voice recognition software, allowing it to operate on spoken command. Now that would be truly conversational!