
Late last week I met with Chris Yeh, VP of Marketing at Tacit Software. I have taken a watching brief on Tacit since late 1999, but its work is of increasing interest to me because of how it supports the seventh pillar in my 7 Pillars model, that of broadening the network of people who can contribute to a team through the automatic discovery of collaboration opportunities. Late last month, Tacit released Version 3 of ActiveNet, its platform for collaboration auto-discovery. Chris kindly gave me some of the background to ActiveNet 3, and confirmed some of the perceptions of Tacit that I've developed during the previous five years.
Update on Tacit
The fundamental idea of Tacit's platform is that there is a high likelihood that people are most expert in the things that they frequently write about and read about. By collating a map of these expertise areas across a large body of people, be they in one organization or even in all government agencies in a specific jurisdiction, there are increasing opportunities for connecting people who need expertise with people who have expertise. Tacit's approach builds on and complements more structured approaches to classifying and categorizing "who knows what" in document databases and best practices repositories, but because of its ability to look across a plethora of applications in real-time, it offers an up-to-the-minute map of who is the best person to speak with.
With earlier versions of ActiveNet, it was my perception that Tacit offered a "broad-based platform" that early adopter organizations could apply to specific functional business challenges. In other words, that organizations could build collaboration auto-discovery applications of ActiveNet to solve their specific problems. Tacit has done something different with ActiveNet 3, in that it has gone to market with three packaged offerings that specifically address problems that global companies all face. In this way, Tacit has made the attempt to put a compelling reason to acquire its platform front-and-center on all discussions, by focusing on the value that it offers within well-known areas. The three application areas for ActiveNet are:
- Pre-Procurement. To help people in organizations discover other people who have previously dealt with a specific procurement challenge, or have prior experience with a vendor. This helps the organization to stop buying poorly performing products from vendors, or alternatively, to pool purchases across different divisions for greater bargaining power. See the Tacit Solution for Strategic Pre-Procurement for more information.
- Research and Development. To help researchers located in different places to find others who share common or overlapping interests, who have previously dealt with a specific problem, or who could make an introduction to a key person at another organization. See the Tacit Solution for Research & Development for more information.
- Business Development. To enable people who are thinking about expanding into a new market or establishing a partnership with a new external party to discover what is currently known by others in the organization. If you forced me to rank the three application areas in order of impact, I'd put this one last on the list, but that may be a faulty conclusion. See the Tacit Solution for Business Development for more information.
Of course, nothing in ActiveNet 3 prevents organizations from using its services and capabilities across other application areas.
The Tacit platform relies on a two-stage process. In stage one, the Tacit platform discovers blocks of text from multiple applications that have an associated author. Noun phrases within those blocks of text are associated with the name of the author. In stage two, there is a continual revision of what Chris called Tacit's "confidence index", that is how confident the Tacit platform is that a specific person is an expert on a certain noun phrase. This is done by looking at how frequently over time a certain person uses that noun phrase in their communications. The index is a dynamic construct (since it is being constantly revised), done for the purpose of being able to state authoritatively at the current point in time that a given person is the right one to talk with.
Some Observations
Here's my reflections based on my discussion with Chris:
- Tacit sources text blocks and associated authorship from email, Lotus Notes databases, document management systems, file servers, and even Groove Virtual Office. With an increasing number of organizations embracing or trying out enterprise instant messaging, archives of IM transcripts will become an increasingly important rich source of noun phrases and authorship for the Tacit platform. I expect to see some formal announcements from Tacit with IMlogic, Face Time and Akonix in the coming months.
- There is value for a person in being proactively notified that noun phrases they are writing frequently have already been discussed by others in the organization. That is, that the individual doesn't have to initiate a query of ActiveNet to discover connections to experts, because ActiveNet proactively alerts them via on-screen alerts, for example. I raised this idea with Chris, and he signalled that Tacit was working on some form of this idea. It is currently possible to subscribe to certain phrases in the ActiveNet platform.
- One of the algorithms within Tacit looks for the "three most expert people" in the organization on a specific topic, and then emails the enquiry to them. In their good time, the identified expert can respond to the request. It wouldn't surprise me if an additional option was inserted into the platform, for the "three most expert people who are currently available", for those times when an immediate answer is preferred by the originator rather than a delayed one. Availability would be determined by referencing a presence server in the enterprise, and experts would be alerted of the request by an instant message. Tacit has built an integration with Yahoo IM (as a proof-of-concept in my view), but it needs structured agreements with Microsoft (for LCS), Lotus (for Sametime), and Jabber for enterprise-oriented integration.
Recommended Next Actions
Organizations that have widely distributed operations and who run procurement activities out of multiple locations, should proceed immediately to talk to Tacit about ActiveNet 3. There are real cost savings to be made through finding people who have previously dealt with a given supplier, or through being able to consolidate a series of orders into one larger order for a bigger discount. The same idea applies to organizations with globally distributed research-and-development operations.
I think that there is great opportunity for a platform like Tacit in government, but at the cross-agency level. In other words, the implementation looks at what people are doing across a whole raft of agencies, and then builds a confidence index. An all-of-government implementation would do much to drive collaboration opportunities that positively impacted citizens.
Final Words
I'm looking forward to meeting Tacit's CEO, David Gilmour, at the Collaborative Technologies Conference next month in New York. If you have a question you'd like me to ask him, please drop me a line, or leave a comment below.



This appears to be possible to replicate with existing technologies, many of which are in the free and public domain, therefore what do you consider is the business resilience to competitor offerings?
Posted by: Ed Daniel | June 01, 2006 at 03:01 AM