Syndication in the Enterprise — it’s not all about you

The pressure is on to find an RSS / News solution for the Enterprise. But it’s not coming from the users (I think this is because users that want RSS just install their own client anyway).Every time it’s been raised internally in the last couple of years it’s come from the “Knowledge Managers” (inverted commas indicate sarcasm at the use of this title) who see it as a new communications channel that can achieve cut-through over e-mail. The two major requirements they present are:

  1. At an Enterprise level we can force users to subscribe to their knowledge channel.
  2. Any news reader has an alerting mechanism (pop ups) so that it gets peoples attention.

Primarily this is because the copious lotus notes databases, intranet sites, e-mail bulletins etc. are NOT getting the attention that they supposedly deserve (I try and argue that they are in fact getting EXACTLY the attention they deserve, but that’s not always well recieved). I quickly point out to them that implementing a news reader for every staff member that allowed us to force feeds on them would be quickly swamped by the competing parts of the organisation with their individual information spam which would turn users off and leave us back in the same mess.As you can see, it’s really not a subscription problem at all— it’s actually an attention one, and in an Enterprise there are two important halves to the attention equation;

  1. Am I as a user being told about the things that are important to me?
  2. And as an Enterprise how can I be sure that users are seeing the things that it’s important that they see?

After thinking about this for a few weeks now, I’m going to express the high level problem domain for the Enterprise like this. We need a solution that:

  1. Allows us to “declutter” e-mail and remove alerts and news from e-mail notification. E-mail should be for inter-personal communication ONLY.
  2. Provide a new channel (Enterprise RSS) for all “bulletin” or one-to-many styles of communication. Satisfy the needs of the Enterprise by allowing sophisticated profiling of both RSS feeds and users, so for example particular alerts can be automatically subscribed to particular groups of users. Most importantly provide detailed reporting and therefore added value on the read counts for each feeds and even the time spent browsing content.
  3. The content aggregator needs to function like Technorati, allowing me to browse feeds by searching and by tags in a central location — I shouldn’t need to go to every corner of the organisation to find feeds, they should be in one place. I also want some way of valuing the feed.
  4. Provides RSS “connectors” to other sources to allows users to aggregate their monitoring behaviours into a single location (big thumbs up to Lotus who have added an RSS feed generator to do this automatically with Domino databases).
  5. We need an Attention Client Engine (like Particls) which can monitor the news reader and provide two sophisticated features:
  6. Learns from peoples behaviours, profile and feeds what is important to them and alerts them accordingly.
  7. Provides this information back up to an Attention Server to aggregate and understand the attention profiles of the user base and then manage sophisticated alerting back to all users of information they may have missed that could be important to them all.
  8. Finally, lets the users have a degree of control over the information they are alerted to, personalising their own attention profile so that they are interrupted with the things that matter to them, while still being able to browse the news client for the things that matter to the Enterprise.

Here’s one attempt at a high level model for this. Feel free to comment on it — I can already think of a few additions, so I’d love to hear from you on yours.


Of course the Attention Engine and the News Client could be combined, but I think back to the values of the APML work group and even some basic architectural principles, there is value in separating these — I can select, train and tune an Enterprise Attention Engine separate from my subscription engine, and I’m not beholden to one subscription engine, or even one attention engine if I can separate the two.I don’t think we are there yet — open standards would need to be created to allow News Readers to publish their attention statistics to an attention engine (ie. even simple things like feed read counts etc.), but a solution like this would begin to radically alter the way in which users in an Enterprise experience information.I think Particls and News Gator are two companies both approaching this same problem space from different sides of the equation — I’ll be interested to see how they resolve the issues and the solustions they propose. So that’s what I’m looking for; I’m still to evaluate a number of vendors more fully, maybe my utopia exists, but I haven’t seen it yet. If you think you have something in this space, then feel free to contact me or comment below — I’d love to see what you’ve got.In the mean time, I’ll resist Enterprise RSS until I can be sure that we don’t just end up with another mess like e-mail has become — there’s only one chance to do this right.