Search Behind the Firewall, or, The Potential for the Information CMDB

…or, Lunch at The Blogging Enterprise 2005.

From talking with a table-mate from National Instruments, I came across the
idea that search is, and/or could become, something like a CMDB for
information in the enterprise:

  • Search can be the one-stop-shopping place for keeping up with
    info as it’s created, updated, and deleted. By caching and indexing
    all the information, you can accomplish the second thing…
  • If you’re looking around for something, you can search for
    it. The CMDB analogy being that, with a CMDB, you don’t have to go
    visit every single machine in your network to find all the XP boxes
    that need SP2 applied, you just ask the CMDB, ’cause it stored all
    that profile info.
  • If “actions”/macros are added to search results, and batches of
    search results, you can move towards a good
    enough
    approach to a compliance
    oriented architecture
    .

Search Actions

That last point is where is gets exciting: what would make search
fully like a CMDB/Systems Management app would be if actions were
linked from your search results and you could take batch actions on
search results.

Tag & Bag Information, or, Regulation, Compliance, and CYA

For example, let’s say you need to retrieve all information related
to a patent lawsuit you’re involved in, for example, about JPEG. It
needs to be marked and bundled up to send to your lawyers. With a next
generation search app, you could write-up search queries like “JPEG or
JPG or ‘image file formats'”, etc. Then you could hook up an action
that would create a zip file with PDFs of each search result, to send
to lawyers.

Since it’d all be involved in a lawsuit, you’d want to assign ID’s
to each item so you could track them. So, your system could do
this. Even better, you could get a history of all the changes that
information had gone through, ’cause your search app had cached each
version of the doc, or, as CVS would do, the diffs between each
version. This would be analogous to versioning your configuration in a
CMDB.

RSS

By slapping RSS on top of search (creating RSS feeds from search
results), you can start to do mash-ups of behind-the-firewall
information. The top contenders are the “classic” examples of
distributing newsletter information to employees, thought leadership,
and statuses that you’d otherwise waste time in meetings
gathering.

Ending the Intranet Ice-age

As I’ve noted
before
, all the services you take for granted on the internet —
Google, flickr, del.icio.us, bloglines, technorati — don’t exist. The
technology on typical intranets is effectively frozen in the late
90’s. What’s exciting about search behind the firewall is that it can
be the platform/beach head for providing all these features.

The Actions Are Endless

The patent CYA is a pretty simple example. But, you can imagine how
much you could do with your information once you normalize and cache
it into a central search repository: like I’m saying, it’s the
Information CMDB.

Lighting the Dark Information

Another interesting analog between a CMDB and search is that the application helps you get value out of what you already have. Also, it helps you stop loosing value from what you have. In the CMDB world, by imposing the regulations that come with that ITIL-mindset, you make IT more efficient and controlled, the goal being to let your IT group be less fire-fighter focused and more “how can we make money with all this IT” focused.

That
is, large companies have tons of information: silo’ed, “dark,” or
otherwise dead and lost. There’s money to be squeezed outta those stones, and search
is looking like a damn good juicer.

Tags: , , , , , .

More Google Enterprise Apps On the Way; Search Behind the Firewall

In a sort of pre-announce, pre-announce, Schmidt said:

The business is growing very quickly. We’ve added all these new features. It makes sense now to just keep adding more Google to the Google-in-a-box. So a reasonable expectation is that more and more of what’s on Google will be available in the blue box. And that’ll keep them busy for years. Because when you do that inside of a company, you have a higher level of security and privacy and all that stuff. … Imagine the evolution of this product line. It gets very interesting. Without pre-announcing anything, it seems like a no-brainer. You have lots of these things sitting around inside all these guys’ networks. Just think of the strategic value of that to Google.

That’d be exciting.

Back in Jan. or Feb., thanks to an internal weblog post I made and an eager to help out VP at work, we got ourselves a Google mini at work. So far, it’s been fantastic. People send in emails all the time saying how much they like and use the search. Having search behind the firewall, real, works-out-of-the-box search is a sort of major event for an intranet.

With Search, Unstructured Intranets are OK

In a company as large as mine, there’s 100,000’s of web pages on the intranet (and 100,000’s more that I filter out of the search). With that many documents, and so many people, it’s difficult for people to add lots of helpful structure to all those URLs. Sure, there are plenty of systems that are highly structured (like project pages), but anyone/team can start up a web server, and often do, to host any sort of document. Worse, once you setup a structured page, it’s care and feeding often go by the way side: internal documents are always a much lower priority than external releases, i.e., money-makers.

From this, we might derive Coté’s Theory of Intranet Content, which goes: “you’ll never be able to structure everything effectively, so you might as well not try…very hard.” Therefore, unless you have search to layer over all that, it’s impossible to find what you want as quickly as you want. Thus, you need search. Q.E.D.

The Google mini has been the answer to that. Indeed, as the theory hint at, it almost makes all the structured sites (and apps we use to create them) kind of gratuitous. Just like GMail, Bloomba, or LookOut: once you add a search box to your email, all those folders aren’t really needed. (At best, folders are just 1-1 tags…but, you want 1-many tag-as-folders, of course.) Once you don’t need folders — because you have search — you don’t really need structure.

Simplifying the Intranet

Of course, some structure is required to simply make creating and hosting content as easy as possible. You need a light content management system. So far, in my opinion, the answer to that light-CMS has is to use wiki’s and optionally,
weblogs*. With a wiki, creating a document is as easy as clicking “Edit This Page.” And a weblog has an inherent stream of content that’s a natural fit for people’s desire to get streams of content about relevant topics/people. Once you add search to that mix, you just about solve all your Enterprise intranet problems:

  • Employees/users can easily/quickly create documents on the wiki.
  • Because the documents are on the wiki, documents can easily refer/link to each other.
  • Because wikis are open, those documents are instantly available to anyone who wants to read them — no more documents being locked up and lost in the email blackhole.
  • Instead of using email lists, or long CC:’s to broadcast out information, have discussions, innovate, or just get people talking to each other, everyone simply broadcasts via their weblog.
  • Bonus Feature: because the wiki, search, and blogs are all web applications, people don’t need to install anything on their desktop except a web browser. Meaning you don’t have upgrade worries, security problems, and (following the Google and open source wiki pricing models) a very simple pricing scheme to follow.
  • Bonus Feature: though most wikis, blogs, and search don’t waste time worrying about internal DRM (“only team A can see these pages, while team B and team C can see these pages, along with person X.”), you can still quickly do information management/deletion when you need to. Because all these services are centralized, if you want to go and delete some sensitive document, you just go into each system and delete it. Compare this to hunting down a sensitive document that needs to be virtually-shredded in every employees inbox, desktop, USB thumb-drive, home computer, etc.
  • Bonus Feature: linked to the previous point, there’s better, even instant, “employee access to data life-cycle management”; that is, when an employee quits, you want to pull all access to sensitive data. In the case of all these web apps, the employee never really downloads any thing, everything is on the intranet. So, to cut off their access to all that info, you just cut off their intranet access.

A few of those points are a bit forced, but I think you get the idea.

(As I’ve posted before, my experience with weblogs behind the fire-wall hasn’t been a huge success quantity wise, but quality wise it’s been great: if it weren’t for the weblogs, we wouldn’t have gotten the Google mini. It’s taken several years for the wiki to really become a core part of the groups I work in, and I suspect the same will be true for weblogs.)

More Blue Box Applications

Of course, what’s even more exciting about the above quote is the promise of having more Google apps available behind the firewall.

First off, we’d want GMail. No one really likes Outlook/Exchange, but everyone uses it (probably because everyone has those big Microsoft site licenses). As many people have noted, if you just throw in Calendaring that works as well as Outlook/Exchange, you’d have the full Email/Calendaring package that you need. And if the Google story continues to be the same, it’ll be both cheaper and more simple than Outlook/Exchange.

Google Groups would be another nice thing to have. There are tons of email lists behind the firewall, and if you use Outlook/Exchange, there’s really no good way to say, “put the archives of this lists on a web page so everyone can read/search them.” Not to mention the fact that once you reach those damn email storage quotas, you’re forced to start blowing away all those emails with all that important info in them that you’ve been trying to save.

And, of course, there’s Blogger.

Indeed, one could see that Google would simply create a “Collab Blue Box” that just loaded all those things together, with search. And, if they priced it as well as they do the Google mini, it’d sure be an intranet app killer. Even Outlook/Exchange if they could get a story that all the check-signers and IT guys believe in enough to make the switch.

Text Ads Behind the Firewall

“Eric [Schmidt] had a great quote the other day. He said, ‘When we said all of the world’s information, we meant all.’ So that means FORTUNEs from the 1930s, TV content…there’s very little content we’re not interested in.”

Of course, the more material that can be searched means more opportunities for Google to place ads next to those search results.

The above from a Fortune article on Google got me thinking: what about text ads behind-the-firewall, at work? Sure, that’s kind of like Text Ad Taboo now-a-days (business allowing other business inside their walls to make cash), but only seemingly.

At our work, there’s advertising all around:

  • The TV in the breakroom always tuned to some news channel…with ads.
  • The Coke machines in the breakroom and elsewhere with their big ads for Coke, C2, and whatever else.
  • All the logos and company names on products.
  • All the advertising we see on web pages when try to search for the impossible.

In short, we’re surrounded by ads at work.

Piggybacking on Intranet RSS and Search

So, adding to the idea of adding RSS feeds to existing middleware products (as Baus summed up my post)…when you get all those RSS feeds, you have “more material” to put Text Ads next to. You just somehow have to figure out a way to allow companies to allow you put Text Ads in there: perhaps sharing in the profits just like I share in the profits of having Google Text Ads on my weblog.

Using Text Ads for Internal Advertising

Indeed, if you applied Text Ads behind the firewall, there’s more, new things you could do:

  • Sell ads internally. Different departments, teams, projects, whoever, could buy ads to advertise themselves internally. At large companies, broadcasting that you exist and what you do internally (never mind externally) can be a challenge. I’m not sure you would use real money to sell the ads. Probably something akin to “blue money.” The “KeyMatch” feature of the Google mini seems to do this.
  • More effectively get internal announcements and news out. Most intranet sites at large companies have news articles about goings on at the company. But, you have to go to the intranet sites and read the news. Instead, if these news items were laced in as Text Ads, they’d show up next to relevant feeds and searches that employees subscribed to and read. Meaning, that the employees who “should” be reading those news items would have a higher chance of reading them.

The Headhunters are Already There

There are, of course, problems with ads that the Business wouldn’t want employees to see: ads to join unions, ads for headhunters, or other ads that would be perceived as damaging the business.

The problem with that line of thought is that employees (as outlined above) are already saturated with ads, and they’ll just be seeing more of them in the future. In the information-worker business (and, increasingly in any other line of work in America), businesses that filter their employees information flow just end up looking stupid and clueless.

When it comes to the free-flow of information, what’s good for the goose is (probably) good for the gander.