0 comments on “Google Web Search is not Enterprise Search…”

Google Web Search is not Enterprise Search…

Summary: If you are in charge of Digital Workplace or intranet projects I bet a lot(!) on the fact that you continuously get the “requirement” to simply launch something like Google. Then the internal search experience will be so much better. This is a short but maybe a helpful one…

If it’s getting serious…the big “G” goes Taxonomy and Refinement, too

Yes, managing taxonomies is an effort. Yes, assigning taxonomy to enterprise information makes is less easy to just “dump” stuff onto a server. No, the application of meta data to information objects cannot be fully automated (yet).

There is a substantial difference between “finding something that somewhat meets my need” (aka Web Search) or “find something specific that is required to enable me to achieve a certain (unmovable) objective” (aka Enterprise Search). That’s why 2’300’000 results and a few “media type” categories won’t hep. You need refiners/filters, which are populated from a taxonomy. So if you’re trying to find a specific “thing” Google changes your search/refine experience as well.

Here’s a little example for the search query “laptop” in Google Web and Shopping search.

SIX15-02 Google Web vs Shopping

You can’t just “make up” stuff in terms of meta data that you apply to products that you want to register with Google Shopping. Similar to the categorisation and description that you have to apply to offers on eBay…if you’ve ever done that you know what I am talking about.

Web Search vs. Enterprise Search: it’s about controlling the “experience of finding stuff”

I’m simply sharing a slide that I’ve created as part of my work at Infocentric. Less for advertising but more of pragmatism reasons.

The Google Misinterpretation

Since it’s little hard to read here’s a link to a JPG.

If you need a few more reasons why taxonomies are essential to successful and user friendly (not publisher super low effort) experience:

  • freedom to combine/aggregate information objects dynamically
  • disconnect information from organisational/corporate structure
  • ability to deliver information to the relevant user profiles (matching information object and profile meta data)
  • option to associate “future” information objects (not in the system yet) to existing content via metadata
  • ability to “follow” subjects instead of people (like you partly deal with Twitter, when you follow/aggregate a #tag)

Is there proof that it works?

Yes. Simply get in touch with Estée Lauder, New York. The have extracted all their assets from the search index, applied a newly designed taxonomy, moved them back into the system and now: all purple roses! Check out the Twitter minutes for #intrelEU or Social Business Collaboration 2015 where they presented that insanely awesome project!


0 comments on “dynamic #personalisation will kill static pages; get your #IA homework done! #socialbusiness #socbiz #enterprise20 #e20”

dynamic #personalisation will kill static pages; get your #IA homework done! #socialbusiness #socbiz #enterprise20 #e20

Summary: There is literally no major site in the commercial web that contains a lot web pages in a static version. What you see is based on current events, information retrieved from cookies, your location, your account details etc.. In the future the same principle will apply to the digital workplace and thereby the future of what we today know as intranet. To achieve that a lot of homework has to be done. Only with a proper information architecture in place one will be able to profit from the potential that personalised and situation aware digital services contain.

“There is no information overload, just a filtre failure.”

Clay Shirkey

With every single day complexity within and around organisation increases. Every single document, email and comment creates a new data object that needs a home, stakeholders and (ideally) someone who needs it and someone who owns it. Today we perceive this growing volume of data and information as information overload. Because of that many employees are hesitant to really dive into “another channel” and add the intranet to their frequently visited destinations – or even turn it into the single most important destination for the personal work day. This is of cours enforced by the fact that the treasure that lies hidden within the corporate information assets is neither transparent nor hard to find. Not to mention that it should reveil itself to a user at exactly the right point it time: when it’s actually needed.

I believe that the future of the digital workplace and thereby the successor of today’s intranets will provide its major value but being directly connected with the person that is right now looking at it. Information from all repositories and the social sphere will look at the user and immediately match itself in relevance and priority to the specifics of that person. As a result not one single user will see the same “page” on the digital workplace. It will always be the combination of

  • relevance to person, role and situation
  • importance of the data object to the user
  • how time critical interaction is
  • how dependent others are on the user’s interaction with that particular data object

Dynamic Pages Illustration

The underlying principle is a powerful search engine that matches data objects and users in real time. It’s the power of search that most users don’t know about because it’s not (directly) connected to the personal action of typing a query into a search box and (hopefully) retrieve the stuff one was looking for.

However, it’s essential that the foundation for this convenient future of digital workplace is laid. “We have to establish an enterprise search function that gives us access to what we have” still is a mission for a lot of IT departments and (in a lot of cases) not more than wishful thinking.

Rubbish in, rubbish out.

Search engines are heavily dependent on their ability to identify the right data objects in order to be able to deliver them to the right person at the right time. That applies to active searches as much as it applies to the creation of fully situation aware and user centric pages. That’s where the homework lies… To build a powerful and more and more automated workspace for information workers companies have to do some house keeping and put effort in getting their information architecture and corporate taxonomy right. Without clear guidance for search engines what to match to what there will be no proper automation – and thereby no real value add for users and motivation to move away from the established ways of bouncing work related information back and forth through the know channels.

In this context I’d like to refer to one of my previous posts as well. I believe that “social” is just another extension to data objects and an improvement to the ability to specify the relevance of an object to the current user. Here’s the articles illustration as a teaser 🙂

Information Objects