Search Engine Optimization for the Enterprise

Michelle sent me an interesting article: Enterprise Search Just not there Yet. She and I both agreed... it is a terrible analysis.

To sum up, a lot of people are complaining that their enterprise search appliances aren't working right. Why doesn't it work like Google? they all say... I can always find relevant information on Google! Well, I got news for you:

People get paid LOTS of money to make sure Google can find their content.

People have meetings about getting high rankings... they hone their content. They obsess about keywords, and making sure the content is written in a readable format. They obsess about URLs, and make sure other pages link to this content. They register with a bunch of indexes, catalogs, and online yellow pages to boost relevance. They set up whole web sites for specific topics, neatly organized with clear, browsable topics. They hire very expensive specialists in SEO, and information architecture. In others words, internet content creators actually care if people find their content!

Google has it easy...

In contrast, how many enterprise employees obsess about findability, browseability, proper language, or keywords? How much of their content is even intended for an outside audience? How many of them even bother to enhance their content with useful metadata like title or comment? How many of them actively promote their content, and ask people to link to it?

Pretty nearly zero...

Without this effort, not even Google's laudable algorithm can find useful content in the enterprise... as is evidence by the general disappointment of the Google search appliance. No auto categorization engine can save you. No search engine will rescue you. No matter what people would like to believe, no software can ever replace a human being who actually gives a damn.

So... how do we fix the enterprise findability problem? It won't happen until people start caring both about them being able to find others' content, and others being able to find their content. I suggest you take advantage of the natural competitive nature inside humans... cash incentives might backfire, but nothing motivates people more than "your hit count is below average."

Start publicly ranking people on how findable their content is, and I guarantee that things will improve.

Comments

It does work

So, my company promotes this very incentive program right now. We are judged on how "useful" our content is, and by content I mean articles and solutions that eventually make their way onto our customer support website for the general public. This usefulness is measured by a few factors.

1. How many customers view that content in a given time period (simple hit count)
2. Customer ratings of that content (ranking and comments about that content)
3. Customer and employee referrals of that content (we can and should link to existing articles in every service request)
4. Service Requests prevented (this is done by a simple survey question)

All of this adds up to compiled numbers across the board. We all want our articles to be useful and widely read or we wouldn't take the time to create them.

We get monthly statistics on these factors which creates a friendly competition amongst the team members.

good policy...

however, that leaves out one very important aspect: browsability.

Individual content creators can only do so much to get high rankings... in order to really improve things on your company's web site, people should start ranking the findability of content, which means holding the webmaster accountable for poor information architecture. These techniques have worked in the past:

1) count the average number of searches per user
2) try to associate that with customer IDs who just "give up" and place a call instead
3) feedback on the search form, like "do these look like what you need?"
4) feedback on the articles, like "is this what you were looking for?"

Once you get too many negatives, its time to kick out the webmaster, and bring in the information architects and usability experts ;-)

analytics

Yes, optimization matters but I think the other thing that hampers findability is being able to -measure- your results easily.

In this, Google is still ahead of the game compared to enterprise search engines. Think Google Webmaster or Google Analytics and you'll see what I mean.

Uh oh, incentives for making info findable - smells like KM!

Fantastic to hear that your organization is aware enough to put these measure in place, rather than raw "contributions to knowledgebase." Awfully easy to load a whole lot of garbage into the system with that approach - and even THAT approach isn't used. The favorite is instead as Bex mentions, which is to pay NO attention to the findability of content (whether searched, linked or browsed).

Anyone doing more sophisticated measures and enforced processes like this, please get in touch with me (@dankeldsen). Always looking for good cases to discuss, writeup, and insert into our magazine, webinars, roadshows, etc.. It's terribly easy to do findability badly - let's see what we can do to show how to do it WELL, and make an impact.

Incidentally, the research report which leads off the article mentioned is our Market IQ on Findability, released (for free) last week. Grab it while it's hot off the digital press, and I'd love to hear feedback (good, bad, ugly, indifferent).

Cheers,
Dan Keldsen

Post new comment

The content of this field is kept private and will not be shown publicly.
CAPTCHA
This form prevents comments spam...