Archiving content: Sorting diamonds from duds

We’ve all had the experience of sorting through our closets and choosing some clothing to discard or donate. You may find a pair of pants that doesn’t quite fit anymore, or a shirt that’s an outdated style, and those should definitely be culled from your closet. It’s an exercise in sorting the good from the bad and the ugly, just like archiving content.

One of the key pieces of any archiving project, whether you’re keeping your knowledge base healthy or your wardrobe fresh, is finding a few gems that may have flown under the radar but are truly valuable. Maybe you’ll find that comfy old pair of jeans you forgot you had, or maybe you’ll see a helpful article that had an unsearchable title and never saw the light of day. Getting value out of something you didn’t realize you had is the best way to maximize your time and money, whether you’re managing a knowledge base or dressing for work.

Identifying archive candidates

The first step in archiving unused knowledge is deciding which articles are in need of review. There are several ways to identify archive candidates, but these two help you get the most value from your time:

Sort by Last Modified

As your knowledge base evolves, you may find the last modified date of a document to be a helpful tool in determining if it’s out of date.

Sort by Page Views

Sorting by the page view count can tell you if an article is helping your readers. If it’s not used, there’s always a reason for that. It may not be helpful, or it may not be findable. Either way, it presents a problem to solve.

Assessing the content

Once you’ve identified your potential pool of archive candidates, the next step is to determine which ones could add value to your knowledge base. Auditing the content prior to archiving it maximizes the benefit of your content.

Review each article and ask these two questions:

  • Is this knowledge useful?
  • Is this knowledge valid?

If the answer to each of these questions is no, then the choice is simple: The content can be archived. If you’ve answered yes to either, you’ve found a diamond in the rough.

Improving the content

Content that is both useful and valid should be reviewed and improved. Having a low use count or a long gap since the last update are both signs that the article is not living up to its potential.

For an article with a low use count, some exploration is necessary. Is it an orphaned page that doesn’t contain any inbound links? Can you find it using your search tools?

If an article hasn’t been edited in a while, can it be made relevant again? Older articles that pertain to outdated processes may help resolve new issues. Content around legacy products may not have been updated in a while, but if the information within has already been polished and refined over time, users of older products may still find them useful.

Discarding the rest

For some content, there’s no hope. A service has been discontinued, a new process has been implemented, or an event has passed. The most important part of any sorting exercise is getting rid of what’s no longer useful. It’s very likely that corduroy bellbottoms will be trendy again, but pushing them aside every morning to get to the clothes you want to wear now is a needless hassle. When this principle carries over to knowledge, it represents wasted time and effort for your customers or employees and can diminish the impact of your content improvement efforts. Readers will wonder why clearly outdated content still exists and may not trust that other information within your support site is up-to-date.

Moving forward: Baby steps

Archiving content is a vital part of keeping your knowledge base healthy and analyzing knowledge that’s fallen through the cracks. Even if your organization is not ready to kick off an archiving program, you can find a smaller measure of success by working these principles into your workflow as you manage, review, and edit content. Assessing unpopular content on a smaller scale is an incremental change that can truly make a difference.

Knowledgebase metrics identify the symptoms, but let’s find the root cause

Content metrics are fairly obvious, right? When you want to measure how useful your support site is, you’ll look at the basics – Call deflection, use count, issue resolution, or even customer ratings on individual articles. These are important metrics, but they have one fault: They define the problem, but not the solution.

What’s the difference?

When you use data to determine how effectively your customers are using your content, you’re getting a high level readout on what your areas of opportunity are. Reporting on these trends will provide you with feedback on what you need to improve, but it won’t provide enough detail for you to make the course corrections that will take your content to the next level.

To get the right diagnosis, you need to ask the right questions!

To get the right diagnosis, you need to ask the right questions!

A deeper analysis of your articles will get you to the root cause of low performing support site content. Concise paths to resolutions, effective article structure, an easy-to-follow brand voice, and media that adds value are all examples of important attributes of a high-quality knowledge object. These, and other attributes, can be quantified and compared to industry best practices or your own top-performing articles. By observing these metrics, you can identify quick wins to make your knowledge more accessible, more useful, and more satisfying.  Here are a few examples of common Knowledge Base woes that you can solve by measuring small aspects of your article quality.

 

This article didn’t help me!

How often do your Knowledge Base users say, “This article didn’t help me!” while the answer is a mere six lines down in the page? If you’re nodding your head as you read this, the answer is simple: Too many!

It’s long been known that most people scan an F-shaped pattern when reading web content. Titles and introductions catch the eye, and then readers skim down the page, pausing at headings and picking up mostly the first few words of each line.

To diagnose this issue, try this: Review a sample of support articles that contain accurate information, but that your readers have called out as unhelpful. Tally how many of them use scannable, quick sentences listed under informative headings. Does the data surprise you?

I can’t find the right article!

Do you hear “I can’t find the right article!” when you know the knowledge your readers need is in there? If that’s a trend in your support content, the quick answer is that your content is either not searchable by your site’s search, or not findable to your readers.

Most Knowledge Base search tools weight the title heavily, so skillful use of keywords will help get your readers to the right article. Once they find the right article, the title confirms for your reader that they’re in the right spot. Good titles are only a few words, but they pack a great punch.

If you want to evaluate how titles contribute to your content’s success, this exercise will help: Identify a handful of articles that have surprisingly low view or use counts, or are related to issues that drive a large amount of customer contacts. Then, take a sample of articles that are used frequently and have a higher resolution rate. Review the titles of each of these articles. What differences do you find? How can you revise the lower performing articles to mirror the top performers?

The two examples above should give you some food for thought on how to measure your support site against industry best practices. Now that you’ve identified both the problem by measuring your customers, and the solution by measuring your content, you’re well on your way to evolving your Knowledge Base to be more useful and more effective.