Illustration by Bronwyn Gruet

Written by Mercer Smith

Knowledge base metrics are a set of quantifiable measures that help monitor your documentation’s performance and prove their return on investment.

Below, we’ll explore several options for metrics you can track, but it’s essential to recognize that they may not all be a fit for your team.

The key to knowing which metrics to use is understanding precisely what you’re looking to measure and improve upon. So the first step to picking metrics is understanding why you’re tracking them. Finding your root cause and motivation helps you understand where to focus your energy and measurement.

Here are the nine knowledge base metrics that are the most useful, along with guidance on how you can select the right ones for your team.

1. Page visits

Do you know how many unique visits your knowledge base receives every day, week, or month? If not, that’s a great place to start. This knowledge base metric tells you:

• Whether customers are finding your knowledge base easily.
• If your knowledge base is helping viewers resolve their issues.
• Which sections of your knowledge base your customers view the most.

Start by looking at the total number of visits over the timeframe. Is it higher or lower than expected?

If the overall numbers are lower than expected, consider adding links to your documentation site in places like your team’s email signatures, your marketing site, or even in navigation drop-downs within your product. You can also implement functionality to search your docs right within your contact form.

If you want people to use your knowledge base before reaching out, you should make your documentation easy to find.

2. Contact rate versus knowledge base visits

As valuable as contact rate is on its own, it becomes even more illuminating when you start to compare it to other metrics.

By looking at the relationship between traffic to your knowledge base and your contact rate, you can understand what impact, if any, your knowledge base is having on customers reaching out to your team.

For instance, if you see an increase in knowledge base traffic and a decrease in customer contact rate, you could assume that your documentation is helping to make some of that change.

Rather than measure the overall contact rate to knowledge base views, you could also count the number of customer support cases before and after implementing a new piece of your knowledge base.

For this knowledge base metric to be maximally effective, you must be able to look at data from the time before your team made changes to your knowledge base, or at least be able to view backdated volume reports. That will help you get the most precise understanding of your documentation’s impact.

3. Top visited articles

After you’ve looked at the total number of visits over time, consider diving a bit deeper and looking into categories and individual pages — are there specific docs that are highly popular and others that are mostly left unviewed?

Information like this will help give your team clues into what is essential for your customers and which docs aren’t very helpful or aren’t easy to find.

It may also tell you how your readers prefer to read documentation and suggest changes that you could make to create more readable content.

For instance, should so many people be reading the same expansive article, or would it be wiser to break it into smaller, more digestible topics? Or is there a product change that you could make to negate the need for this doc altogether?

When you know the types of information your customers read about most, you’ll know where to focus your attention and energy when creating new docs or revamping older sections of your knowledge base.

4. Conversations resolved on first contact

This metric is calculated by tracking the number of conversations or cases resolved on the first response by sending a document from your knowledge base. The inference is: If they could have found that document on their own, they wouldn’t have had to reach out to your support team in the first place.

The higher this metric is, the more it suggests that you could focus on helping customers find the articles in question to boost your deflection rates.

Recommended Reading

5. Failed searches

This metric tracks the number of searches that customers have input into your documentation search box that turned up without any viable results to show.

While the actual number of failed searches can be helpful, viewing the phrases that customers searched for without results can also be a helpful knowledge base metric for your support team.

Suppose you start to see a trend in failed searches or note high counts for specific terms or phrases. In that case, it’s time to consider rewriting the docs you have, adding additional docs, or perhaps rewording doc titles and keywords to ensure that they get captured when your knowledge base is returning results.

The rewrite could be as simple as changing a proprietary name for something in your product to something more generalized or adding a subtitle that uses the searched-for phrase.

6. Usage of help docs in support replies

Beyond tracking the efficacy of using knowledge base articles in support responses, knowing how often your team uses them can also be a helpful knowledge base metric.

Keep track of how often your docs are referenced in support responses to understand if there is growing confidence in the quality or breadth of your documentation. You may also consider including the use of help documentation as part of your quality rubric if you have standardized quality assurance as part of your support strategy.

For more direct tracking, consider focusing on which types of documentation are being shared most or if there is a spike in specific docs your team is sharing.

As you track this metric, consider how the usage of a specific help doc changes as it grows older.

Suppose you notice a tendency for your team to heavily use docs when they’ve first been implemented, only to lag as they grow older. In that case, it may mean that you need to put better knowledge base management in practice to keep your docs updated regularly.

7. Survey responses on knowledge base pages

Most knowledge base software offers the option to include a survey that asks customers how helpful the page’s content was. The survey could be a popover after the customer has reached the bottom of the page or inline text that includes a link or clickable pictures.

Keep track of your average responses over specific periods and whether the scores improve as you make changes to your docs. To hone in even more, you can track ratings for particular pages or sections within your documentation.

This metric is a tangible way to measure if your readers find use in your documentation and get direct feedback on what you could improve.

8. Contact rate

Calculate your customer contact rate by comparing your number of active customers to those who contact your support team for help every month. As companies grow, many run into this problem: Their number of users increases, and the volume of their support requests skyrockets.

In an ideal world, it’s best to gain product users without increasing your report requests too rapidly — you want to see “hockey stick” growth for your support requests, rather than a straight line into the top edge of your chart.

One of the best ways to achieve this type of growth is through self-service support: documentation, bots, and in-context product guides. So, one of the key (and most basic ways) to measure your knowledge base’s performance is by keeping track of your contact rate.

9. Average age of the last update

Nothing feels worse than finding the knowledge base article you’re looking for, only to discover that the company posted it back in 2006.

You should update both internal and external knowledge bases regularly to ensure that you maintain a level of trust and that you’re providing the correct information to your customers. After all, if they get to a page that has the wrong information on it, they’re going to end up contacting your team anyway.

Some information may only need to be updated once a year. However, if you can, try to set your articles to “expire” after a specific period and require an owner to look at it, verify its accuracy, or update it as needed.

Some knowledge base software has this built in, but the process can just as quickly be done using a spreadsheet and an automation tool like Zapier to record when each doc is updated. Then your team can sort by the oldest date in the sheet and have a list of docs to work off.

A secondary metric to consider in this same vein is the percentage of knowledge base items that your team has updated over the last month. Comparing the measurement month over month will give you good insight into just how up to date your knowledge base is.

Creating your knowledge base structure and then writing articles for it are the first few steps, but knowledge base metrics keep you on the right path as it grows. Use analysis to understand your areas of opportunity and take appropriate action to improve your help center’s effectiveness.

Not only will you continue to boost external trust and loyalty with your customers by having excellent self-service options, but you’ll have a fantastic resource for internal knowledge sharing and learning as well.

While all of these can be measured using the basic functionality of your knowledge base software, you may also want to look into deeper resources like Google Analytics or Mixpanel. It’s super helpful to keep your knowledge base metrics aligned with other points of data cross-functionally. The more you measure, the more you know