Forums/Product news and updates/Announcements

Improve Self-Service Support with Search Analytics

Pierre Grenier
posted this on March 12, 2012 15:18

We are really excited to announce our newest reporting feature: Search Analytics.  With this feature you will be able to gain insights into how comprehensive and effective your knowledge base is at helping customers find solutions.

Self-service support flow

Before we explain the Search Analytics feature, it is important that we review the typical self-service support flow (however, if you would rather get into the details right away, please read our excellent user guide on the subject).


searchflow.png

Essentially, a customer who is interested in self-service knowledge base will typically follow the above flow: search, view results, read a topic, and repeat until a satisfactory solution is found.

As we developed this feature, we kept that customer flow in mind to design a feature that could easily identify the potential issues that may come up during self-service.  Specifically, there are three potential issues during this process:

A : No results are returned to the user.

B : User does not click any results.

C: User opens a topic after reading a topic.

As you use the feature, keep this above process in perspective as it was a guiding factor in our design.

Why do self-service?

With this feature we are not advocating the implementation of a pure self-service process.  Our position is that some requests are better handled through a knowledge base.  In fact, no customer likes to wait hours before finding the solution to a simple question.

Search analytics provides you with the tool to improve on that self service.

Key features

  • View all your search traffic over the last 30 days.  Search traffic also includes searches completed on any of your live feedback tabs.  Expanded graphs allow you to get to the specific numbers.
  • Assess knowledge base and search coverage issues: searches with no results, with no clicks, or with tickets created. Graph those issues against total volume search to quickly assess self-service improvements over time.

    search_analytics_graph.png

  • Inspect your top 500 searches to pinpoint problems within your knowledge base.

    search_analytics_table.png
  • Easily export your search content for in-depth analysis in a third-party software (CSV export).
  • Identify the most important topics for each of your top 500 searches.

Benefits & Uses

For this feature, you will be able to:

  • Measure your knowledge base effectiveness for providing self-service support.
  • Quickly find customer-pertinent topics that are not covered by your knowledge base.
  • Find topics that need to be updated in your knowledge base.
  • Identify most important topics for your top user queries.

Uses

We understand that you might use this feature in unexpected ways.  You have several ways to filter and sort the Search Analytics data within the application.  If this is not sufficient, we provide you with the ability to export your top 500 searches (over the last 30 days).

Here are some use cases our beta customers were most excited about.

  • Find popular queries that do not return any results.  These queries indicate that people are looking for content that is not easily available in your knowledge base.  You might need to create a new topic or tag an existing topic with the terms that your customers use.
  • Look for popular queries that have a low click-through rate (CTR).  These indicate that the search results are not relevant.  Similar remedies as above can be employed.
  • Look for queries that typically lead to the creation of tickets.  Although there might be relevant results, these topics lack specific knowledge regarding a subject.

Lastly, we would like to hear from you about the creative ways you use this feature.  Please share your suggestions with the community on this thread.

Availability & Permissions

This advanced analytics module is available for Plus+ and Enterprise customers.  Also, the feature is accessible by administrators, moderators, and those who can view reports.

To open the search analytics dashboard, select Manage > Reporting, then click Search. Search data includes only end-user searches, not agent searches.

If you want to trial this feature, please contact sales@zendesk.com.

Getting Started & Learn More

There is no set up required to use this feature.  If you are a Plus+ or Enterprise customer, this feature is enabled by default.  Simply visit your forums to get started.

However, if you want to disable this feature, go to Settings > Channels, and then click Web Portal and look for the Search & Forums Analytics option.

You may consult our excellent user guide for a quick overview of how Zendesk Forums Analytics works.

 

Comments

User photo
Anthony Madama
brightidea

Great stuff! - Are there plan to push a lot of these metrics within the GoodData integration?

March 15, 2012 08:13
User photo
Julia Antenor

I am really loving this! Keep up the good work!

March 15, 2012 09:21
User photo
Travis Swientek
Rackspace

Hello! This is a great dashboard to have for a quick glance, but I'd really like to see the ability to switch it to show the last month. All metrics, throughout our company, go based on a specific month. ZD system admins would love running this report every 30 days instead of daily, anyway!! hehe :)

March 15, 2012 10:00
User photo
Pierre Grenier
Product Manager

Hi guys, glad this feature is helpful.

Anthony... about Gooddata, yes this is something we would really like to do.  No ETA yet though.

Travis, ability to specific a time range would prevent your admins to remember to run the report at a specific day during the month.  I agree.

March 15, 2012 10:48
User photo
Travis Swientek
Rackspace

@Pierre... Was speaking to the Zendesk engineering team. Sure it takes a lot of processing power to generate those stats daily. 

March 15, 2012 10:49
User photo
Thomas Andersen
tv2net

I would also like to see historical data, so we can compare e.g. searches with no result month to month and see if we are getting better.

March 16, 2012 03:42
User photo
Sergei Shablovsky

Great work! But...

May be You know (and describe in "" book well) no one ordinary web user "is interested in self-service knowledge base will typically follow". No one ordinary people love to spend time at the front of helpdesk. Facbook, YT or any else entertainment web place age better at time spend point view.

Only geeks, like developers, technician, like me and You love to "teach every pixel" of ability or feature. 

In now time ordinary people just love drop phone call or mail. Rather spend 20 sec to forum search.

I understand that this is uncomfortable truth for each of us. But this is era of post-computing. This is era driven by consumer interest rather technologies. This is why iPad are better then everyone other computer worldwide. Like we this truth or not.

So... Statistics are great. How do You about dynamical data? I mean much popular topics are on top, and make search in Feedback tab very powerful. Very-very powerful.

So this very decrease tickets to support. This is important. Very important. 

May 16, 2012 15:00
User photo
Stephanie Vaughn Hapke
AOL Inc.

Although I love this feature, I could really use the ability to narrow the time period, like maybe the last week, last three days, yesterday.

May 17, 2012 19:42
User photo
Mark Fly

Since discovering this later than sooner I would like to know if I can get some data back farther?  Perhaps the last 6 months? Since we are getting ready to revamp all of our Forum Media, this would be essential.

If not, Zen, maybe something else in mind?

 

 

 

May 25, 2012 12:39
User photo
Pierre Grenier
Product Manager

Well I must say that the feedback is really consistent.  Most of you want more flexibility on the time frame (not just last 30 days).

@Mark... I do not have any other alternatives at the moment, sorry.

June 06, 2012 13:00
User photo
Brian Tobin
evernotesupport

Looks like we can now run custom date reports for Knowledgebase analytics. Thanks for implementing.

February 21, 2013 18:07
Topic is closed for comments