5

Fine Tuning: Insights reports to transform the way you do support

In this Fine Tuning session, we’ll look at how we can use data and reporting to help accomplish the below support issues and more:

  • Give great support to every segment of customer
  • Uncover the issues that are driving your support costs
  • Identify and prevent potential problems before they occur with proactive messaging

Missy is a Data Analyst on the Customer Success team. She comes to Success from Support where she was a Product Champion during the Insights launch. Missy loves collaborating with customers to make compelling metrics for their support organizations.

See all the Fine Tuning series discussions .

Getting to Know Insights

Zendesk Insights is the business intelligence (BI) platform that helps you make more sense of your data. Along with pre-built dashboards designed around best practices for customer service reporting, Insights allows you to create dashboards that are fully customized to reflect the business data you store in Zendesk custom fields., When you understand how to customize Insights, you can connect your customer service activity to compelling business outcomes like retention, renewal, and revenue—empowering you to. Insights is powered by GoodData .

This Fine Tuning session will focus on three ways to add greater context to your customer service data.  We’ll provide detailed instructions on how to create reports that will help you

  • Improve support operations,
  • Discover what drives support costs, and
  • Mitigate potential problems through proactive customer engagement
  • These examples are drawn from our white paper on Customizing Zendesk Insights .

Before you get started creating a custom report, we recommend navigating to the native GoodData UI.

Camera Obscura—a vendor of high-end photography equipment and maker of critically acclaimed photo editing software—relies on Zendesk for customer service*. As their business grows and their product line becomes more differentiated, Obscura executives know they need a more detailed picture of their support operations to maintain a high-quality and cost-effective support experience for their entire customer base. Let’s start by taking a look at how Obscura’s support operations team uses Insights to monitor support performance across their key customer segments.

*The reporting techniques discussed in this article are grounded in real customer use cases. The company Camera Obscura, however, is totally made up.

Part 1, 8 am: Provide great support for every customer segment

When Obscura’s customers purchase a new camera, they spend a good chunk of change. While some of those customers only carry their new toys to low-impact excursions, like weddings or baseball games, others like to shoot photos from the mountainside or ocean cliffs. Because customers face different levels of risk that their camera will be damaged, Obscura offers two different warranty plans: a limited lifetime warranty that covers any damage to the camera, and a complete lifetime warranty that also covers accessories. There are also many customers who choose not to purchase a warranty at all.

To make sure that all customers have a great support experience, regardless of their level of coverage, Obscura breaks down their support KPIs by warranty type. With a custom user field to show the requester’s warranty, building the next three reports is a piece of pie.

A clearer picture of the Obscura customer base emerges immediately. Complete and Limited warranty customers drive over 90% of ticket volume. While the Complete coverage claims are simple to evaluate, the Limited coverage claims are more nuanced—driving their resolution time to an extreme 48 hours. In response, Obscura support managers assembled a task force of advocates to develop new macros that will make the limited warranty claim evaluation process more efficient. Customer Satisfaction is a volatile metric, especially for no-warranty customers. Why? Advocates frequently have to say “no” to these customers’ requests for returns or exchanges. With a representative sample of tickets, support managers have engaged the marketing team to more carefully set customer expectations during the online purchase process.

Some questions we might ask after looking at these reports include:

  • What is happening with the full resolution time of tickets from customers with the limited warranty?
  • Why is satisfaction for the customers with no warranty higher (minus Week 40) than the other two paid plans?
  • How do we feel about the no-warranty plan having the lowest reply time?  Is there a reason for this?
  • Why are 50% of the tickets we receive coming in from Complete warranty customers?

Possible actions to take:

  • Use a drill-in to examine the About Field for the Complete category in Ticket Volume by Plan to determine if there is a certain subject that is being miscommunicated to your VIP customers.  From here you can up-level agent training in these areas, make macos to deflect common topics, or add documentation to your knowledge base to increase self service.
  • To Drill in, navigate to the report and click the show configuration panel

  • Select “About” under the drill-in panel, click apply

  • Drill into the _Median Support Times by Plan _by _Ticket Group _(same process as the last bullet point) to see if these tickets are getting caught in a support group and take actions accordingly.
  • Hire more people
  • Reallocate agents to different specialization groups
  • increase documentation and training

Reporting Recipe

To replicate this dashboard, you’ll need to create three separate reports. Also note that this recipe uses a custom user field called Warranty .  Here is an example of a report recipe that involves creating a custom field .  Obscura used the Zendesk API in order to pump in their customer’s warranty purchase information to the User profile page.

Ingredients:

  • 1 custom user field for ‘Plan’
  • 3 new Insights report

Add a custom user field

  1. Click the Admin icon () in the sidebar, then select Manage > User Fields.
  2. Click add custom field and select the custom drop-down list option; call it something like Warranty .  For more information, see Adding a custom ticket field .
  3. After creating the custom field, you’ll need to generate tickets that use the custom field.  After your next Insights sync (daily for our Professional customers, hourly for Enterprise), the field will be added to your Insights project as an Attribute.

Navigating to GoodData

  1. Click the Reporting icon () in the sidebar, then click the Insights tab.
  2. Click the GoodData link in the top right corner of the panel, then select Reports > Create Report.

Now we’re ready to start building reports!

Report #1 Ticket Volume by Plan

WHAT

  • # Tickets

HOW:

  • Plan type (Custom Organization Field)

FILTER

  • Last six weeks by week/year (ticket created)
  • Click Select from a list of values (including date ranges).

  • Click Week (Mon-Sun)/year (Ticket Created) as the attribute.
  • Select the right hand side ‘Floating date range’
  • Enter the number 5 into the text field
  • Click apply

Configuration Panel

Chart type: Stacked bar graph

Report #2 Median Support Times by Plan

WHAT

HOW

  • Plan type (Custom Organization Field)

FILTER

  • Last six weeks
  • Click Select from a list of values (including date ranges).

  • Click Week (Mon-Sun)/year (Ticket Created) as the attribute.
  • Select the right hand side ‘Floating date range’
  • Enter the number 5 into the text field
  • Click Apply

Configuration Panel

Report Style: Bar Graph

Report #3 Satisfaction by Plan

WHAT

  • % Satisfaction (built-in metric)

HOW

  • Week (Sat-Sun)/Year (Event)
  • Plan type

FILTER

  • Last six weeks
  • Click Select from a list of values (including date ranges).

  • Click Week (Mon-Sun)/year (Event) as the attribute.
  • Select the right hand side ‘Floating date range’
  • Enter the number 5 into the text field
  • Click Apply

Configuration Panel

Report Style: Line Graph

Part 2, 11 am: Where do support dollars go?

In the first reports, we looked at support outcomes like resolution time and satisfaction, and broke them down into their component parts to better understand them. While reporting on outputs is necessary and good, it isn’t very actionable: outputs don’t tell you where to act to change your support operations.

Inputs are different. Breaking down the resources and workload that go into your support process can cast light on specific areas in your business that need immediate attention. Since the support team engages the customer at the end of their purchase cycle, their interactions contain valuable feedback for every other team that affects the customer experience, from product design to customer education.

Obscura categorizes each support ticket by its topic using a custom field called About . They also track the time agents spend working on tickets using the Time Tracking App . Since advocate salaries are roughly aligned with their status as a Level 1, 2, or 3 support agent, they can easily create a custom metric to calculate the money spent on customer support—and break that down by issue type. The report below provides a clear picture of where support dollars are spent, highlighting areas where other teams can implement changes to reduce the amount spent on support.

Right away the biggest cost drivers are clear: agents in all tiers are spending lots of time on warranty claims. This aligns directly with our discovery above that warranty claims drive high resolution times, and adds weight to the argument for better warranty processing business rules.

The Editor Pro software is the next culprit: does the head of user experience have any idea that the installation and upgrade process alone drive almost $15K in support costs per quarter?

The Customer Education team is hungry for topic ideas for their instructional YouTube series. Obscura can fill the demand for more assistance on portrait shots and night photography while cutting costs in customer support. That’s two big wins for the price of one!

Some questions to ask:

  • Why is the installation such a pain point for our customers?
  • Why is our most technical team handling such a large portion of the billing tickets?
  • What is going on with the lens and camera to perpetuate so many claims?

Possible actions to take:

  • Bring installation issue to Product.
  • increase training on installation for agents
  • add seats for you accounting team to handle the billing issues
  • Train Tier 1 team members on exporting images, these seem to be escalated a lot.

Reporting Recipe

This recipe combines a useful app from the Zendesk marketplace and a creative application of the GoodData platform to deliver answers to the million dollar question: What are my support cost drivers?

Ingredients:

Adding a custom field

  1. Click the Admin icon in the sidebar, then select Manage > Ticket Fields.
  2. Click add custom field and select the custom drop-down list option; call it something like About .
    For more information, see Adding a custom ticket field .

After creating the custom field, you’ll need to generate tickets that use the custom field.  After your next Insights sync (daily for our Professional customers, hourly for Enterprise), the field will be added to your Insights project as an Attribute.

Install the Time Tracking app and create custom metric to track time

  1. Download and install the time tracking app here .
  2. Click the Reporting icon () in the sidebar, then click the Insights tab.
  3. Follow the instructions from this recipe to develop a metric called Total Time Spent (TTS).

The final metric will look like this:

Create three custom metrics to track time in group

Using this metric we’ll make three metrics that use the TTS metric.

  1. First migrate to Manage > Metrics > Create metric.
  2. Click custom metric.
  3. Create the following metric.
  4. Create two more metrics like this one, changing only the Hourly Rate and Attribute Value (e.g. 20 > 21; Tier 1 > Tier 2).

Report #1

  1. Click the GoodData link in the top right corner of the panel, then select Reports > Create Report.
  2. Use the three custom metrics to build the report.

WHAT

  • Tier 1 $
  • Tier 2 $
  • Tier 3 $

* These are the three custom metrics we made in the steps above

HOW

  • About Field

FILTER

  • Last quarter

  • Click Select from a list of values (including date ranges)
  • Click Quarter/year (Ticket Created) as the attribute.
  • Select the right hand side ‘Floating date range’
  • Select Last Quarter
  • Click Apply

Configuration Panel

Report Style: Bar Graph

The final product

Right away the biggest cost drivers are clear: agents in all tiers are spending lots of time on warranty claims. This aligns directly with our discovery above that warranty claims drive high resolution times, and adds weight to the argument for better warranty processing business rules.

The Editor Pro software is the next culprit: does the head of user experience have any idea that the installation and upgrade process alone drive almost $15K in support costs per quarter?

The Customer Education team is hungry for topic ideas for their instructional YouTube series. Obscura can fill the demand for more assistance on portrait shots and night photography while cutting costs in customer support. That’s two big wins for the price of one!

Part 3, 2 pm: Prevent problems before they occur

Making a sweeping change to an important product line can send a wave of tumult through your customer base. Naturally, the company relies on support to intercept the confusion and pain. In the past, teams would wait to act until tickets began to flow in, deflecting the onslaught of questions and complaints with one-size-fits-all canned responses. The cost to support is monumental, and customer satisfaction is invariably impacted.

What if there’s a better way? What if instead of taking customers by surprise, asking agents to work overtime as they buckle under exploding ticket volume, and dragging down support KPIs, you could act proactively to manage customer expectations and assist them through the transition, garnering great satisfaction scores in the process?

That’s exactly what Obscura’s support team did. When their software team announced the impending launch of the next version of Editor Pro software, support managers used the Proactive Ticket App * to run a campaign that notified customers in advance. The tickets they created on behalf of customers included a full explanation of the switch, the required actions the customers needed to take, and an open invitation to reply on the ticket with any questions. They created an Insights dashboard to monitor the success of their campaign. When the engineering team finally made the switch, support felt confident that they had reached the vast majority of impacted customers already.

*We love the Proactive Tickets App, but it is a Zendesk Labs project in active development. It is not officially supported by Zendesk.

Armed with these reports, support managers can see that over 80% of affected customers have already been engaged:

  • 34% of tickets are Solved—this represents the proportion of customers who have responded to the proactive engagement, have been helped by a support agent, and have resolved any outstanding issues.
  • 28% of tickets are Pending. The customers behind these tickets are in ongoing conversations with support, and have received all the information they need so far.
  • 19% of tickets are Open. This number is more dangerous, as these customers are waiting to hear back from an agent.
  • 19% are New. This is very problematic: these customers are replying to ask for help, but the tickets have not yet been assigned.

Support can also monitor the quality of the customer experience provided by the proactive campaign by setting customer satisfaction goals using bullet charts. The chart above shows the team falling short of their goal of 95% satisfaction. After the campaign, managers can mine the satisfaction survey comments for feedback about how to improve the proactive experience.

They can also keep track of which tickets are mired in long agent/customer conversations. Many comments can be a leading indicator of an unsatisfied customer: drilling into this report to locate the at-risk tickets for escalation to a senior advocate allows managers to avert bad satisfaction ratings.

Some questions we may ask when looking at these reports include:

  • Why are so many tickets in pending?
  • What is going on with the 37 tickets with > seven agent comments?
  • Why have almost 20% of our customers not seen this message?

Possible actions to take:

  • Increase agents assigned to handle the EOL tickets. This will help to address the number of open tickets
  • Make strides to alert the customer of the EOL.  This will help to address the 19% of customers who have not yet opened the ticket - For VIP customers, have an account manager reach out directly to the customer
  • Drill into the satisfaction score to determine whether low ticket satisfaction is due to qualms with the EOL or level of support
  • Engage the Support Manager to address tickets with greater than seven agent comments

Creating This Report

Ingredients

Adding a custom field

In this scenario, Obscura has a custom checkbox field called Editor Pro .  If a customer has this software, the checkbox is checked.  They’ll use this checkbox to determine who will receive the proactive ticket.  Let’s first walk through the steps to set up this custom field.

  1. Click the Admin icon () in the sidebar, then select Manage > User Fields.
  2. Click add custom field and select the custom check box.
    For more information, see Adding a custom ticket field .
  3. After creating the custom field, you’ll need to generate users that use the custom field.
    After your next Insights sync (daily for our Professional customers, hourly for Enterprise), the field will be added to your Insights project as an Attribute.

Sending out the Proactive Tickets Campaign

  1. Follow the instructions here to install the Proactive Tickets App .
  2. Create a Customer list .
  3. Use this customer list to target a subset of customers.
    Your campaign page will look something like this:
  4. Send off the campaign. Take note of the tag that is generated for the campaign. In this case it’s eol_campaign.

Navigating to GoodData

  1. Click the Reporting icon () in the sidebar, then click the Insights tab.
  2. Click the GoodData link in the top right corner of the panel, then select Reports > Create Report.

Building the custom metric to count tickets with tag: eol_campaign

  1. Migrate to Manage > Metrics > Create metric.
  2. Click custom metric.
  3. Create the following metric.  Click Add when finished.

Building the custom metric for Satisfaction Target

  1. Repeat the steps above under " Building the custom metric to count tickets with tag: eol_campaign" to open the metric editor
  2. Build this report to enter the target satisfaction as a decimal.

Let’s get building!

Report 1: EOL Tickets by Status

  1. Select Reports > Create Report.
  2. Fill in the What, How and Filter for your report as follows.
  3. Click Done when you’re finished.

WHAT :

  • # Tickets

HOW :

  • Ticket status

FILTERS :

  • Numeric ranking filter where we filter by Ticket ID and select the metric # Tickets w/EOL_campaign is equal to 1

Configuration Panel:

Report Style: Stacked bar graph

Recipe 2: Actual vs. Target CSAT

WHAT :

  • % Satisfaction Score
  • Target^

HOW :  nadda

FILTER :

  • Numeric ranking filter where we filter by Ticket ID and select the metric # Tickets w/EOL_campaign is equal to 1

Configuration Panel:

Report Style: Bullet

Inspiration: https://support.gooddata.com/entries/23812742-Customer-Satisfaction-Bullet-Chart

Recipe 3: Tickets with > 7 comments

WHAT :

  • # Tickets

HOW : nadda

FILTER :

  • Numeric Range filter:  attribute to filter is ticket ID, metric is # Public comments is greater than or equal to 7
  • PLUS sub filter attribute * User Role is Agent *
  • Numeric Range filter: filter Ticket ID where metric # Tickets w/eol_campaign is 1

How to see what tickets have > 7 comments

  1. Click Show configuration > drill in settings > select drill in by Ticket ID
  2. Click the number on the dashboard (here the number is 37)

Some questions we may ask when looking at these reports include:

  • Why are so many tickets in pending?
  • What is going on with the 37 tickets with > seven agent comments?
  • Why have almost 20% of our customers not seen this message?

Possible actions to take:

  • Increase agents assigned to handle the EOL tickets.  This will help to address the number of open tickets
  • Make strides to alert the customer of the EOL.  This will help to address the 19% of customers who have not yet opened the ticket
  • For VIP customers, have an account manager reach out directly to the customer
  • Drill into the satisfaction score to determine whether low ticket satisfaction is due to qualms with the EOL or level of support
  • Engage the Support Manager to address tickets with greater than seven agent comments

Do you have recipes of you own that you’d like to share?  If so, submit them to our community forum and get free SWAG.  If you post something, be sure to mention in in this discussion so we can take a look.  Sharing is caring!

13 comments

  • 0

    Happy Monday, Everyone! 

    Part one is posted.  We look forward to your comments today.  Cheers :)

  • 0

    Part two is up!

  • 0

    I'm encountering an error trying to create the to track time in group metric: "Unexpected end of string. Expecting one of: ELEM_IDENTIFIER, FUNCNAME1, NEXT, OBJECT, STRING, -, RANKFUNCNAME, IF, PREVIOUS, COUNT, RUNFUNCNAME, NUMBER, IDENTIFIER, FUNCNAME2, CASE, FUNCNAME12, PERCENTILE, TEXT_ELEM, THIS, SELECT, (" anytime I try to add the Total time spent (min) to the custom metric equation. I also already use this metric to track total time spent on tickets is various other reports. What am I missing?

  • 0

    Hey Leigh,  Thanks for the comment!  Do you mind posting a screenshot of the metric you're building that's generating the error?  Unfortunately error messages in themselves aren't always the most helpful.  Also, is the Time Tracking metric you're using identical to the one in this recipe?  

  • 0

    Here is my Time Tracking Metric:

    And my error:

    I type in "SELECT (" and then when I choose "Total time spent (min)" under metrics and click "Add" I get the above error.

  • 0

    Oh!  Gotchya.  Thanks for the reproduction steps :)  You'll want to enter the entire metric into the metric editor before clicking 'Add.'  In order to select the Total time spent (min) metric, double click it then move on to the the next part of the argument: /60)*20 WHERE ...  

    Here's a link to a video from GoodData on making custom metrics.  I think you'll find it useful here.  Of course, let us know if you run into additional road bumps and be sure to let us know how the metric works out for your team.  Thanks again!  

  • 0

    Thanks, that's embarrassing. :) Guess it's been a while since I've messed with the custom metrics.

  • 0

    Part three is up!  Feel free to chime in with questions or comments.  Thanks!

  • 0

    @Leigh, not at all.  This stuff is confusing and easy to forget if you're not in it everyday (even then it's easy to forget ;))  Thanks for engaging with your Q.

  • 1

    Missy, finally had the time to sit down and read this thread. Good job!

    I recreated much of it but added two additional charts. Since I store the "plan" against the organization I was able to create a comparison of the Ticket volume by plan chart to show the percentage of customers in each plan. 2% of my customer base that I call "Platinum" create 11% of my tickets. Demanding!

    The second chart I added was the count of tickets for each type of ticket, "About". This added another dimension to the report in that not only did i know where I was spending the most amount of dollars but also could compare that to the number of tickets that made up that cost. I am thinking I may add a third chart here to show which type of ticket costs the most by doing some arithmetic on the two sets of data.

  • 0

    Hi Missy, I'm not sure if what we are trying to do falls under this fine tuning session or not, but I could use some help.

    We have several resellers/partners and have created several reports to show how many tickets have been submitted for each reseller/partners. Now i'm trying to create a report that shows % of tickets being submitted for each reseller/partners. In other words, out of 500 total tickets submitted last month, 10% were ANCILE direct. 

    Not sure how to go about creating this custom metric. I've tried reading thru this but not sure it applies to what i'm trying to do.

     

     

  • 0

    Hi Susie,

    Great question! Since I may need some more information in order to assist you, I'm opening up a ticket on your behalf. I'll see you there!

    Desiree

  • 0

    Hi Desiree,

    Thanks for replying. I've tried a few other things since this post which I'll tell you in the ticket.. see you there!

     

    Susie

Please sign in to leave a comment.