Identify High Volume Customer Complaints

While most organizations may have preponderance of data about the root-causes that result in customer support calls, they may not have this call-data classified by customer complaints as defined earlier.

This is how you can determine your high volume customer complaints.

  1. List complaint candidates: Make a list of 10-20 problems that are possible candidates for high volume customer complaints. If you are planning to create interactive guides for multiple products, you will need to create this list for each product.

    Please write them down on a spreadsheet – one sentence each – less than 10 words. Remember, a complaint is a customer’s answer to the Agent’s opening question “How can I help you?”

    You can create this list from following sources:

    • Gut-calls of subject matter experts, experienced agents & customer support management.
    • Reviewing the phrases customers frequently search for in your knowledgebase.
    • Review list of most frequently viewed knowledgebase articles.
    • Review the categories that result in most trouble-tickets in your CRM/Helpdesk system.

    Be very careful in crafting the words you use to describe these customer complaints. These words should resonate with the customers facing the problems & make them want to engage with the corresponding Yonyx Interactive Guides (when published) by clicking on these complaint descriptions.

  2. Validate problem candidates This is an extremely important step. It not only helps you determine which complaints dominate customer calls – but what percentage of total calls are due to these customer complaints. This latter helps you determine total available opportunity for serving customers through the Yonyx Interactive Guides you decide to create & publish.

    One great way to validate the problems you have listed as candidate high volume customer complaints is by surveying your customer support staff. Below, we describe how you can create a survey in literally 10 minutes & capture invaluable validation information from your team of front line support staff.

  3. STEPS TO CREATE A SURVEY
    1. Create an account (if you don’t have one already, its free) at www.surveymonkey.com
    2. Login and start creating a survey.
    3. Create a multiple-choice question. Your survey will consist of only one question (see below). Add this given question (use the language used here) and add the 10-20 candidate customer complaints you have listed so far as the answers to this question.
    4. Configure the survey type – so that a visitor can vote multiple times for each survey.
    5. Edit the survey.
    6. Click on Change Settings to enable multiple responses from a user.
    7. Choose the option to allow multiple responses from a user.

    8. Send the link to all Agents participating in the survey – use language similar to the one shown below announcing the survey.
  4. Invite Agents to take survey
    Feel free to use this language when inviting Agents to participate in the survey.
    Dear Friends:

    We are requesting you to participate in a survey by clicking on the link below. Unlike traditional one-time surveys, this one will be on-going for XX days (Cover enough days to have 1000 data points – e.g. if you have 50 members taking 20 calls per day, one day worth of survey would be enough, but if you have 5 members taking 30 calls each, you will want to run the survey for a week). This is a Single Question survey with multiple answers.

    [Replace with your survey link] http://www.surveymonkey.com/s/V5FRJHG

    How should I take this survey?
    At the end of each customer call, we want you to click this survey link & simply choose one of the answers that best describes the symptom or complaint (not the underlying root-cause) the customer had called about. Each time you vote by selecting a given answer, you will see the running results that reflects the combined inputs of all the team members taking this survey.

    How often should I vote?
    You should only vote once for each customer call. At the end of the next call, you should vote again by hitting the browser back button and choosing one of the answers again.

    Can I vote for all my calls at the end of the day?
    No. Please vote at the end of each call.

    What if none of the reasons given suits a given caller?
    If at times you find none of the reasons given reflect what the customer called about – please write a quick sentence and submit by choosing “Other” as your response in the survey for such calls. Please make sure all other options are unchecked.

    Why are we doing this?
    We’re conducting this survey to improve our customer service by prioritizing our efforts on the problems that affect most of our customers first.

    Who should take this survey?
    Only members of the level 1 support team should take this survey. If you are not a member of level 1 support, please do not take this survey.

    We really appreciate your support. If you have any further questions, please drop me an email.

    Signed …

  5. Collect Survey Results
    Now analyze and revise the complaint list if necessary. Re-run the survey if a revised list was necessary.

    Now that you have a list of high volume customer complaints – you know two important things:

    1. Which complaints dominate the reasons why customers call support?
    2. What percentage of all calls do these complaints represent?

    If a handful of customer complaints don’t represent nearly 50% of your front line (level 1) support call volume, you should re-look at your complaints list. It is possible, in this case you have created a list of very detailed problems or you have not captured the list of truly frequently recurring customer complaints in your list.

    Look at the comments from surveyed agents & consider adding new problem(s) to the list to replace some of less frequently occurring ones & re-run the survey.

  6. Now that you have a good list of high volume customer complaints, the next step is to create a Yonyx Interactive Guide for each complaint.

Share this article

Back to top