Case 3. Analysis of conversation scripts with the help of Power BI.
Hi, this is Alexey Vertkov from the company "numbers". Let me introduce you a report that helps analyze conversation scripts.

Every company has a sales department that must:

  • Sell, and as expensive as possible.
  • Try to increase conversions by affecting overall metrics.
  • Retain customers as new customers always cost more.
  • Increase the main source of profits growth.
Any sales department, always or in most cases, uses a conversation script - a pre-built logical chain of answers, depending on the situation. A good manager follows this chain, but a bad one does not.

To understand how a manager works, a 100-point call scoring system is usually introduced. Initially, the quality assurance department listens to conversations of every manager. Then it assesses each question. The core of the subject is that the employee of the department must follow the rules: introduce himself, warn that the conversation is recorded, must not stumble, must not be rude, conduct a conversation using the script (predefined rules and conversational tactics), tell the date of the next call, and so on.

If there are a lot of screw-ups, then it may be not the employee, but eventually the system to be blamed. Therefore, evaluating any situation the quality department begins to select various motivational or other strategies, such as training, course implementation, or bonuses.

A high percentage of "no success" on a call or incorrectly closed objections requires a detailed study. Existing data in companies do not give the desired result, as they discover only superficial problems. A point is created at which the process freezes.

Being a classical case, it is for this that the company applied for the implementation of analytics that could clearly represent the work in the heart of the sales department.

The task:

The scheme of work of the company, that needed this report, is almost identical to the described above. In this scheme, we needed to adjust several points that affect the reliability of all possible data:

  1. The calls of each of the managers were listened to by a person from the quality assurance on our side. The number of calls for each of the managers comprised from 1 to 5.
  2. We decided on the list of verification questions for the existing script. The total of questions is 32, 1 check.
  3. We chose the method to evaluate every call. There are two grading systems: binary, where a fixed number of points for each question are given, and weighted, where key points can be obtained from certain questions that need to be emphasized. We agreed on the binary system.
Then we fine-tuned the CRM-system. Namely, it was decided to transfer the items for putting marks in the form of additional fields in AMO, since all data was stored and maintained in tables, which was extremely inconvenient. We duplicated questions from the scripts to the deal profile, where the manager listens to the call and puts down yes / no, filling out this profile. Accordingly, you needed to configure Power BI, analyze the information and make the summary of data.

We found it necessary to include the following pages into the report:

  1. Analysis of key issues.
  2. Report with comparison of results by questions and managers in the view of two different dates.
  3. Control of the execution of the required number of checks by the quality assurance department itself, for the relevance of the analytics results.
Tools:

  • amoCRM (it is also possible to use Bitrix24) – tracking system.

  • MS Power BI – visualization instrument.

  • myBI Connect– information retrieval and collection.



Page 1.
Here we see the data broken down by calls:

  • Dynamics of average marks.
  • When the mark was put.
  • The number of lead for each manager. It is possible to set the volume of leads in order to see the relevant assessment, and the system decomposes this data and shows the result that needs to be achieved.
  • Number of leads per month.
  • Employee, responsible for the lead.
  • Average mark.
Issues by groups and managers represent information that is of particular interest in this report.

This table displays all questions that were rated with the 0 mark by the quality assurance department. That is, the employee did not use it, did not sell it, did not say, and so on. For each of the questions, the number of errors is summed up, and the system decomposes them, showing the most problematic zones by groups.

Here we also observe: problems by groups, by managers, the dynamics of ratings, the number of ratings/marks, the number of evaluated leads, the average rating by leads.

There is a table in the middle of the page that distributes leads into groups:

  • Up to 65 points.
  • From 65 to 79 points.
  • From 80 to 100 points.
This data will help you look at the work from a different point of view, that is:

If the calls are made flawlessly, then the employee falls into the section from 80 to 100. Here we observe that the average check for payment is 50% more than the previous one. If the call is bad, then the average bill is doubled, and the rest of the data is also increased accordingly. We conclude: here we put down the KPI for employees.




Page 2
This worksheet compares results for the week and includes the data:

  • By managers.
  • By the group.
  • By each question.
The dashboard is particularly useful for the quality assurance department because it is possible to compare the previous week and the present one to catch different indicators on issues in order to further regulate or encourage this. For example, to make a conclusion that training is required, or that the process is slowed down by a certain manager, or something else. All innovations can also be compared in the next week in order to understand whether such a solution is relevant. It is a perfectly tuned tool for correcting errors and identifying employee weaknesses. Data are presented as a percentage, and by delta.


        Page 3
        It is intended to control the quality assurance department. KPI by the volume of checks. It helps to find gaps when checking calls.

        When setting a KPI for the volume of calls heard, we see:

        • Planed/actual of calls heard.
        • Planed/actual in the average cut by manager.
        • Planed/actual for the month for the entire sales department.
        We added the list showing the number of checks performed for each manager; derived an average rating, having distributed by dates.

        In a separate table, there is a check for filling in the fields when listening to a conversation using a script, which is especially important for correct information.

        To the right we embedded a link for quick transition to amoCRM, to the deal profile, to verify the data.

        Conclusions

        Companies do not understand where to look for issues and gaps in the sales department. Negative reviews start falling in. Managers do not achieve their goals. And then they come to the conclusion that existing situation does not meet their needs. And, step by step, they come to the implementation of the dashboards.

        In this report everything is automated except for listening to calls. Unfortunately, there is no such mechanism that can work with semantic loads. Therefore, the quality assurance department practically plays the key role here, and the dashboard provides the most useful information:

        1
        The manager sees how employees work.
        Just to know that the employee is in his workplace is one thing. Another is to see all his work in detail, evaluate it, look for weaknesses, find and work on them, subsequently improving the overall performance.

        2
        Assessment by calls, groups and managers.
        Assessment by calls, groups and managers.
        3
        Comparing time periods by managers or by groups
        Every new week gives a new result. To understand which of the results was better, just click a couple of buttons in the report. Based on the comparisons, you can draw a lot of useful conclusions about the scripts.

        For example, if a manager falls into the group from 80 to 100, and the average check grows, then an atmosphere of motivation and healthy competition is created for him. And if you add the other two groups into 80-100, then the revenue will increase by 75%:

        • Deal conversion gets better.
        • Deal cycle also gets better.
        • Average check grows.
        Here is a direct road to setting KPI by calls.

        Fill out our brief. And we will:

        • Study the issue.
        • Implement the solution.
        • Show figures.
        I also want to
        Our partners
        Our company is the official partner of Microsoft Power BI. All models are built on the basis of automatic downloading and modeling of connections of the MyBi Connect service. The main specialization of the company is the amoCRM CRM system.
        Find out your real numbers
        One click to easily understandable business analytics
        By submitting the application, you give your consent to the processing of personal data and agree to the privacy policy.
        Made on
        Tilda