.png)
Prioritize - Key driver analysis
With the key driver analysis we try to pinpoint what really matters to your customers and what can be linked to their score, namely what is related to people recommending your company (NPS), your customer satisfaction (CSAT) and customer effort score (CES). We give you a prioritization of which issues you should tackle first: the ones your customers are unhappy about and have an impact on your metric (NPS / CSAT / CES) score.
PAID MODULE:
This feature is not available for all customers. Contact your customer success manager or support@hellocustomer.com. if you're interested in using this.
IN THIS ARTICLE
1. How to get there
Option 1:
From the home screen of your Hello Customer account, click on the Prioritize icon
- In case of tiles: below the touchpoint name:
- In case of list: next to the touchpoint name
You will automatically end up on the Key driver analysis page of the touchpoint of your choice.
Option 2:
Step 1: On the home page of your Hello Customer account, click Prioritize on the left side:
Step 2: Choose the right touchpoint:
2.The key driver analysis explained
Do you want to know exactly what you need to work on? Don't look any further, the key driver analysis will tell you. In the graph, there is a clear overview of all the topics people are talking about and how they relate to the score these people gave on the metric question. In a blink of an eye, you will see on which topics you need to focus first. The table below is perfect for data lovers! Here you can find all the numbers in detail. On top of the page, you will find a summary of the score of the selected metric.
IMPORTANT
Keep in mind that only the category paths with at least 1% of the mentions are included in the analysis. So if you can not find a certain category path that you spotted on the analysis page, it is probably due to this limitation.
A. Main elements of the key driver analysis
To build a key driver analysis, you first need to select:
- For which metric you want to make a key driver analysis. The metrics you will be able to choose are dependent on the touchpoint you selected in the left navigation bar. In case there is more than 1 metric question in this touchpoint, this will be a dropdown, otherwise it's a preselection.
- For which touchpoints
- You will only be able to select touchpoints that contain the metric you selected
- Please note that the touchpoint you selected in the left navigation bar will always be included in the analysis
- The date range you can select on top of the page is dependent on the touchpoint in the left navigation bar. Therefore, it's best to start from the touchpoint with the oldest start date
IMPORTANT
Only administrators and root team users (= users on the highest team level) will be able to select multiple touchpoints. It's also important to note that root team users will only be able to:
... make an export if they have export rights for all the selected touchpoints
... refine the results if they have metadata rights for all the selected touchpoints
... open the side panel if they have at least conversation view access for all selected touchpoints
The key driver analysis is based on this main elements:
- The average sentiment of all mentions of the specific topics. This ranges from -2 (very negative) to +2 (very positive).
- The correlation (impact) between the sentiment of the topics and the score on the metric question. This ranges between 0 (the expressed sentiment is totally unrelated to the assigned score) and 1 (positive remarks about the topic are accompanied by higher scores and vice versa)
- The amount of mentions which represents the total amount of people talking about this subject.
In the graph, the average sentiment is shown on the X-axis and the correlation on the Y-axis. This leads to 4 different levels of priority you should give to the topics:
- Fix this, now! - These are the topics you should focus on first to improve your NPS, CSAT or CES.
- Promote - These are the strong points of your company. Keep them top of mind with your customers to remind them what you're good at.
- Keep in mind - Do not focus too much on these topics. Your customers are not really happy about them, but overall, they do not really affect their satisfaction
- Amplify - These are things that customers take for granted. Yes, they are happy about it, but they only think it's normal. It does not improve their overall satisfaction.
The size of the bubbles in the graph represent the amount of mentions. When hovering over a bubble, you’ll see more information about this specific category path such as average sentiment, correlation and amount of mentions. You will also see the potential gain for this category. Learn more about this in the next section!
In the table, all topics are listed with all details about amount of mentions, correlation, average sentiment and the potential gain. Learn more about the potential gain in the next section. Next to each topic, the quadrant of the category path is specified so it is immediately clear which priority you should give to this topic. It is possible to sort on each column by clicking on the column title. It is also possible to make an export of this data to use it for internal reporting.
B. The potential gain
The potential gain shows how much your selected metric can improve by turning unhappy customers into happy ones. There are a few steps in calculating the potential gain per category path:
- First of all, the difference in NPS / CSAT / CES between people who are (very) negative about a topic and people who are (very) positive about a topic is calculated.
- Next, this difference is made relative to the number of mentions for that category. The more negative mentions, the higher the relative difference, because there are more people who can be converted into happy people.
- Finally, the total potential gain is divided between all topics. The total potential gain is the total possible increase in metric. This is different depending on the metric.
- NPS: 100 - current NPS. This can range between 0 (if the current NPS is 100) and 200 (if the current NPS is -100).
- CSAT: 100 - current CSAT. This can range between 0 (if the current CSAT is 100) and 100 (if the current CSAT is 0)
- CES: 7 - current CES. This can range between 0 (if the current CES is 7) and 7 (if the current CES is 0)
C. The raw feedback and automated summaries
When clicking on a category path in the table or in the legend of the graph, a side panel opens where you can read all the open text feedback linked to the selected category path. You can also use the blue text bubble to open the side panel.
In case you do not have the time to read all the open text feedback, you can generate a summary. This will give you a summary of all the open text feedback linked to the specific category path for which you opened the side panel. Please note that there is a limit of 100 summaries per month across all the users within your environment.
3. Filtering the key driver analysis
If you want to see the key driver analysis for a certain group, you can make use of the filters. The filter panel has several sections:
- Category filters
- Refine results
- Filter sets
If you selected all your filters, or if you do not wish to filter, you can collapse the filter panel by clicking on the filter icon to make more space for the widgets on the page.
A. Category filters
With the category filters, you can choose which topics you want to show in your key driver analysis based on different criteria
- Amount of mentions: if you want to exclude topics that are not mentioned a lot, or if you need to focus on topics not frequently discussed, you can use the slider to adjust the number of mentions
- Categories: possibility to filter out one or a few specific topics. This can come in handy if you want to take a look at for example internal KPI's on which you want to follow-up. If you made Isaac category groups, these will also be available in this dropdown. Read more about this here
B. Refine results
When filtering the analysis, again only the category paths with at least 1% of the mentions are included in the analysis. It is possible that certain category paths (dis)appear compared to the unfiltered key driver analysis, since this 1% is always calculated based on the totals of the current selection of metadata values and on the selected date range.
NOTE:
To get the most out of the Smart filtering, it is necessary to make your metadata smart in the participant attribute mapper. There you can define what data types your metadata are. Learn more about it in this article.
However, do not worry. If you do not map your metadata keys in the participant attribute mapper, they will be shown as text in the smart filtering as described below.
1/ Boolean
Booleans are keys with values that are either true or false. Most of the time, they will represent an answer on a yes or no question. Examples are:
- First time right = Answer on the question "Did we solve the problem in our customer service center with one e-mail / phone call?"
- Bought online = Answer on the question "Did this customer buy something online?"
- Car insurance = Answer on the question "Does this customer have a car insurance?"
In the participant attribute mapper, you can define which values represent the yes answer (true) and which ones no (false). In the Smart filtering, you will be able to filter on yes or no. If there are several values that represent yes / true, they will be combined into one option to filter on.
2/ Number
Numbers are keys with numeric values. These can be numbers with or without decimals. Examples are:
- Age = The age of a person
- Pieces = Amount of items purchased
- Years as customer = The amount of years the customer visits your shop / has an insurance at your company / ...
Numbers are shown in the smart filtering as sliders to make it easy to select a certain range. You can also fill out the minimum and maximum value you want to select. It is possible to fill out the same number as min and max to filter on one value.
3/ Currency
Participant attributes with data type currency, are keys with values that refer to amounts in either €, $ or £. Examples are:
- Purchase amount = The ticket value of the last purchase
- Contract value = The value of for example an insurance contract
Just like for numbers, for currencies sliders are foreseen in the smart filtering to make it easy to select a certain range. Again, you can filter out one value by filling out the same number in the minimum and maximum box.
4/ Date
Participant attributes with data type date, are keys with values that refer to a certain date on which something happened. Examples are:
- Birth date = The date when someone was born
- Date visit = The date the respondent visited your shop, agency or restaurant
Dates can be filtered by using a date picker. It is also possible to fill out the start and end date manually.
5/ Text
Text can be anything that is not a number, nor a currency, boolean, date or time. Examples are:
- Employee = The person who helped the respondent on the phone, in the shop, ...
- Type of customer = The type of customer the respondent is, for example B2C or B2B
Also for text, we improved the filtering. All values are shown in a scrollable list and you can use the search bar to look for certain values.
6/ Team filter
For the metadata key that is set as teams filter in your touchpoint settings, we have something special. In smart filtering, we allow you to filter on the whole team structure that is set up in the settings.
Example:
- Let's assume this is the team structure set up in the settings. The teams on the lowest levels are the values you send along with your respondents to link them to the correct team:
- In the smart filtering, this will look like this for admins and people part of the root team (highest team):
Of course, all users only see the part of the team structure they have access to. If a certain user only has access to one team on the lowest level of the structure, no team filter will be displayed in the smart filtering.
7/ Special cases
For the standard attribute Gender, we will always display three possible values in the smart filtering:
- Female
- Male
- Other
In the participant attribute mapper, you can define which values represent female, male and other in your specific key.
C. Filter sets
To make it more convenient to filter on certain combinations of values, we have the option to work with filter sets. When you are filtering on several values from either the same or different participant attributes, you can save this combination as a filter set.
The values you added to the filter set will all be automatically selected when you click on it again. Simply click the bin next to the filter set's name to remove it from your list if it is no longer useful. All of the tabs on the analysis page can utilize the same filter sets.
NOTE:
Filter sets are user specific, so the only one who will be able to use and see the filter sets is the user who created it.