![logo_moinai-full-transparent-1.png]](https://helpcenter.moin.ai/hs-fs/hubfs/logo_moinai-full-transparent-1.png?height=40&name=logo_moinai-full-transparent-1.png){kind=logo}
Transparency and measurability show which answers from the AI chatbot work well and where there is room for improvement. For each editorial topic, you decide whether to measure success with a feedback question or a conversion goal. This makes it easy to see whether information was helpful or whether users completed the desired action. Alternatively, the CSAT rating captures satisfaction with the entire conversation.
Contents
1. Choose the right measurement method
2. Feedback question
2.1 How it works
2.2 Analyse results
3. Conversion goal
3.1 How it works
3.2 Analyse results
4. Change the measurement method
5. CSAT rating
5.1 How it works
5.2 Activate the CSAT rating
5.3 Customise the prompt message and label
5.4 Analyse results
1. Choose the right measurement method
For each editorial topic, decide whether to measure the quality of the answer with a feedback question or a conversion goal. As a rule of thumb:
| Type of request | Measurement method | What is measured |
|---|---|---|
| Informational request with information provided | Feedback question | Whether the information was helpful to users |
| Transactional request with information provided | Conversion goal | Whether users completed an action, e.g. clicked a link |
2. Feedback question
The feedback question suits informational answers that give users information. It shows directly whether the answer helped.
2.1 How it works
A feedback question with Quick Replies appears at the end of the answer, for example “Did this answer resolve your query?” with the Quick Replies Yes and No. With one click, users indicate whether they are satisfied with the content of the answer.
2.2 Analyse results
In the Statistics area, Positive / Negative Feedback shows the total amount of positive and negative feedback.
Top Intents and the related focus report show the ratings per topic: positive in green, negative in red. Hovering over a bar displays the exact number. For more details, see Statistics and focus reports.
3. Conversion goal
The conversion goal suits topics that aim to lead users to a further action, for example visiting a landing page via a URL button in the answer.
3.1 How it works
The goal the answer aims to achieve is tracked as a conversion. The following count as a goal:
- URL buttons: Outgoing links in the last dialogue step automatically count as the goal of the answer.
- Follow-up questions: A follow-up question can also count as a conversion, for example a redirect to a form.
Important: If an answer ends with a follow-up, only the follow-up counts as the goal, not the URL buttons in the answer.
3.2 Analyse results
In the Statistics area, Conversions shows how many users successfully converted via a topic in the selected period.
Top Intents and the related focus report show the share of completed conversions per topic. Hovering over a bar displays the exact number.
Recommendation: Conversion tracking combines all URL buttons in an answer into one goal. To track them separately, add a ref parameter to each link, for example ?ref=chatbot-pricing. An analytics tool such as Google Analytics then shows the clicks per link.
4. Change the measurement method
The measurement method is set individually for each editorial topic. The Feedback question is selected by default. AI agents always use the feedback question.
Important: When the CSAT rating is active, the star rating applies to all topics and AI agents. Feedback questions and conversion goals are then not available as separate options.
- In the Hub, open the Topics menu.
- Select the topic.
- Open the three-dot menu.
- Under Success measurement, select Feedback question or Conversion goal.
Note: Selecting Conversion goal automatically removes the feedback question from the answer editor. Instead, the happy path is activated and plays out once the goal is reached.
5. CSAT rating
5.1 How it works
CSAT stands for Customer Satisfaction Score and shows how satisfied users are with a conversation overall. Feedback questions and conversion goals measure the quality of individual topics. The CSAT rating, by contrast, covers the entire conversation with the AI chatbot.
The rating bar appears at the top of the chat after every answer from the AI chatbot.
Users give a rating from one to five stars. They can change their rating at any time during the conversation, with each new rating replacing the previous one. As a result, each conversation always has exactly one final rating. The current rating appears next to the AI chatbot’s name in the header.
5.2 Activate the CSAT rating
You can switch the CSAT rating on and off in the Hub under Conversation management > Rating.
Note: When the CSAT rating is active, the star rating applies to all topics and AI agents. For more details, see Change the measurement method.
5.3 Customise the prompt message and label
Prompt message (optional)
An additional message can point users to the rating bar. Under Conversation management > Rating, click the plus icon to add this message. The text is fully customisable, for example: “Please rate our conversation so far at the top of the chat.”
Label next to the stars
The text next to the stars can be customised, for example to match the tone of your website. The Customer Success Team can make this change for you. Possible labels:
- “Rate this conversation”
- “Rate this chat”
- “How satisfied were you with the conversation?”
5.4 Analyse results
CSAT results appear in the Hub under Statistics. For more details, see Statistics and focus reports.
