Copilot in Microsoft Clarity: Everything About Its AI Capability + Strategic Usage Tips

Microsoft Clarity has a feature named «Copilot in Clarity,» which leverages Artificial Intelligence (AI) to offer an intuitive conversational experience to understand behavioural analytics data. It does this by leveraging the GPT model to provide insights.

The Microsoft Clarity Copilot feature significantly reduces the time needed to analyse and understand how users’ interactions with your website impact business and metrics, enabling you to build helpful user experiences for customers.

This guide will introduce you to various functionalities that Microsoft Clarity Copilot can facilitate within your Clarity projects, including:

• How a single click helps generate session insights, providing summarised and concise takeaways from user session recordings.
• Quickly get actionable insights to help you understand user behaviours and trends on your website pages.

Future Copilot updates, as indicated by the Microsoft Clarity team, will allow users to:

• Interact with analytics data conversationally, receiving simple and concise responses.
• Utilise natural language to query the data of the Google Analytics property connected to your Clarity project to identify trends.

Additionally, I will outline some strategies to effectively use this feature, which will help you derive better insights. Throughout this article, we will also explore some limitations of the Microsoft Clarity Copilot feature, clarifying what is possible and what is not.

How to Enable Copilot in Your Microsoft Clarity Project

To activate and utilise Copilot’s AI capabilities in your Microsoft Clarity project, ensure that Microsoft Clarity is first installed on your website.

Next, verify that the Copilot feature, which is typically enabled by default, is enabled for your Microsoft Clarity project. 

You can check whether the Copilot setting for your project is enabled by navigating to the «Settings» view within Clarity and clicking on the «Copilot» tab.

In the «Copilot» view, you can confirm if «Copilot» is enabled and review whether the onboarding checklist for additional AI functionalities within Microsoft Clarity Copilot is marked.

If the Google Analytics item on the checklist appears unchecked, it’s simple: Your Microsoft Clarity project and your Google Analytics property are not linked.

For guidance on integrating these tools, refer to this article, where I covered everything you need to know about the natively and manually Microsoft Clarity and Google Analytics integration process.

Now, let’s look deeper at the possibilities of using Copilot to make sense of Microsoft Clarity data.

Here Are The Capabilities of the Copilot AI Feature in Microsoft Clarity

The Copilot AI feature in Microsoft Clarity enhances your ability to analyse and understand analytics data. 

Here are some of the ways to leverage the features of Copilot in your Microsoft Clarity project:

1. Session Insights: This function watches and summarises the information from session replay recordings and provides summarised insights.
2. Heatmaps Insights: This can provide critical observations and takeaways from heatmap behavioural data, helping you understand user engagement on your website.
3. Grouped Session Insights: This feature aggregates a group of session replays and summarises website session data, facilitating a deeper understanding of user journeys.
4. Clarity Chat: The proposed feature will let you interact with your Microsoft Clarity analytics data through conversational human language.
5. Google Analytics Chat: Allows for querying Google Analytics data using natural language, integrated natively within your Clarity project.

To understand each functionality’s potential and strategic applications within the Copilot feature of Microsoft Clarity, let’s explore each capability and discuss effective usage strategies.

Session Insights:

Session recordings are among my favourite features in Microsoft Clarity. They recreate user visits as videos, enabling a comprehensive understanding of the user journey across your website.

Clarity captures all visible page information (the DOM content) and user actions such as mouse movements, clicks, and scrolls, reconstructing these elements step-by-step in a video format that displays user interactions across different devices.

However, given the constraints of time resources, reviewing session recordings can be time-consuming, especially on high-traffic websites.

Beyond utilising filters and segments in Microsoft Clarity to hone in on specific recordings that provide needed insights, the session insights feature, powered by Copilot, offers a time-saving and productive alternative.

This AI-driven feature of Microsoft Clarity helps generate concise summaries of user journeys described in plain human language.

Using Session Insights In Microsoft Clarity:

You can access session insights in two ways within Microsoft Clarity:

1. View in Session list
2. View in More details

1. View in Session List:

Navigate to the «Session Recordings» view within your Microsoft Clarity project. In the «Recordings» section, look for sessions that meet the Microsoft Clarity session summarisation condition.

Click the «chat-like» icon labelled «Generate Session Insights» next to the desired session.

This action produces a summary of the user’s activities during that session.

You can provide feedback on the generated insights and copy the description for further analysis. I added more context to this later in the blog.

To close the generated session insights summary, click the «Generate session insights» icon again to toggle off the summary display.

2. View in More Details:

This functionality is also accessible in the «Recording» view. While watching a session recording, click the «More Details» option in the top bar or session card.

Doing that opens a new view where you navigate to the «Insights» tab. Copilot will then generate a summary and critical takeaways of the user session.

Although you can provide Clarity’s Copilot with feedback regarding the summary it generated, copying it is not permitted, as you would with the «View in Session List» approach.

Limitations of the Session Insights Feature of Copilot

• Sessions lasting less than or equal to five seconds do not qualify for Session Insights due to the short duration, which results in greyed-out, non-clickable options.
• The sharing and downloading of AI-generated session replay takeaways and insights are currently not supported.

Let’s now explore the capabilities of «grouped session insights.»

Grouped Sessions Insights:

Grouped Sessions Insights expand upon individual «Session Insights» capabilities by summarising multiple session recordings simultaneously. This feature allows you to generate insights for any group of session recordings filtered through Microsoft Clarity tools, including «Smart Events«, «Custom Tags» you’ve created, device types, UTM campaigns, and more.

Accessing Grouped Sessions Insights:

To use Grouped Sessions Insights, navigate to the «Recordings» view. In the top right corner, you will find a button labelled «Summarise Recordings,» which presents you with two options:

1. Summarise Top 10 Recordings:

This option allows for a quick summary of the top 10 recordings in the playlist view.

It uses the current order of the session recording list to determine which sessions «Copilot» will watch and include in the summary it generates.

Selecting this generates an immediate summary in the left panel, revealing insightful patterns and key takeaways from users’ actions during those visits.

To view the recordings referenced in each summary, click on the numbers at the end of each callout or open the «Summarised Recordings» module.

Clicking the «Summarised recordings (10)» will show you the sessions included in the summary insights.

2. Summarise Custom Selected Recordings:

This option enables you to choose specific recordings to summarise. Once selected, you can add session replays to your summary queue by clicking on them.

While you can select more than ten sessions, there is an upper limit beyond which no additional recordings can be added.

After selecting the desired sessions, click the «Generate» button, and the Microsoft Clarity Copilot AI will provide an insightful summary of user behaviour patterns and recommended actions.

Here is an example where I selected over 50 session recordings, which wouldn’t have been possible if I had chosen the «Summarise top 10 recordings.» 

Also, note that you can click on an already selected session to remove the session.

To exit the summary generation, click the «Exit Selection» button in the top right corner.

Limitations of Grouped Session Insights:

• It is not possible to summarise all project recordings at once.
• Sharing and downloading of insights is not supported.

Strategic Tips for Using «Session Insights» Capability of Copilot in Microsoft Clarity:

To ensure that the insights generated by Microsoft Clarity Copilot are valuable, consider the following strategic tips:

1. Use Filters: Apply filters to narrow down your data to specific segments of website visits.

2. Use Session Filter Controls: Utilise session filter controls to exclude sessions that may be too brief or irrelevant for generating meaningful insights. Consider filters such as:

1. Session duration: to exclude overly short visits.
2. Session click count: to focus on sessions where users are actively engaged.
3. Session page count: This includes sessions involving navigation across multiple pages.

3. Use Exclude Filter Conditions: Exclude sessions using «Custom Tags«, «Smart Events», etc., that you do not need to be included in the analysis Clarity’s Copilot should perform when generating insights.

4. Use User Actions Filters: Prioritize sessions where users performed noticeable actions, such as scrolling or cursor movement, which indicate active engagement rather than inactive sessions. 

By effectively applying these strategies, you can improve the insights the AI capability of the «Session Insights» and «Grouped Sessions Insights» features in Microsoft Clarity can generate, which helps you gain a deeper understanding and actionable insights from your website analytics.

Heatmap Insights:

Similar to the «Session Insight» feature, Microsoft Clarity’s Copilot’s Heatmap Insights capability enhances your analysis efficiency by summarising heatmap data.

Clarity’s Copilot, «Heatmap Insights,» provides behavioural insights and takeaways that will help you better understand user interactions and improve web page structure, content, and design.

To access «Heatmap Insights,» navigate to the «Heatmaps» report tab and select the page for which you want to analyse heatmap data.

You will find a «Summarise Heatmaps» button in the top right corner. Clicking this button activates the Generative AI, which quickly analyses and summarises insights across different Microsoft Clarity heatmap types for a specific page, covering all device types.

Upon clicking the button, you will instantly see a summarised view of all heatmap data for the selected page URL in the left panel.

Clicking the «Summarise Heatmaps» icon again will close the summary panel and toggle off the insights display.

The AI model’s summary includes anchor links that, when clicked, will direct you to the specific elements referenced in the web page «User Behaviors» summary insight.

The takeaway insights can sometimes be generic, and the AI may recommend making changes that are already present on the web page.

Limitations of Using Copilot in Heatmaps:

• Heatmap Insights are not supported for mobile apps, but according to the Microsoft Clarity team, it’s coming pretty soon, as mentioned in their announcement for Copilot on Clarity Mobile App Projects.
• The feature does not currently process data from Area heatmaps or Conversion heatmaps.
• Sharing and downloading of insights is not supported.

Strategic Tips for Effectively Using the «Heatmap Insights» Capability of Copilot in Microsoft Clarity:

To ensure that Microsoft Clarity’s Copilot provides you with valuable insights and actionable recommendations, here are some strategic tips to optimise your use of the Heatmap Insights feature:

1. Toggle Off URL Parameters:

You should ensure you’ve toggled off the «URL Parameters» option as default.

Doing this will help prevent Microsoft Clarity from treating pages with query parameters as separate web pages from the same web page without parameters.

This distinction is vital for maintaining a clear and consolidated page performance analysis. In exceptional cases where URL parameters might be needed to identify the actual page content, carefully consider when and how to enable them to ensure data relevance and accuracy.

2. Strategically Apply Filters When Needed:

The key to using filters effectively is to be strategic about their application.

If your goal is to measure the general effectiveness or perception of a particular page offering across all visitors who have seen it, avoid using filters so that you can truly get a holistic view of user interactions on that page. 

This approach allows Copilot in Clarity to give you a diverse understanding of how engaging, usable conversion is designed based on the insight generated.

Conversely, suppose you are interested in a more segmented analysis, such as understanding how a specific subset of users interacts with the offering page. In that case, you can selectively apply filters like «page» and «user actions».

There is a lot more that you can do. However, doing this will help you narrow down to page visits with meaningful interactions, enabling Copilot to generate more specific and actionable user behaviour summaries based on the filtered visits’ heatmap.

Clarity Chat:

This Microsoft Clarity feature will allow you to interact with your data using natural language, processed by Clarity’s GPT model. This will enable you to obtain simple and easy-to-understand information based on the queries you’ve supplied, its AI, and the data it has about your website.

Currently, the Copilot Chat widget is located in the «Dashboard» section at the top right. Clicking on the «Copilot logo icon» will open the Chat panel, where you can start querying the behavioural data collected by Microsoft Clarity.

When the panel opens, it offers three main functionalities:

• Ask Questions: Use pre-filled questions for quick answers to specific analytical queries.
• Summarise Your Data: Generate summaries of collected data using pre-filled prompts.
• Get Recommendations: Obtain insights on improving the digital experience.

You can type your question into the chat field or click on the Prompt Library icon to browse pre-made sample questions.

After entering your question, click the Send icon to submit it. Note that there is a 500-character limit, and it may take a few seconds for Copilot to generate a response.

You can adjust the chat panel size by hovering over the left side and using the mouse to resize the chat window.

When Copilot generates a response that includes filters, you can click to expand the filters, allowing you to see and validate the filters applied to the result generated by the AI.

In the expanded view, you can see the filter it applied.

You can also select the Dashboard icon to filter your dashboard or the Recordings icon to navigate to the Recordings tab with the same filters applied.

It’s important to note that Microsoft Clarity Chat uses Generative AI models to power Copilot features. These models can sometimes generate inaccurate information, which is why every response it generates includes the label «AI-generated content may be incorrect.»

You have the option to provide feedback using the thumbs up/down button and can copy the response to your clipboard by clicking the Clipboard button.

Understanding your business, how Microsoft Clarity works, and honing your prompting skills will be very helpful in using Clarity Chat productively. If you’re unsure of which prompts to use, you can click the view prompts option, represented by a book icon, and browse through the categories to see pre-made prompts.

Keep reading to access a prompt library that will enhance productivity for you, your team, your clients, or your business by leveraging Copilot’s AI capabilities in Clarity Chat.

Things to Note When Using Clarity Copilot Chat:

Copilot may not answer every query as it depends on the available dashboard data, but a little conversation with the Copilot chat widget reveals this.

Although it mentioned supporting Smart Events, my test reveals that it doesn’t support API events.

You can instruct Microsoft Clarity to format its answers in specific ways, provided the format is supported.

The chat feature allows you to provide a URL that you will need for specific answers to your analytical questions related to sessions, users, or behavioural questions, and it will generate relevant answers.

Limitations of Clarity’s Copilot Chat:

• Clarity Chat Copilot relies on data displayed on your dashboard and currently does not support sharing links to specific Recordings or Heatmaps.
• Once a chat session ends, you can only start a new chat and not return to previous chat threads.
• The chat feature does not support sharing or downloading insights.
• Certain data points, such as custom tags, API Event types of smart events, Microsoft Clarity Funnels data (although you could still structure your prompt in a funnel manner approach) and content insights data points.
• Reloading the Clarity project interface during a chat session will clear the chat.
• You cannot save prompts for later reuse, so keeping a document with frequently used prompts is advisable.
• Clarity Chat Copliot AI has a strong capability for summarising insights and providing answers to questions related to the data collected in your Microsoft Clarity project. However, when it comes to recommendations, they are not always tailored or specific to your business. I believe it needs more information and business context to offer personalised recommendations. Nevertheless, screen recording data points can still provide valuable tailored insights.
• You can’t ask it about a specific user even if you provided the Clarity user ID or custom ID.
• The current setup limits chat threads to a maximum of 10 questions. Once you reach this limit, you’ll need to click the «broom icon» to start a new chat session.

Here is proof that it doesn’t support «custom tags.» For example, when I asked a question about «tools and templates» pages, I used a «custom tag» with the key «pageType» and the value «tools & templates» to send this data to Microsoft Clarity.

It still went ahead and used the page URL in the filter.

Here’s how Clarity responded to my question about a user’s visit, even after I provided an ID.

Mistakes to Avoid When Using Clarity Copilot Chat:

• Always provide data context in your prompts to ensure accurate results and that Copilot is aligned with you. However, you can still see the date range that Copilot used when analysing the data for insights.
• Avoid asking multiple questions in one prompt; break them down into sub-queries. During my tests, I noticed that the Microsoft Clarity Chat AI attempts to reapply the filter on subsequent chats that reference the previous response it generated.
• You should always validate the filters applied by Clarity Copilot to the response it generates to ensure accurate insights.

For instance, when I asked Clarity Chat about the performance of blog posts, it applied a filter where the visited page contains «blog,» while the actual URL structure is «/post/.» Additionally, it failed to utilise the custom tag.

To help you get started, I’ve created a repository of prompts that you can use across various industries. This resource will be regularly updated to ensure you have access to the latest and most effective prompts. The product is still in its beta phase, and I look forward to seeing how it evolves in the future.

When using Clarity Chat, please note that after reaching 10 chat questions, you won’t be able to continue the thread. To start a new chat session, simply click the broom icon.

Google Analytics Chat:

Similarly, the «Google Analytics Chat» feature within Copilot mirrors the functionality of Clarity Chat by enabling users to chat with their Google Analytics (GA4) data in natural language.

You must integrate a Microsoft Clarity project with a GA4 property and activate Copilot for that project before you can use the «Google Analytics Chat.» Also, the feature hasn’t been rolled out yet.

Considerations for Using Copilot in Microsoft Clarity:

When generating summary insights, Copilot in Microsoft Clarity considers the website content, user interactions, and «Smart Events» implemented in your Clarity Project. For instance, consider a session where I remained on a single page but triggered various e-commerce-related «Smart Events.» This additional context significantly aids the AI in producing a more comprehensive session summary.

While Copilot showcases how AI can enhance analytical productivity, you should keep several points in mind:

1. Generative AI models power Copilot features and can sometimes produce incorrect or misleading information. You can help improve the system by providing feedback via the thumbs-up/down buttons.
2. Effective use of filters, Microsoft Clarity custom tags, and segments can significantly refine your focus on relevant session recordings and heatmap segments, simplifying the search for needed insights.
3. Innovative applications of session insights can be achieved by combining them with the right prompts and additional context in AI tools with internet connectivity.

Here is an example of how inaccurate the insights it generates can be.

Additionally, if your site uses consent banners and the Clarity tag activates before consent is granted, the privacy banner interactions might distort the insights provided. Conversely, if the Clarity script only runs after visitors grant consent, it will not record or include in the summary any actions performed on the site prior to consent being given.

And, if you have undertaken the additional task of configuring Clarity to collect data «compliantly» after a user declines consent, using the «Consent APIs» without using cookies, Clarity’s recordings will not be linked into multi-page sessions until the user’s consent is acknowledged. However, heat maps and other data types, excluding session recordings, will still be accessible.

The implementation of consent management in Clarity effectively recognises when an end user revokes cookie consent after initially accepting it. This prompts Clarity to cease writing cookies for that user.

Closing Remarks on Copilot in Microsoft Clarity

This guide should have given you a comprehensive understanding of Copilot in Microsoft Clarity, including its capabilities, activation process, limitations, and tips on using this feature better.

I will continue to update this article as Copilot in Microsoft Clarity evolves, and I welcome any feedback or questions via LinkedIn or the comment space below. Until then, happy copiloting your business towards greater Clarity.