Skip to main content
AI Comment Comparisons in Engagement Results

Quickly compare employee comments from top & bottom scoring cohorts with AI to uncover key similarities and differences in engagement surveys.

Jessie Walsh avatar
Written by Jessie Walsh
Updated over a week ago

Note: Our new AI Comment Comparisons feature is now available to all customers in Open Beta! We’re eager to hear what you think as your feedback will directly help shape and refine the experience. While it's not fully finalized yet, you can already start using the AI Comment Comparisons in your reports to discover new insights.

Got suggestions or feedback? We’d love to hear it! Just reply with "Ask a Person" in a support conversation, and you’ll be connected to a Product Support Specialist.

What are AI Comment Comparisons?


AI Comment Comparisons let you quickly compare employee comments from the top and bottom scoring cohorts based on your engagement index factor. This eliminates the need for manual analysis, helping you identify key themes and insights faster.

How is it Generated?


When you select a demographic, the demographic values are sorted by index factor score. Then, the comments from the top and bottom groups for the questions related to your selected index factor are compared. Finally, a comparison summary is generated for easy analysis.

What Surveys Are AI Comment Comparisons Available On?


While AI Comment Comparisons are best used for Attributed (Snapshot) Engagement surveys they are available for all survey templates (live or closed) under the Engagement section of the template library.

As survey results come in, admins can access this feature from the Admin Report > Heatmap View > apply a demographic filter to access.

How can report viewers access this feature?


Report viewers with access to a shared/filtered report that includes the following settings can access AI Comment Comparisons:

  • Advanced report type

  • AI Comment Comparisons enabled

  • Additional Report filters have been enabled.

Once a demographic filter is applied on the Heatmap page of the report, the AI comment comparison banner will appear for viewers.

Requirements for Generating Comparisons


To generate a comparison, each cohort needs to have at least 3 "meaningful" comments from 3 unique employees. A "meaningful" comment must meet the following criteria:

  • More than 2 words

  • Contains at least one topic tag (excluding "NO_COMMENT")

  • Includes a sentiment tag (where relevant)

Additionally, Comment Group Minimums (CGM) will be applied at the quartile level. Both cohorts need to meet the CGM requirement before a comparison summary is generated.

How to Use AI Comment Comparisons


AI Comment Comparisons help you compare the comments left by employees across your survey. It spotlights similarities and differences between comments from demographic groups scoring highest and lowest on the index factor (i.e., Engagement). The tool will help to surface inconsistencies in the comments about the employee experience, helping you pinpoint areas that you might want to prioritize for investigation.

Example: Your company just wrapped up its annual employee engagement survey. This year, the survey results include a new feature: an AI-driven comment comparison tool. You're tasked with identifying differences in the experience of the most engaged and least engaged employees.

While the overall Engagement score is moderate, the comments reveal stark differences between the most and least engaged employees. The AI comment comparison highlights:

  • The top cohort frequently praises flexible scheduling, team support, and recognition of personal needs.

  • The bottom cohort highlights long hours, workload issues, and unclear expectations regarding time off.

You include the AI comment comparison in a concise report for senior leaders. The stark contrast between the two groups provides interesting insights that warrant further investigation: are flexible policies and workload management contributing to higher engagement? You suggest specific follow-ups, such as reviewing workload distribution and reinforcing policies for work-life balance.

If they want a drill-down comparison of specific demographic values/groups eg. Finance vs Product, you’re prepared to investigate with them on the platform.

With the insights prepared, you feel confident presenting a clear narrative about what differentiates the experiences of engaged and disengaged employees. Senior leadership now has the actionable insights they need, and you’re ready to tackle the next challenge.

Best Practices for Reviewing Comments


While comments offer valuable context, they should be viewed as supplementary to quantitative results. Here’s how to use them effectively:

  1. Start with the numbers: Look at the quantitative data to get an overall understanding first.

  2. Dig into comments: Use comments to add depth, but remember that they may not always represent the majority opinion. Culture Amp's 2024 research that employees tend to comment more when they’re dissatisfied, and negative comments are usually longer than positive ones.

This means comments can skew perceptions, so balance them with your overall survey results.

Enabling and Sharing AI Comment Comparisons


Admins can include AI comment comparisons in filtered reports (also called shared reports). To do this:

To create a new filtered report with AI comment comparisons:

  1. Go to your Surveys > Share Reports page.

  2. Click Create a Filtered Report and select the “Advanced” report type.

  3. Enable AI comment comparisons, then enable any relevant filters.

  4. Click to Enable report filters and select any additional demographics to filter by.

  5. Follow the steps in our Report Sharing Guide to assign report viewers and share the report.

To add AI comment comparisons to an existing report:

  1. Go to your Surveys > Share Reports page.

  2. Locate the report, click the more dropdown next to the report name, and select Edit Configuration.

  3. Choose the “Advanced” report type.

  4. Enable AI comment comparisons

  5. Click to Enable report filters and select any additional demographics to filter by.

  6. Follow the steps in our Report Sharing Guide to assign report viewers and share the report.

How we are using machine learning and AI


We have partnered with Google to utilize their LLM foundational models in our AI products. This partnership gives us access to best in class technology that will power Culture Amp products to bring the market-leading HR products to our customers.

There are 3 foundational product pillars that we use in our AI product design:

1. Trust - Customers look to us for help with sensitive workplace topics. As such, we consider privacy, security, ethics, bias, and regulatory concerns in everything we do. Via thoughtful design, we'll disclose where and how AI is applied, attempt to explain results, and accept and respond to feedback.

2. Control - AI capabilities should never assume control, exclude people, or remove the agency of a person in the course of their work. Employee experience is inherently human and AI should support but never direct. Even the best available data and technology cannot replace the contextual awareness, specific experience, and collective impact of groups of people working together.

We always encourage you to explore your comments in detail and use the comparison as a guideline, to get the maximum value out of our AI comment comparisons feature.

3. Value - Product capabilities should always be clear on customer value, and that's just as true for the applications of AI. We ensure that our applications of AI have a clear benefit and are fit for purpose before unleashing them upon the world. Our AI technology aims to enhance the experience of users (people, in all their roles, from employee to manager to HR professional, to exec) by reducing unnecessary toil, surfacing insights, connections, and meaning, and augmenting people’s skills, ultimately allowing them to focus on high value, high impact activities.

Our AI comment comparisons feature has real power to reduce both time and cognitive load when interpreting qualitative data, providing substantial value to our users.

AI Security


Culture Amp prioritises customer data privacy and security throughout our AI capabilities in the same way we do across the rest of our platform. The privacy and security of your data is of the highest priority to us - and we apply the same strict protection measures with your employee feedback data as we do across our entire product offering.

Note: None of your employee data will be shared with Google Vertex. Our customer’s employee privacy and data security remains a top priority for us. No Culture Amp customer data will either be used or retained by Google to improve their models or services.

Responsible AI


Culture Amp's Responsible AI (RAI) means we continuously evaluate our existing and new practices to responsibly and ethically build technology, use data, and prepare for future regulation and customer needs.

FAQs


Are AI comment comparisons available in shared/filtered reports?

Yes, you can enable AI comment comparisons in leader and demographic-based shared reports.

Can AI comment comparisons be enabled for shared reports that have already been published?

Yes, admins can enable AI comment comparisons for reports that have already been published.

Are AI comment comparisons available when additional demographic filters are applied?

Yes, when you apply more demographic filters, it updates the top and bottom quartile groups, which changes the comments being compared. This helps identify new differences and similarities between groups.

Can AI comment comparisons be exported?

Currently, AI comment comparisons can’t be exported directly. However, you can copy/paste the content or use your browser’s print feature to save it as a PDF.

Which survey templates have AI comment comparisons available?

AI comment comparisons work best with Engagement/Snapshot surveys, but admins can use them across all survey templates in the Engagement section—whether live or closed.

Does the survey have to be launched by a certain date for this feature to be visible?

No, the AI comment comparison feature is available for both past and future surveys automatically.

How accurate are the AI-generated comparisons?

The AI-generated comparisons are generally quite accurate, but we recommend using them as a guideline. Since the comparisons are powered by Google’s language model, they’re based on approximations and may not always be perfect. It’s always a good idea to explore comments in detail as well.

What if the AI comment comparison is not what I expected?

If you notice something unexpected in the comparison, please share your feedback! You can do so by contacting your Customer Success Manager or by replying with "Ask a Person" in a support conversation to speak with a Product Support Specialist.

Will the AI comment comparison remain the same if I come back to it a few days later?

Yes, it will remain mostly the same as long as you haven’t updated any of the underlying comment sentiment or topic data.

If I change the sentiment or topics on comments, will it change the AI comment comparison?

Yes, any changes to comment sentiment or topics will update the AI comparison. The tool will regenerate a new summary based on the updated data.

Will any customer data be shared with Google Vertex?

No, customer data will not be shared with Google. We take your data privacy seriously, and no customer data is used or retained by Google to improve their AI models or services. Your data remains secure and protected.

Can AI comment comparisons be turned off at the account or survey level?

No, AI comment comparisons can’t be disabled at the overall account or survey level. However, you can turn them off for specific filtered reports. To do this, go to your Surveys > Sharing page, click the more option under the report name, and disable the AI comment comparisons option. Keep in mind, AI comment comparisons only work with Advanced report types that have both AI comment comparisons and additional filters enabled.

Can I provide feedback on the AI comment comparisons?

Yes! You can give the comparison summary a thumbs up or thumbs down.

Are AI comment comparisons included in an export?

Not right now, but you can use the copy button to add the comparison to your clipboard.


💬 Need help? Just reply with "Ask a Person" in a support conversation to speak with a Product Support Specialist.

Did this answer your question?