Group customer questions by theme with Theme Analysis

The theme analysis feature groups customer questions into sections to help marketers save time when reviewing feedback via the alby platform.

First, you set the themes you’d like to split the questions into, then run an analysis of the questions your customers are asking. The questions will be sorted into those categories. For example, if your customers are frequently asking questions about customer support, you can create that theme to see all matching questions.

This grouping helps you see similar questions and may identify gaps in your outreach to customers. Theme analysis can turn hundreds of questions into structured insights about how your customers see your brand.

You can also see the conversion rate of customers asking questions related to that theme.

How to use

To use the theme analysis feature, click Theme Analysis in the alby dashboard. When setting up your theme analysis, create the themes to group questions in.

If you don’t know what themes to start with, you can analyze the data without configuring themes and alby will recommend options for you. 

Configure themes

1. Click Configure themes in the Theme Analysis page. The theme catalog dialog opens to show all your themes, empty if you’ve never made any before.

2. Add as many themes as you want to group customer questions. The more themes you add, the more your customer questions are getting split into different groups. These themes will depend on your brand type, but a few common themes may include:

   1. Compatibility

   2. Fit and sizing

   3. Shipping & Support

   4. Price and value

   5. Recommendations and comparisons

   6. Assembly, installation, and setup

   7. Battery life

   8. Subscription fees

   9. General product questions

   10. Other

3. Click Save. Your themes have been configured.

Run analysis

Once you’ve configured your themes, you can run an analysis of your customer questions.

1. Set time window for the analysis (defaults to the last seven days, maximum one month).

   1. Depending on how many questions your brand gets via alby, you may adjust the date range to include more or fewer questions. The further back you set the start date, the more questions will be included, but some may be outdated depending on any changes you’ve made to your site or products.

2. Add any product IDs (optional)

   1. This allows you to filter and narrow down the questions received just on those specific product pages. For example, if you’re wondering what customers are asking about a popular pair of skis, you could input that product ID to only show analysis for those skis.

   2. When adding product IDs, use the external item_group_id values, depending on how your catalog is structured. You can also validate the product IDs before running an analysis, to ensure the IDs were entered correctly.

3. Add a widget filter (optional)

   1. Only analyze questions customers asked on a specific alby widget.

4. Click Analyze. Depending on how large your set timeframe is, the analysis may take a few minutes. When it’s ready, the Results tab will populate.

Theme analysis results

When viewing the results of an analysis, you can see the total messages, total conversions, and conversion rate grouped by the themes you’ve created.

For each theme, you can click Latest threads to see the customer conversations related to that theme.

You can export the results of your analysis by clicking Export CSV.

Compare analyses

When viewing an analysis history, you can select two analyses and compare the results to see differences in timeframes using the same filters.