Harmoni Discover lets you find stories in the data faster by profiling groups of interest and comparing them with others. By profiling a target group against a set of descriptor variables you can understand the unique characteristics of the group.
Discover presents a table that lets you see the descriptors that best describe your target group and the comparisons between your target versus other groups.
If you don't see the Discover drop zone in the analysis area, contact support@infotools.com to have the personalization key turned on.
In this article
1. What does Discover do?
Discover allows you to compare a primary group against other groups based on specific attributes, helping you identify key characteristics of your target group.
The Target Group
This is the specific group you want to understand. It could be users of your brand, people in a certain demographic, or customers in a specific market. All comparisons in the analysis are based on this target.
If the target group is your brand, Discover profiles the brand, determining the items that distinguish it most, when compared with all the other brands in the analysis. The profile (or descriptor variables) you use to compare the brands might be demographic, psychographic, behavioral, or some combination of all these.
The Descriptor Variables
These are the attributes you want to use to build the profile of your target group. You can select any combination of demographics (age, gender), psychographics (attitudes, values), or behaviors (purchase habits) from your data.
To start using Discover, first define a list of attributes to use for profiling.
Learn more about setting descriptor variables.
2. The Discover Table
With discover descriptors defined, Harmoni presents a table based on your selected target group highlighting the most relevant information.
The variables at the top of the table are those that most distinguish your chosen target group from the rest. It also sorts the groups across the table, for example, brands or markets, so that the most similar to your chosen target group are closest and the most different ones are furthest away.
3. How does Discover work?
Discover works by performing a series of statistical tests on each descriptor, comparing the cell value for each group with the value for those, not in the group (the rest). These values are compared using Bayesian Statistics to calculate the probability that the group value is greater (or lower) than the rest values.
The Discover table is organized to give you immediate insights:
First Column (Target Group)
- The target group is shown in the first column, next to the Total column.
- All comparisons are based on this group.
Ranked Descriptors (Rows)
- Descriptor values are ranked by their probabilities within the target group.
- Ranking compares the Target group to the Rest (all other groups).
- Descriptors for the target group that have the highest probability of being greater than the rest appear at the top of the table.
- Descriptors for the target group which have lower probabilities of being greater than the Rest appear at the bottom of the table.
- Average probability descriptors appear midway, but may be suppressed in larger tables with many descriptors.
Sorted Comparison Groups (Columns)
- The similarity is calculated using Robinson's agreement metric between the target and comparison columns.
- Discover displays the target group in the first column, with other groups sorted by similarity to the target group. For example, if your groups are brands, your target brand is in the first column (next to the Total column), and the brand next to it is the one with the most similar profile according to the descriptors down the left of the table. The brand furthest away from your brand is the least similar.
Now that you know a little more about how Discover works, have another look at the table above to see the USA as the target market and which descriptors rise to the top, differentiating USA from other markets. Based on the sort order of the columns, you can see which market is most similar (the closest column), through to the one that is least similar (furthest away).
Important Note
The Total column displays the average across all the groups and is used in the statistical testing. Each group is tested against the total.
The target group is compared to the rest of the groups (excluding the target) and is used to rank the descriptors and the similarity of the groups.
Where to from here?
Learn more about Discover: