Behavior - Contrast View
Learn how to create, understand and compare using the contrast view
Last updated
Learn how to create, understand and compare using the contrast view
Last updated
Before we deep-dive in to understanding the report, it is important to understand what a contrast view is. In Consumr.ai there are several types of reports like Intent, Behavior, Feedback, Social and Market. We have dedicated instruction manuals and interactive videos for each of them that you can go through to understand more about each report. Reports exists across different modes in Consumr.ai, like Insight, Pre- Flight and In-flight. In this manual, we will share more information on Contrast View that exist in the Insight mode.
While creating an insight, we use inputs of several types like Interest, Keywords, Audiences, URLs etc. and create a report which is an independent insight of one or more inputs that you select in isolation or combination. They are called Individual Reports. Consumr.ai also provides a comparison view where you can compare 2 to 3 reports or inputs at the same time, that is nothing but a compare view.
A Contrast view can be initiated at the back of an existing Behavior report. While the team is working on solutions to allow comparisons across different report types, compare view is available only on Behavior reports.
To get started select an existing Individual behavior report or initiate a new one and locate the compare button on the report page as given in the screenshot above. To learn how to initiate a new Individual Behavior Report click here.
Clicking the compare button would mean that the report you are in, will become the base report for the comparison. In simple words, the next report you will be selecting will be compared with this report output. On clicking the compare button, you will get a pop up to select a report or audience to compare with.
1. Select the Contrast Button on top
2. If you have an existing report you want to compare it with, select the “Historical Reports” option.
3. If you want to run a new report, select the “Audience” option, and select and audience to run a report real time.
4. If you select the audience option, it will give you a search box to either select a 1st or 3rd Party audience.
In the screenshots above, you can see how the selection options differ between selecting historical reports and an audience.
Once you select your reports or audiences, you get to choose the value for the audience. The higher value audience is by default the base report during comparison.
This means that in the rank flow graph, all spaghettis use the order of interest from the base audience. This part will be explained in detail, when we discuss the rank flow charts ahead in the document.
Comparing reports ideally requires intensive data gathering, processing and calculation, hence it is ideally recommended to pick one audience to do a real time run and then compare them with the base report. However, you can run more than one audience at a time. This section helps you understand what to expect:
Historical Reports: You can compare a maximum of 3 reports together (including the base report). Since historical reports are already generated it reduces efforts significantly. The system in this case only must focus on creating a compare view. Hence, the output comes to you faster.
Real Time Audience: When you select an audience from the pop up, Consumr.ai must extract the data, clean, process and calculate it real time. Since data is gathered from various sources, there is a dependency on the sources and how fast they respond.
There are times when the sources fail to deliver data. In such cases, Consumr.ai must keep requesting for data multiple times before giving up. When you choose this option, we recommend you choose one audience and then compare it with the base report, if you need a quicker result.
However, you can also choose two audiences at once. In this case, Consumr.ai run 2 separate individual reports and then compares it. This increases the chances of rejection from the data sources. To counter these failures Consumr.ai intelligently passes the request for information through a well times queue, to avoid rejection and failures. Due to this work flow we now have almost 0 failures but requesting for 2 real time audience reports increases the processing time significantly. If you need your reports quickly, either choose 1 real time audience or create individual reports for all the audiences you need and then compare them using ‘Historical Reports’ option.
After you submit the inputs, you will be asked you rename and save the report and a pop up will appear to guide you through the time that it will take to create this report. Once the reports page appears, the data will be displayed on various cards grouped in sections, for you infer from.
As in the Individual Behavior Insight, (refer to the manual to learn more), this section is the first on the list of displays to help the user get the synopsis of the report.
However, instead of one Summary card, you will see two. One for each audience selected for comparison. If you look at the screenshot above you can see that while your idea audience that is designated as High value has a persona of a woman from New York, USA while the other audience has a persona of a Male from Bengaluru, India. The age groups are almost similar. While the first audience appears to be focused on self-care, the second audience is focused on relationships, home, and work. Two vastly different and contrasting traits.
This is the first graphical chart of the report. On the top horizontal axis are the audiences. The High value audience sits on the left. On the left vertical axis are the different age and gender ranges. You will find that the Rank flow charts are used across all comparison reports, so once you learn how to infer from it, all the reports become extremely easy to understand.
In the chart above you can learn that Female 25-34 ranks at the top of the high value audience but is number 3 in Low. On the other hand, Male 35-44 ranks # 1 in Low audience, but is on number 4 for High. We can also learn that in High value audience (as the persona card explained), top 2 segments are younger as compared to the Low value audiences. This assessment helps you understand a lot about how these audiences differ from a demography standpoint and guide you to make the right strategies.
This chart compares top cities for each report to learn the nuances between audiences at a geographical level. The above example compares audiences from two different countries. Therefore, this chart will not be available for such views, as the comparison will not have any conclusive inference.
However, to understand the graph is not overly complex as we have employed The Rank Flow chart to show the differences in cities. City data can be interpreted on basis of City vs city or regions vs regions (like New England and Southeast for US) or even based on coasts like East Coast and West. This data, when coupled with Themes and Interests, can help give you some very meaningful insights.
There are 2 levels to this visualization. At Level 1, are the themes that give you an idea of how Thematically your audiences differs. The first big shift is Console Games between High and low. The top 3 themes show us how both the audiences are different. The first audience is all about Games, Drinks, Fitness and Beauty. The second audience is all about Food, Fitness Home and Education. If you now connect these learnings with the age and gender chart, you can deduce that these changes in interest may be attributed to the age gap. As the audience matures their focus changes from Games and Drinks to Food, Home and Education.
There are some clear differentiators, however it is not wise to click on any of the themes to drill down. For example, drilling down in to Console Games won't give you any conclusive results since both audiences are poles apart when it comes to that Theme. Perhaps 'Drink' may show how these audiences are different from each other. Let us click on the dot of 'Drink' in the High Audience Column and we get the following result.
The differences adds great amount of character to both the personas. High Value & younger audience prefer Potent or casual drinks as compare to Low Value but Older audience that prefers expensive drinks like Wine and Coffee.
As described above, if you want all elements of an Advanced AI Behavior report analyzed, you have to ensure that you select a report that has Advanced AI enabled
Following the core principles of data, we strive to make every data point actionable. At the bottom of this report, you will find a button that will carry forward the data to Pre-flight mode from where you can create plans and ideas to make this data more actionable.
To conclude this manual, Contrast view is a compare mode between 2 to 3 audiences that helps you decode the nuances between the audiences and understand which aspects really make a difference. Try comparing your behavior reports using this mode to build out your strategies backed by real time data.
