Table of Contents

Using CaliberMind Answers to Understand Funnel Impact

Angie C. Updated by Angie C.

Most businesses organize their sales and marketing practices around the idea of a funnel. The general hope is that more leads coming into the top will lead to more customers coming out of the bottom. Monitoring and understanding the various parts of a company's funnel is an essential practice for sales and marketing professionals to understand what worked, what didn’t, and how things are going right now.

The most common question CaliberMind Funnel users ask is “What things are having an impact on getting potential customers to move through my funnel?” This question touches on several potential areas of interest including specific campaigns, channels, industries, and more. Traditionally, getting insight into this question requires wading through multiple dense tables of data and supplementary analysis. 

With CaliberMind Answers, we aim to dramatically simplify this process by providing a flexible and interpretable way to highlight exactly what is and is not working to drive journeys through the various stages of your funnel.

How Do I Set This Up?

The Funnel Impact Answer within CaliberMind is designed with ease of use and flexibility in mind. With a few simple configuration steps, you can get an idea of what tactics and characteristics impact success in your funnel the most.

Below we detail the various configuration options available to users.

(1) Name

This field allows users to give their answer a unique name to identify them across CaliberMind quickly.

(2) Funnel

The name of the funnel the user wants to learn about. These names are defined in the underlying Funnel configuration.

(3) Stage

The name of the funnel stage on which you wish to measure impact. For example, setting this value to “Marketing Qualified Account” would be primarily concerned with what impacts the most reaching Marketing Qualified Account. 

Currently we support stage selection for all stages save for the first stage of a funnel. This is due to the interstage nature of studying events once a journey has begun. Future enhancements will include the ability to analyze the funnel entry stage as well.

(4) Time Scenario

This parameter allows users to define the most appropriate period for analyzing their funnel. Long time ranges, such as This and Last Year, are useful for understanding broad historical trends. In contrast, shorter time ranges, such as Past 90 Days, are more appropriate for analyzing the impact of recent decisions.

(5) Metric

This parameter allows users to control what metrics they use to measure impact. 

  • # of Successes: This is a count of all journeys that exhibited the feature of interest that reached a given stage. This measure is useful when a user is concerned primarily with funnel volume.
  • % of All Successes: This is the proportion of all journeys that reached a particular stage that exhibited the feature of interest. This measure is useful when users are interested how prevalent certain attributes are at given stages.
  • Probability of Success: This is a measure of how likely journeys are to reach a certain stage given that they are tied to a specific feature. This measure is useful for understanding the likelihood of moving through the funnel given the presence of certain tactics or characteristics.
  • Impact on Odds: This is a measure of how the presence of a feature affects reaching a certain stage compared to examples where that feature is missing. This measure is useful for understanding the scale of impact in relation to baseline activity. Specifically, this is the ratio of odds of success for cases with our feature to the odds of success for cases without our feature.
These values can typically be thought of as “Reaching MQA is 4 times more likely when this feature is present”


We are interested in understanding how various campaign types impact reaching the Marketing Qualified Account (MQA) stage of my funnel. Over the past quarter, we’ve observed the following:

  • 200 journeys reaching the MQA stage.
  • Of those 200 journeys, 42 of them could be tied to Content Download.
  • Overall, we saw 100 journeys tied to Content Download. 42 of these reached MQA with the other 58 failing to reach this stage.
  • Overall, we saw 500 journeys not connected to Content Download. 158 of these reached MQA while the other 342 failed to reach the stage.

With these values, we can easily calculate the metrics for Content Download as described above.

  • # of Successes = 42
  • % of All Successes = 42/200 = 21%
  • Probability of Success = 42/100 = 42%
  • Impact on Odds = (42/58) / (158/342) = 0.7 / 0.46 = approximately 1.5.

These metrics can be interpreted as saying we have a 42% chance of reaching MQA when we observe Content Download. This leads to an approximate 1.5 x improvement on the odds of reaching MQA compared to our baselines.

The current recommendation is to use Probability of Success as the initial metric when exploring funnel impact.

(6) Impact Range

This parameter controls how we connect various events to funnel journeys. Currently there are two options available.

  • Stage-to-Stage: This option only looks at what occurs in the directly preceding stage when figuring out stage-level impact. This is useful when users care about understanding how timing of certain tactics could help move journeys forward.
  • Cumulative: This option looks at all events from initial funnel entry up to the stage of interest. This is useful when users care less about timing but more about the composition of successful journeys as a whole.


We are trying to understand how various channels impact reaching the stage 4 of our funnel. The Stage-to-Stage option would only look at the impact of channels immediately preceding stage 4. If there are no backfills, this would specifically be events occurring between stage 3 and stage 4. The Cumulative option would look at the presence of channel interactions between the start of the funnel and stage 4.The current recommendation is to use Cumulative as the default impact range to avoid potential issues surrounding backfilled stages.

(7) Feature Type

This parameter controls what is set as the event-level characteristic of interest. Depending on the selection, users are able to explore the impact of marketing tactics, industry dynamics, technographics, and more.

  • Campaign: The name of specific campaigns set up by the user.
  • Campaign Type: A classification used to group similar campaigns together. This is dependent on the user’s underlying data. Examples could include things like Paid Social, Organic Google, Content Download, etc.
  • Channel: The channel that received credit for driving a particular interaction. Examples could include Direct Mail, Paid Search, etc.
  • Department: The department associated with people who are interacting with a specific journey. Examples could include Marketting, IT, Sales, Finance, etc.
  • Industry: The industry associated with companies interacting with a particular journey. Examples include Advertising, Industrial Equipment, Finance, etc.
  • Job Level: A standardized measure of the seniority associated with people who are interacting with a specific journey. Examples include VP, Functional, CXO, Director, etc.
  • Title: The specific title of people interacting with a given journey. Examples include Senior Director of Operations, Vice President of Demand Generation, Chief Marketing Officer, etc.

(8) Maximum Number of Features

This parameter controls the maximum number of features returned for stage-specific impact assessment in the resulting exploration visualization. Rankings are determined by the impact metric for the specific stage that has been configured.

How Do I Use This?

Once you have set up your Answer, you will have access to a curated exploration page focused solely on your version of the question at hand. This is in addition to the generated insight that is delivered to you. This page provides both high-level and deeper insight into your data. The screenshot and breakdown below provide supplementary information about using and interpreting this information.

(1) Answer Name and Configuration Panel

This section provides the current configuration for this Answer. Often times this information is useful in differentiating use cases and providing additional context for discussion.

(2) Answer

Rather than just give users a multitude of reports to wade through, CaliberMind aims to provide a curated experience for understanding and interpreting the data at hand. This section provides a high-level answer to the core underlying question being considered. This statement is limited to 1-2 lines of concise information. This is the same value that is listed on the My Answers page for this specific question.

(3) Feature By Stage Heatmap

This heatmap shows the metric of interest for each combination of feature and funnel stage. This allows users to easily discern which characteristics are important for each stage and compare them across their funnel. Some key notes regarding the usage and interpretation of this graphic:

  • Values within a given stage column refer to the impact on arriving into that stage. For example, a value of 42 total successes in the Marketing Qualified Account (MQA) column is referring to the number of journeys with a given feature that made it to the MQA stage.
  • As mentioned in the configuration section, we currently support analyzing all funnel stages except the initial entry stage. This is due to the interstage nature of studying events once a journey has begun. Future enhancements will include the ability to analyze the funnel entry stage as well.
  • The features included on the vertical axis are the top features ranked according to their impact metric for the configured stage of interest. 
    • Ex: If Sales Accepted Opportunity (SAO) is selected as the stage, Channel is selected as the feature, and 5 is selected as the maximum number of features, the vertical axis would be the top 5 channels by impact to SAO
  • Each cell is classified into one of the following categories based on comparisons within a given stage.
    • Normal Impact: The metric falls within the normal range for a given stage.
    • Above Normal Impact: The metric is significantly higher than the normal range for a given stage.
    • Below Normal Impact: The metric is significantly lower than the normal range for a given stage.
    • Negative Impact: The metric indicates that the feature in question has a negative impact on the stage in question.
    • Not Enough Information Available: There isn’t enough information available to classify the metric in question.

Note that Negative Impact and Not Enough Information Available only occur for the Impact on Odds metric.

(4) Feature by Stage Metric Table

This table is an extension of the data shown in the heatmap. This data is more comprehensive since it does not limit by the maximum number of features selected. The table columns are described below:

  • Stage Name: The stage of interest for a given metric.
  • Stage Order: The associated stage order for a given stage name.
  • Feature Column: This column is dynamic depending on the Feature Type that was selected (Channel, Campaign, etc.).
  • Metric Column: This column is dynamic depending on the metric selected (# of Successes, Probability of Success, etc.)

By default the metric column shown is the one that is configured for the Answer. All other metrics of interest are available in this table by selecting them from the table menu in the top right of the widget.

(5) Filter Tray

Options in the filter tray allow for exploration and on-the-fly changing of a small subset of Answers configuration. Here, we enable the ability to change the funnel stage of interest as well as the feature type. These changes do not affect the underlying configuration.

How did we do?

Setting Up Answers - Start Here!

Using CaliberMind Answers to Understand Campaign Performance