Conversions - Your first Conversion Analysis

This is part 2/5 of a series explaining Conversions:

• Introduction to Conversions
• Your first Conversion Analysis
• An overview of your Conversions results
• Digging into & sharing your Conversions results
• Digging into your Conversions results II

Overview

The more sessions you have captured, the more signal Conversions will have to detect significant customer impacts. There’s no absolute minimum number of sessions that will work for all situations, but we’d recommend having at least a few hundred sessions in which a customer engages deeply with your funnel (i.e. not just sessions of customers bouncing at the first step!) before seriously diving in to Conversions.

You’ll start by using the Conversions Composer to select exactly what you want to analyze.

Running a Conversions Analysis

In this section, we’ll walk you through how to get results from running a Conversions Analysis, along with some technical clarifications along the way for your reference. If you’re in a hurry, we have a separate “quick start” guide and you can catch up on these details later.

Create or Select Conversion Funnels

The first step in running an analysis is to create a new conversion funnel or select a saved funnel. For this example, let’s create a new funnel by clicking Create custom Funnel.

A new Conversion Funnel needs a name and at least two events to define the steps. While you’re setting up your funnel, we show you a preview of the results of your funnel over the last 30 days (how many users made it from Step 1 to the final step) so you can confirm you’re building the funnel correctly. If your funnel has no users completing the funnel, you won’t see any results.

You can also edit or delete any existing Conversion Funnels within the Conversions Funnel Builder.

We allow you to specify in your Conversion Funnel if you’d like to look at funnels where all the events happened “within the same session” (the default) or “across any number of sessions”. We recommend the default unless it would be typical for your users to take multiple sessions to complete your Conversion Funnel.

• If you select “across any number of sessions”, the analysis will consider the number of unique users who have participated in the funnel flow, even across two or more separate sessions.
  
  • This means, a user’s funnel steps and signal or friction events can occur in different sessions.
    
• If you select “within the same session”, the analysis will consider the number of unique users who have participated in the funnel flow, where enforcing the funnel steps and friction occur in the same session.
  
  • A user’s funnel steps and signal or friction events must occur in the same session.
  • Users who participated across any number of sessions may be accounted for in both the affected and unaffected unique user counts and conversion rates, within the analysis. This is because these users may experience a signal during one session and not in another.

Note: by default, funnels will be shared and visible to your whole organization.

Choose “Performed By” Cohort

Next, select the cohort of users you’d like to pass through your funnel. Cohorts correspond to a Fullstory Segment that you have previously saved, and therefore defines people who ever met the segment criteria within the date range you select in the next step. The time range you select is independent of whether a customer met the segment criteria at the time they engaged with the funnel (although it often overlaps) and it also overrides any date range that was set in the original saved segment (but does not alter the original segment).

For now, you can pass through Everyone as your cohort. If you’d like to get more granular later, try creating a segment like “Mobile users,” or all users from a specific country or region.

Choose date range

The date range selects when the events defining the Conversions Funnel must have occurred, and is defined according to the local browser time of your customer. Additionally, all the “Performed By” Segment criteria must hold within this range for a session to be counted in the analysis.

If you want to see the effect of recent changes (such as a new feature you released last week), try the “Past 7 days” or a shorter time frame. For a good overview of longer-running issues, try “Past 30 days”. Keep in mind, if you select the custom date range, the range must be 30 days or less. 

Add or remove signals to be analyzed

Are you in a situation where you know you can’t work on page performance issues this month, and you want to focus on issues you can address right now? You can choose to streamline the analysis to focus only on your most relevant signals (uncheck Page Performance).

You can also create your own signals to get other types of insight. If you instrument Custom Events to detect points of friction (for instance, errors that happen in the signup or checkout process) you can select those custom events as signals for Conversions to examine. In a similar way, you can set up Watched Elements on your error modals or specific CSS selectors as additional signals. Conversions supports string and boolean property types.

Analyze

Click Analyze to begin your analysis! This could take a minute or two depending on how complex the analysis is that you have constructed. We will explore the analysis results in our next article, linked at the bottom of the page.

(Optional) Save your Analysis

Now that you have built and configured the perfect analysis (funnel + user cohort + signals) for your new business use case, you may have a desire to share these results with other people on your team or to come back and check this analysis again later to see how things have changed. Be sure to click Save Analysis, which will prompt you to name your new analysis. After choosing a name for your analysis, click Save to complete the process.

Go back and view your analysis from the Saved Analyses view. When viewing a saved analysis, you have the ability to save changes to that analysis, save changes as a new analysis or to delete that analysis by clicking the trash can button.

In summary, building your first Conversion Analysis will:

• Automatically prioritize opportunities based on the strongest signals we detect from user behavior in the funnel.
• Help you dive into specific results to learn more about the opportunity, such as how many and how severely users are impacted.

Now that you understand how to begin exploring your funnel conversions, you may have questions about the results. You can read more about Understanding your Conversions results.