Drop-Off Analysis: How to Find Friction Points
How do you conduct drop-off analysis to identify friction points in the user journey?
This is the main question we cover in the article, so if you’re after the answer, you’re in the right place.
Let’s get right to it!
Drop-off analysis (short summary)
- The drop-off rate is the percentage of users who don’t complete an action or process.
- To calculate it, divide the number of users who finished the activity by the number of those who started.
- The first step in the drop-off analysis process is defining the conversion events, like completing the registration forms in the sign-up flow.
- If your analysis drop-off rates for specific user personas or segments, use filters to ensure you’re including the relevant data.
- Next, generate the funnel report in your analytics tool to visualize the conversion/drop-off rate data.
- Apart from the conversion and drop-off rates, pay attention to other metrics, like time to convert and the total number of conversions.
- When you know where users drop off, conduct additional research to identify possible causes and create hypotheses.
- To test the hypotheses, run experiments, like A/B or multivariate tests.
- If your hypothesis is true, release the improvements to all users in the relevant segment.
- Finally, monitor the impact of the changes and iterate for further improvements.
- To be able to conduct drop-off analysis easily and efficiently, use the right tools, like Userpilot. To learn more about its analytics and experimentation features, book the demo!
What is a drop-off rate in SaaS?
A drop-off rate is the percentage of users who stop using the product at different stages of the user journey or touchpoints.
In a SaaS product, we often talk about drop-off rates in the context of:
- Website conversions – when users don’t complete the purchase, book the demo, or sign up for a free trial.
- Free trial conversions – when users fail to convert to paying customers.
- Onboarding process – when new users don’t complete the onboarding process or don’t reach a specific milestone, like the activation point.
- Feature usage – when users stop using a feature.
How to calculate drop-off rates?
To calculate the drop-off rate, divide the number of users who abandoned a particular step, process, or action by the number of users who started it within a specific period, and multiply it by 100.
For example, if 200 users start your onboarding process but only 174 finish it, the drop-off rate will be 13%: 26/200 x 100=13%
What is a good drop-off rate in SaaS?
The drop-off rate is the opposite of the conversion rate, which is the percentage of users who reach the next stage or step.
Let’s look at a few numbers to give you an idea of what to aspire to:
- For website conversions, the average conversion rate for B2B SaaS businesses is between 0.9 and 2.3%, depending on your niche. So a drop-off rate lower than 97.7% could be considered good.
- An average free trial conversion rate is 17–25% and goes up to 50% when credit card details are required at sign-up.
- An average freemium conversion rate is 1-10%, depending on the niche and product.
- The average activation rate is around 36%.
- Median gross retention is around 91%.
All of these rates depend a lot on the unique characteristics of your product, industry, and sector. That’s why it’s good practice to measure your current rates and work to optimize them.
How to perform drop-off analysis
Drop-off analysis allows you to answer 2 main questions: where are users dropping off and why?
The guide’s meant to be short and sweet. For something a bit more comprehensive, sign up for our webinar where Adam Thomas and Lusine Sargsyan are going to cover the process in more depth.
Now, let’s have a look at how to do it.
Use an analytics tool to simplify the process
The first step in the process is choosing the right tool for the job.
Here are 3 analytics tools that cover the whole user journey.
- Google Analytics – for tracking web conversions and user behavior on the website.
- Userpilot – for user behavior analysis inside of the web app.
- Mixpanel – for tracking user behavior in mobile apps.
In addition, you may need a software application with session recording capability so that you can collect granular data about user behavior on a page or screen. This functionality is coming to Userpilot in Q2 2024 and for now, Hotjar and FullStory are robust solutions.
Define the conversion funnel with custom events
The actual drop-off analysis starts by defining the funnel events. These are the user actions that mark conversion from one stage of the journey to another, and as such their unique to each product and often user persona.
For example, your activation funnel may consist of three events: starting onboarding, creating a project, and adding a contact.
Here’s how you do it in Userpilot:
1) In the sidebar menu, click on Analytics and choose Funnels.
2) Add your events by clicking on + Select Event.
3) From the menu below, define the Conversion Criteria:
- Choose either users or company-level.
- Decide if the events need to happen in a specific order (for linear user journeys with a set sequence of events) or any order (for non-linear journeys).
- The conversion time – the window within which users need to complete the actions to be included in the funnel, e.g. 1 day, 1 week, 1 month, etc.
Optimize the conversion funnel for different user personas
As mentioned, the conversion funnel may be different for each user persona. This means you may need to apply additional criteria to what’s included in the funnel and what’s not.
In Userpilot, you can do it with filters. You can filter data by User and Company properties, like sign-up date or plan type, or choose specific segments you created before.
For our analysis, we’ve set the filters:
- Signed Up – exactly 20 days ago.
- Plan Type – equals (name of your plan).
- Web sessions – 5.
Visualize the funnel stages
Once you define the events, set conversion criteria, and filter the data for specific user personas, just run the query to visualize the funnel stages.
The funnel chart consists of bars, each representing the number of users at each stage of the funnel. As users drop off at each stage, the bars get narrower.
Sometimes just one look at the chart is enough to find the stage in the journey where the drop-off rate is high because the next bar is dramatically narrower.
In Userpilot, you can also break down the data based on a specific filter, like the operating system.
Identify key metrics for analyzing funnel performance
To benefit from funnel analysis though, just looking at the chart isn’t enough, so analyze the funnel metrics.
For example, there may be no massive drops between stages in the funnel, but if you look at the average conversion time, you may discover it takes users longer to convert than it should because of friction.
What other metrics can you find in the conversion chart?
- Event completion rate.
- Total number of conversions.
Create hypotheses to analyze funnel drop-offs
Once you identify drop-off points or areas for conversion rate optimization, create hypotheses on the potential causes.
For example, ‘users don’t complete the onboarding checklist because there are too many tasks’.
For that, you may need to conduct further research. Depending on what you’re trying to optimize, this could involve:
- Path analysis – to track all the user actions leading up to the conversion event.
- Heatmap and session recording analysis – to analyze user interactions with the webpage or app UI.
Test hypotheses with other user behavior data
Hypotheses are just speculations, so the next step is testing them. This normally involves some sort of experimentation.
For example, if you think that there are too many tasks in the onboarding checklist, you can create different versions of the checklist and run A/B or multivariate tests to see if they reduce the drop-off rate.
In Userpilot, you can conduct 3 types of experiments:
- Controlled A/B Test – to test a new version against the original.
- Head-to-Head A/B Test – to compare the performance of 2 new versions.
- Controlled Multivariate Test – the multiple versions against the original.
To test the changes to the onboarding checklist, we would use one of the controlled tests.
For experiments on more than one page, use dedicated experimentation software like Optimizely.
Implement improvements to the conversion process
To limit risk, experiments are conducted on a small (yet representative) user group.
However, once your experiments confirm your hypothesis, you can roll out the changes for all your users (or relevant segments).
Monitor and iterate
Even if your experiments showed that a specific change is beneficial, you still need to track the impact of the changes.
In practice, this means repeating the drop-off analysis process.
Continuous tracking of relevant metrics and analyzing user behavior allows you to identify additional improvements that will reduce the drop-off rate even further.
Conclusion
There you go! Here’s how you conduct drop-off analysis to help users achieve their goals and improve the business performance of the product.
And if you want to find out more about drop-off analysis in Userpilot, book the demo!
Try Userpilot and Take Your Product Analytics to the Next Level
