We’ve Analyzed 5,808 Funnels. Here are the Top 5 Mistakes You Should Avoid
Arik Marmorstein, Wed May 19 2021, 11 min


As marketers, we’re obsessed with growth, traffic, high-quality visitors, and of course, lots of conversions. Funnels are the most powerful tool for connecting all your marketing efforts together and knowing exactly where and why you lose visitors. 

Most websites lose over 99%(!) of their visitors before the conversions. This means the vast majority of those high-intent visitors you fought so hard for end up getting lost. In most cases, optimizing your funnel is the easiest way to grow. Think about it, if you manage to increase the conversion rate from 1% to 1.5%, that means 50% growth! 50% growth without increasing your budget at all. 

At Oribi, we analyze website behavior and work with tens of thousands of companies. During the last year, our conversion experts team analyzed 5,808 marketing funnels. This article summarizes the most common mistakes, how to avoid them, and of course, the best practices for building the right funnels, optimizing them, and mastering your visitors’ flow. 

Mistake #1 -  Not building your funnel properly

Best Practice #1 - Follow the following three steps when building your funnel

Setting the right steps on your funnel is critical. Defining too many events, too little or the wrong steps can lead to misleading results. Here are the three steps to building your main funnel right. 
Step 1. Identify the key milestones you want your website visitors to pass before they convert. 

I recommend starting with a funnel of 4-6 steps. Once you have optimized all the steps in that funnel, build a more granular funnel that includes all the minor details. 

A common mistake we see here is not including the most important steps in your funnel (such as the thank-you page which indicates the completion of the funnel) or starting your optimization journey with a 12-step funnel. 

Let’s take Airbnb’s funnel for example:
  1. Visiting the home page 
  2. Visiting the “search properties” page 
  3. Visiting a property page 
  4. Clicking “book it” 
  5. Visiting the checkout page
  6. Getting to the “thank you” page after a purchase 

We can, of course, take that sales funnel and make it much more granular, taking into account every little step a visitor passes until a booking is completed (e.g. typing your name, adding credit card details, etc.). But, I’d recommend starting with this “high-level” funnel.  

Step 2. Once we have identified our key milestones, it’s time to mark them on your website analytics tool so you can access them easily. 
This is the typical funnel for many websites. 
  1. Visits to the most popular landing page
  2. Visits to the main sales page
  3. Clicks on the purchase button
  4. Visits to the thank you page
Below you can see the process for pinning one of the above events using Oribi (I can track any event, even retroactively)
New event builder
C. After marking the events, it’s time to visualize the funnel using your website analytics tool. We do this in order to identify what steps we should work on. 

Below you can see an example of my own site’s funnel (although it does not include the final step).
Screencast - funnel creation

I will not get into how to set up a conversion funnel in Google Analytics since the process is a bit daunting. 

Banner 1 - Funnels blog post

Mistake #2 -  Not treating your funnel optimization as a long-term process

Best Practice #2 - Constantly optimize your funnel, there’s always room for improvement. 

The term ‘conversion funnel optimization’ means we are constantly working on reducing the number of people who drop out of our funnel.
We don’t do this by pouring more visitors into the top of the funnel—that would be user acquisition. 
Instead, our focus is on getting the most conversions from our existing visitors—fixing the “holes in the bucket”.  That process takes time. 
We see companies who start the process and stop following a quick win or give up following a small loss. That’s a shame. 

Mistake #3 -  Not treating your funnel optimization as an experimental process

Best Practice #3 - Use our “AASTL” framework when optimizing your funnel

Your funnel optimization is a scientific experiment that includes 5 steps (“AASTL”). 
Step 1- Analysis
Here, we identify which step of our sales funnel most visitors drop out. Identifying this drop will help us focus our efforts on the most impactful phase. 
We should prioritize the steps on which visitors have higher intent. Let’s have a look again at Airbnb’s example:
  1. Visiting the home page 
  2. Visiting the “search properties” page 
  3. Clicking “book it” 
  4. Visiting the checkout page
  5. Getting to the “thank you” page after a purchase 

As it’s normal for many visitors to come to a site with low intent, the biggest drop will usually occur between steps 1 and 2. The visitors who drop out here most likely did not come with real intent to book a listing. Because we’re more interested in visitors who come with higher intent to actually book something, we’ll want to focus on the biggest drop between steps 2-5 (or even 3-5).
The most common mistake we see here is focusing on the biggest drop (between steps 1 & 2) without justification. 
Step 2 - Make Assumptions as to why people drop out.
Use your intuition and, if possible, gather qualitative data (talking to people who dropped out or watching what they do) and quantitative data (surveys or tests) to make educated assumptions as to why people drop out. 
Pro tip: You can use Oribi’s Visitor Journeys in order to gather granular qualitative data about what users do on your site.  
Oribi's Visitor Journeys Screencast
Step 3 - Solutions

Make a list of potential solutions and prioritize them based on a combination of 3 factors: 
  • Impact - if they work, how much they would move the needle
  • Complexity - how much effort the change would require
  • Confidence - how confident you are that the solution will work

Obviously, the solution with a mix of high impact, high confidence, and low complexity is the one you should prioritize. 
It’s helpful to ask yourself why each solution would or wouldn’t work when you are trying to determine the level of complexity and confidence. 

Step 4 - Test and compare results 
After you have implemented one change for each step, wait enough time to get a decent amount of data and then compare the results. 

A common mistake we see here is making more than one change for each step, which results in not knowing to which change you should attribute the success or failure. You could change the CTA, rewrite the main value proposition, or move an element to a different position. Just keep it to one change at a time. 

Another common mistake we see is not giving the “test” enough time. Even if you have lots of website visitors each day, give it at least a week so that day-based behavior change won’t affect the results. People behave differently on different days. For example, they are more likely to buy on Tuesday.

Step 5 - Learn
It’s tempting to skip this step. After all, you got the results. But this phase is part of what differentiates good marketers from great ones. If you learned that something doesn't work with your audience, you’ll know not to use it in the future, and vice versa. 
Since so many companies skip part or all of this process, I’d like to give an example of how to implement it. 

For the sake of example, I’ll use Spectroomz, which helps autistic adults learn digital marketing. 

The site has several funnels, but the main one takes people from a quiz that helps autistic adults find the most suitable remote jobs through the main sales page, and then on to clicking the purchase button (which is where I’ll stop the analysis). 

I know this is my main funnel thanks also to the Event Correlations feature which shows me that people who visit the quiz are 21x more likely to click the purchase button. 
A screenshot from Oribi's event correlation section
Now, let’s optimize that funnel.

Step 1- Analysis

This is my funnel for April 18th-24th:
Screenshot of sales funnel on Oribi
We can see the drop between steps 1 & 2 is 86% and 54% between steps 2 & 3. 
Although I wrote that we shouldn’t focus on this step initially, I actually am going to focus on optimizing the drop between steps 1 & 2 for three reasons: 

A) People who visit the quiz get there organically following a Google search about remote jobs for autistic adults. They have a very specific (and strong) intent. 

B) The biggest drop is there.  

C) It’s a relatively short funnel.  

Step 2 - Assumptions.
Intuition - People who visit the quiz search for either remote jobs or what remote jobs would be a good fit for them. The current call-to-action (CTA) that leads them to the home page (which is the main sales page) doesn’t give information about how taking the courses correlates with getting a remote job. 
This is the current CTA:

screenshot of CTA before the change

Qualitative data that supports my assumption - I interviewed several site visitors who went through this funnel. They all actively confirmed my intuition (meaning that I didn’t tell them what my assumption was). 

Step 3 - Solutions
Solution 1: Change the CTA to include a testimonial. 
Solution 2: Add another CTA at the end of the quiz itself (the quiz is done on Typeform).
Solution prioritization spreadsheet
Step 4 - Test change and compare results. 

This is the change I implemented:
Screenshot of CTA post-change
And these are the results a week later
A screenshot of the sales funnel on Oribi after the change
As you can see, I increased the conversion rate between steps 1 &2 by 1%. 

Step 5 - Learn
I learned from this experiment that visitors in the first step have a strong intention to find a job, and associating that intention with what the site has to offer will get more people to complete the funnel. 
This knowledge will help when using different marketing channels for the audience of this site. For example, when using email marketing to get more people to subscribe to the service, the emails will indicate that the courses actually help people get remote jobs. 

Get Access to our AASTL Template

Banner 2 - blog post

Mistake #4 -  Not segmenting your funnel optimization

Best Practice #4 - Segment your funnel optimization by traffic source, platforms, or Location

If you have significant traffic coming from multiple sources (social, organic, paid, etc.), optimizing one funnel for all traffic sources is a mistake. 

Visitors from Facebook are likely to behave differently than Google organic visitors (who usually have higher intent). 

Depending on your main growth engine, start your funnel optimization with your main channel, and then analyze the funnels for other significant traffic sources.  

This is also applicable to the type of platform (mobile vs. desktop) or location of the traffic (visitors from the US will behave differently than those from Australia). 

You can do all these breakdowns easily with Oribi’s Super Funnels breakdowns.

Oribi's super funnels screenshot
The main segments you should focus on are channels, geos, platforms (mobile vs. desktop), and UTMs for different types of campaigns (newsletters, Facebook campaigns, etc.).
When you launch a new campaign or send a newsletter, compare the funnels between different segments. Using Oribi, you can do this with a single click. 
For example, if you see that mobile visitors from the US convert better, you can focus on this audience and/or improve your desktop experience to increase conversions there. 

Banner 3 - funnels blog

Mistake #5 -  Not using sub-funnels

Best Practice #5 - In addition to the main funnels, test sub-funnels to increase conversions. 

In most cases, you will have 1-3 main funnels. These could be your sign up funnel, purchase funnel, or schedule a demo funnel, to name a few. While you should focus on your main funnels, there are additional funnels — sub-funnels — that you should optimize to increase overall conversions. 
For example, testing your main funnel with a different landing page would be a sub-funnel. Adding the pricing page into the middle of your funnel is also a sub-funnel worth testing. 

The Visitor Journey Aggregator in Oribi is a good way to discover sub-funnels. After choosing your main conversion event, Oribi will show you common journeys visitors take on their way to convert. If you see, for example, that a lot of visitors visit the pricing page before converting, it might be a good idea to include that page (or info from that page) in the main funnel.

Visitor Journeys Aggregator GIF

Now it’s your turn!

Funnels are the most popular feature in Oribi. They enable top companies to optimize their funnels efficiently and successfully. Optimize your own funnel using Oribi or schedule a tour with one of our conversion experts to get yours started.
Banner 4 - funnels blog

Arik Marmorstein is Oribi's product marketing lead. He has 10 years of marketing experience.