How to find funnel drop-offs fast and stop losing conversions

How to find funnel drop-offs fast and stop losing conversions

Martin Bolf
Martin Bolf (Smartlook Team)  |  Last updated: Feb 14, 2023
11 mins read
When you pair user recordings with funnels, you can quickly find drop-offs and understand why users don't convert.

Funnel drop-offs — the moment when users abandon a conversion process on a website or mobile application — often indicate areas where users are experiencing problems with your product. Finding those drop-offs and fixing the problems that cause them can be a great way to improve your conversion rate.

However, most tools for finding funnel drop-offs have two big limitations:

  1. They don’t tell you why visitors dropped off. For example, it’s easy to visualize your funnels with Google Analytics and identify the pages where most users fall off (as in the screenshot below). But how do you know what caused those drop-offs? To get that information, some people combine Google Analytics with tools like Hotjar for screen recordings and heatmaps. Looking for drop-offs this way is slow and painful because you often have to watch hundreds of recordings to find one that answers your question.
Checkout Behavior Analysis in Google Analytics
  1. Gathering enough data for new funnels requires a lot of time and website traffic. With most analytics tools, you have to define a funnel and then wait for new website traffic before you can make statistically significant conclusions. During that time, drop-offs remain undetected, meaning you’re missing out on potential revenue.

In this guide, we’ll show you how to avoid these issues with Smartlook, our analytics tool for websites and mobile apps, which pairs funnel analysis with always-on screen recordings so you can immediately find drop-offs, even retroactively. 

With this process, you’ll be able to quickly find drop-offs that affect your revenue and, more importantly, see why they happened. 

The benefits of combining funnel analysis with screen recordings

Knowing a funnel’s drop-off rate isn’t very useful by itself. But when you understand why someone dropped off, you can optimize the user journey and make changes that may increase conversions.

Let’s say we’re looking at the flow between clicking the “Pay Now” button and the “Thank You” page of an ecommerce store. In Smartlook, you can create such a funnel with two events:

  • Event 1: Users click on the “Pay Now” button (you can select the button with our no-code picker to define the event).
  • Event 2: Users land on the “Thank You” page (visited URL event).

In theory, there shouldn’t be drop-offs between these two steps — people who picked an item and entered their credit card information clearly want to buy. In reality, you can often find drop-offs at this stage. 

For example, this screenshot of Smartlook shows a checkout funnel with a 16.41% drop-off rate between these two events.

Smartlook Checkout Funnel: Visitors, Dropoff, Conversion Rate

This is where screen recordings shine. By clicking the “Play” button underneath each step of the funnel, you can go directly to the screen recordings of all the visitors who dropped off at that stage. This lets you see the experience and behavior of those 16.41% of users without sifting through all other recordings of people who went through the same funnel. 

While we’re on the topic of efficiency and time-savings, let’s see how Smartlook avoids the second limitation we mentioned, i.e., the time it takes to collect enough data for new funnels.

Historical data: Finding drop-offs in new funnels without waiting for traffic 

For most websites, it takes several weeks to get enough funnel traffic to make statistically significant conclusions. And with most analytics tools, you have to go through this waiting period for every new funnel. During that time, you might be missing out on conversions and consequently, revenue.

To avoid this issue, Smartlook can take your historical data and turn it into insights, screen recordings, and heatmaps that you can analyze retroactively. 

Here’s how:

After you install the Smartlook snippet, Smartlook starts collecting every user interaction on your website in the background via always-on screen recordings. That’s why when you define a new funnel, it automatically gets populated with historical data, as far back as your data retention plan goes. 

As a result, you can immediately locate drop-off points and start optimizing your funnel, without waiting for new traffic. This feature also comes in handy if you want to add a step to an existing funnel, as it only takes a few clicks to add the step.

Now, let’s see how you can create and monitor funnels in Smartlook.

How to create essential funnels and monitor sharp drop-offs

Finding a website’s overall drop-off rate (and conversely, its overall conversion rate) is simple. You just need to know how many people visited the site and how many of them converted. 

But that’s not particularly useful since users go through a lot of steps in their journey. Each step can push them further along the conversion path or cause a drop-off. That’s why it’s better to look for drop-offs between consecutive steps inside each funnel, like in the examples below.

Before we dive in, keep in mind that the funnels shown here consist of standard events (like URL visits, button clicks, and text inputs), which you can set up in Smartlook without any programming. 

Custom Event with Custom Property in Smartlook

When you have your events, simply put them in the required order to create a funnel. 

Once you do that, Smartlook automatically creates the funnel visualization, like in the GIF below.

Also, for each step of the funnel, Smartlook can break down users by country, device, operating system, and other criteria. 

Break down users by country, city, state, device, browser, operating system, etc.

This lets you easily identify issues that are specific to a category of users.

The checkout funnel

We already touched on this, but drop-offs during the checkout process are a common way to lose potential customers. 

Here’s how you can create a funnel in Smartlook that tracks your users’ entire checkout process, from the homepage, through adding an item to the cart, to clicking on “Pay Now” and getting a confirmation of purchase:

  • Event 1: A homepage visitor clicks on the shop (select the button with our no-code picker to define the event).
  • Event 2: The visitor searches for an item (typed text).
  • Event 3: The visitor clicks “Add to cart” (clicked on a CSS selector).
  • Event 4: They click on the “Pay Now” button (clicked on text).
  • Event 5: They land on the “Thank you” page (visited URL).

Many ecommerce sites will see a large drop-off between users who add an item to the shopping cart and users who purchase. To further dig into why customers who were interested in an item didn’t buy, you can add more events between steps 3 and 4.

You could create separate events for everything that happens on the checkout page, like selecting shipping options, adding discount codes, or filling in billing info. One of our customers did this and discovered that customers were abandoning their carts because their shipping costs were much higher than expected. After they lowered their shipping rates, sales increased by 161%.

While each step in this funnel can reveal user experience issues that give you clues about why people aren’t converting, it’s especially vital to investigate any drop-off between events 4 and 5. Anyone who clicked “Pay now” and didn’t land on the “Thank you” page probably encountered a technical error.

If you find drop-offs here, you need to dig into the recordings and see what’s going on. There are two common problems you might find:

  • JavaScript errors. Smartlook automatically detects those and shows them in DevTools, so you can quickly send them to your dev team.
Dev errors in Smartlook
  • Problems with the payment method or payment provider. Maybe some users entered the wrong card info, or your payment provider couldn’t accept their payment. Again, that’s easily identifiable with screen recordings, as shown in the screenshot below.
Detailed payment recording in Smartlook

Whatever the case, you want to get on this right away. If left unchecked, these problems can add up to thousands of dollars in lost revenue. 

Let’s say your average order is $25 a day. If technical issues prevent just one user from checking out per day, that adds up to $9,125 of lost revenue every year. Imagine if you’re selling expensive items or subscription plans — the losses can easily be in the hundreds of thousands. Not to mention, the experience of trying to buy something unsuccessfully is infuriating. 

Email newsletter subscriptions

A common conversion funnel that marketers care about maps the journey from reading a blog post to subscribing to a mailing list. 

If you’re trying to grow your email base, you likely have a form where people can leave their email and a call-to-action (CTA) button to subscribe to your newsletter. 

Here’s what the funnel optimization process might look like for that flow:

  • Event 1: Users arrive on the landing page (visited URL containing “blog”).
  • Event 2: They fill out the form field for entering their email address (typed text containing the @ symbol.).
  • Event 3: Users click the CTA button (select the button with our no-code picker).
  • Event 4: They get a confirmation or arrive on a “Thank you” page (visited URL).

If there’s a drop-off between events 1 and 2, you might want to analyze which areas of your pages garner the most eyeballs with Smartlook’s heatmaps. Once you know that, you can try increasing the conversion rate by moving your form and CTA to more visible areas.

Smartlook's Heatmap in Action

After updating the design, you can also create a new heatmap and contrast the user behavior with the old one, just by changing the date range. 

It’s also crucial to look for drop-offs between steps 2 and 3, since in theory, a user who types in an email should always click the subscribe button. In reality, technical issues can prevent some users from subscribing, like false warnings that their email is invalid. Also, some users may get distracted by other elements on the page and forget to subscribe.

Similarly, there should be no drop-offs between events 3 and 4. Users who click on the button after entering a valid email should get a confirmation or land on the “Thank you” page. 

In both cases, screen recordings can help you quickly learn why these unlikely drop-offs occurred.

Mobile and SaaS application onboarding tutorials

Most mobile apps and SaaS products include a product tutorial for first-time users, but many users don’t complete the tutorial during their free trial. This is a big challenge for product managers since it makes it difficult to convert people from free to paid. 

By mapping the entire tutorial to events in Smartlook (you’ll need a developer because you can only track interactions on mobile devices with custom events), you can find user behavior insights by watching session replays of people who quit the tutorial early.

Tutorial Recording in Smartlook

The image above shows a moment during a Smartlook screen recording of a mobile user’s journey while navigating an onboarding sequence. During the roughly 10-minute session, the user completed 91 events.

Let’s say the overall drop-off rate for this funnel is 90%. That might make you consider redesigning or removing the entire tutorial. But if the average user completes almost 100 events before quitting, there’s probably some value in the tutorial. 

If you look at the drop-off rate between each step, you might learn that the sharpest drop-off happens near the end of the tutorial. Perhaps there’s a technical error happening between those steps, or the users simply think they’ve learned everything they need at that point. Again, to get a better idea of the user experience and make more informed decisions, you want to analyze the drop-offs between steps of the funnel, not just the overall drop-off rate.

Getting alerts for sharp drop-offs 

You should always monitor your essential funnels (those closest to a purchase event or another business goal) and be the first to learn if there’s a sharp drop-off. 

Smartlook has a special feature for that, called anomalies. Once you set up an anomaly for an event or funnel, you’ll get alerts in the app or via email whenever sharp drop-offs (or increases) happen. These alerts can be triggered in two ways:

  • By a conversion rate dropping outside of a defined range, e.g., below 2%. 
  • By a percentage change from your baseline, e.g., 50% lower than normal. 

When anomalies are set up, you’re instantly alerted about critical errors, and you can address them much faster than if you wait for customers to complain about your checkout being broken. That’s why anomaly alerts can potentially save you a ton of conversions — if you act on them quickly. 

Next steps: Set up Smartlook, define your events, and get started with funnel analytics

We just covered a lot, so here’s how you can put everything into action:

  1. Sign up for a free 30-day Smartlook trial — no credit card required.
  2. Install our code snippet by following the instructions in our Help Center. As soon as you do that, Smartlook will start recording everything your visitors do.
  3. Define events. Again, events are all user interactions that can be measured. You can track standard events (URL visits, button clicks, text inputs, and a few others) without any programming skills, or create custom events with JavaScript to track pretty much anything else.
  4. Add your events to a funnel. To create a funnel, choose two or more events and put them in the order you believe your users follow.

Finally, if you want a more detailed presentation of Smartlook that’s tailored to your business, schedule a demo with our team.

Martin Bolf
Martin Bolf

is the product manager at Smartlook. Martin is enthusiastic about delivering the best possible customer experience. Prior to joining Smartlook as a product manager, he used to work as a consultant for Oracle NetSuite. Martin has a deep professional interest in biometric signing and work digitalization. He is also an NFL enthusiast and likes to enjoy good food (ideally while watching NFL).

0 %