Search tips: bounce rate, form abandonment, and JavaScript errors
Tips and Tricks · 7 min read

Search tips: bounce rate, form abandonment, and JavaScript errors

Caitlin Brett
Posted June 13, 2017

Troubleshoot bounce rate and form abandonment, analyze product engagement, find software bugs, and find when customers get lost

New FullStory customers are often amazed when their first sessions roll in. They watch each and every one, proclaiming, “This is incredible! I know exactly how my customers are using my site!”

But watching every playback quickly becomes overwhelming, and after the 1,000th session appears, our customers begin asking us, “So, what exactly should we be looking for? I can’t watch all of these sessions.”

Without further ado, I present the 5 searches we think all of our customers should be using. These are not, by any means, the only five FullStory searches you should run, but they may be the most universally beneficial. No matter what kind of product or service you offer, chances are these searches will turn up some tasty morsels of customer insight.

1. Why do users sign up but not do _________?

This is a major pain point for all our product management users. You got your customer to sign up, but they didn’t do that very important thing you wanted them to do next. (I left this blank intentionally.)

If you’re one of our e-commerce clients, maybe your customer never clicked “check out now.” A SaaS company? Perhaps the customer registered but never created a new project or never completed the onboarding questionnaire.

You may already know what the blank is in this case, because you’re tracking it in an analytics tool like Mixpanel or Google Analytics. But these kinds of tools only show you the number of customers falling out of the funnel. FullStory shows you why and how to improve.

To execute the search in FullStory. First, move from “Everyone” to “Signed-up” — you can do this in the left sidebar of FullStory:

Next, in the OmniSearch bar, search for the CSS selector, action, or whatever it was you wanted them to do. E.g. “add to cart” works for our example and we select Clicked text: Add to cart:

This creates the following combined search:

Step 1 for setting up this search — be sure to complete the next step!

Don’t miss this next step! The default here is to find users who did [the thing], as captured by the “Has” selector. We’re looking for those other users, so you need to toggle that to Has not:

Modify this search to find a user who “is signed up” but “has not“ done [whatever it is you want them to do].

From there, you’ll get all the sessions from your signed-up users who never did [that thing you wanted them to do].

NOTE: This search relies on you to identify your users with the FullStory JavaScript API; it won’t work if you’re not, because all of your customers show up as anonymous or unregistered. (If this sounds like a description of your FS dashboard, I highly recommend you visit the above link, because FullStory works so much better when you know who your customers are.)

2. Why do my visitors bounce?

Bounce rate may be one you’re probably already tracking in Google Analytics, but for a couple important reasons you need to make this a FullStory segment, as well:

  1. Google Analytics doesn’t do a definitive job of measuring bounce rate— at least not the way we think of bounce rate (read on that here). In particular, Google Analytics does an especially poor job on single-page apps. GA doesn’t know how long the customer spent on your site — only which URLs they visited. So whether a visitor spent 3 seconds or 3 hours reading your landing page doesn’t matter: if they only visited your landing page, GA calls that visitor a bounce.

  2. But why is your visitor bouncing? If they can’t find the information they’re looking for, watching a FullStory session will reveal what you can do to help. If they’re scrolling to the pricing section and then immediately navigating away, FullStory will show you that, too. And will make great fodder for your presentation to the CEO next week.

To look for bounce in FullStory, in the OmniSearch bar start typing “total active …”

From there, you need to toggle the “is at least” to “is at most” and set it at the time you want to consider a bounce. E.g. users arrive and are only on a single page for “is at most 30 seconds:”

And of course, you can look at the above sessions by whatever Event you want — e.g. people who were referred by Facebook … but weren’t active more than 30 seconds.

FullStory can handle whatever bounce rate definition you want — and cut it by referrer.

Bounce rate, perfected!

3. Which form fields don’t my customers want to fill out? (Form Abandonment)

Ever wanted to understand why users give up on filling out important forms on your site? What’s going wrong? What questions are tripping them up?

Simply start typing “form” into the OmniSearch bar and then click “Abandoned any form.”

FullStory has a built in ability to pull up sessions from users who abandoned forms (Read about Form Abandonment on FullStory here). It’s as simple as either typing “form” in the OmniSearch bar or you can find it under the “FRICTION” in the dropdown that appears when clicking into the OmniSearch bar.

The resulting search is only the beginning, of course. You might then want to cut your abandoned form sessions by other parameters.

The abandoned form filter on FullStory.

4. Who isn’t finding what they’re searching for?

You have a search bar so people can easily find the things they want to buy or the answers to their questions. Naturally, you’d want to know if they can’t.

You could create a Changed search focused on your search field, which will result in a segment of all customers who’ve typed something into search.

Type “changed” into the OmniSearch bar, click “Changed anything.”

Here’s the Event Filter that will appear for “Changed.”

The “Changed” event filter. You’ll still need to type something — try typing “search.” Alternatively, use Inspect Mode and find the relevant elements for the search bar on your site. Once found, just create a new search. A walk through on how to do this here.

Start typing the CSS selector for your search field use Inspect Mode to point and click on it, directly.

You can mine these sessions that includes searches to see what kinds of questions users are asking.

To take things even further, add an additional filter to identify users who go on to contact support. It’s a strong signal that the information they’re seeking doesn’t exist, or, at least, is hard to find.

To spot these occurrences in our own app, we added a Clicked search on the button text “Contact Support.” Check it out:

Bear in mind the above search parameters are specific to our own instance of FullStory. Tinker with it on your end and you’re virtually guaranteed to come up with some salient insights.

5. Are any parts of my site broken & throwing errors right now?

This may be the easiest and most impactful search of the bunch.

Error clicks occur whenever a customer’s mouse click or tap corresponds with a JavaScript console error on your site. This could mean anything from simple type errors to critically important broken checkout buttons. Pulling error clicks is extremely easy. Just start typing “error” in the OmniSearch bar, click “Error clicked anything:”

The resulting Event filter will look like this:

In the process of gathering examples for this very blog post, one of our engineers unearthed a bug that prevented a FullStory customer from being able to open the “Search Everyone” feature. They clicked the button, and nada — no search dialogue popped open like it should have.

It took us less than 10 seconds to find, and it’s no small bug! Now it’s on our engineers’ task queue, and we can proactively reach out to the customer to apologize for their sub-optimal experience.

So what are you waiting for?

If you’re a seasoned FullStory user or just getting started, these searches will give you a boost in productivity. I recommend you head over to your FullStory dashboard and give them a try.

Return to top

Stay up to date with FullStory by signing up for our newsletter.