User Onboarding
Header image featuring woman eating, sleeping, and working out

This is page four of a handbook on Growth Hacking. Begin here.

User onboarding is critical

Welcome people into your product in a way that excites them to become lifelong customers. Excitement is achieved through delightful onboarding experiences.

You could have the world's best landing page, but if your onboarding experience is poor, people fall off before actually buying from you.

"Onboarding" is the user's experience between:

This experience consists of two phases:

These phases are unique to new users. New users must be treated differently than established customers. 

On this page, I cover the first phase. You'll learn:

A brief warning

Most SaaS apps lose 95% of their new users after 90 days. That is insane.

When you lose a user's interest during onboarding, your first impression is complete, and they typically never return.

During onboarding, nearly everyone is skeptical of your product. Consider how many services they sign up for each month. And how many services waste their time with marketing tricks and poor products.

Yet, onboarding is usually the most half-baked aspect of SaaS products. You get spammed with tooltips then dumped into a boring dashboard. 

Why? Because there’s often no individual on a tech or marketing team who's responsible for it. Either there are marketers acquiring users or product managers developing features. No one is sitting in between to onboard those users.

But, onboarding is a feature unto itself. It warrants resources dedicated to its thoughtful design, development, and optimization.

Onboarding strategy

To create an onboarding experience that helps people fall in love with your product:

  1. First, identify your product's value to the user.
  2. Then, map out the user's journey toward that value.
  3. Identify the obstacles that exist along the way.
  4. Address them.

You address obstacles by making them:

Before we learn to address obstacles, let's learn to identify them in SaaS products.

Obstacle identification

Most obstacles will be obvious to you. They are the steps in a user's journey that:

But some obstacles are hidden. To identify hidden ones, record users using your app: You can literally record people's browser activity using a tool like Full Story

Watching recordings reveals — in context — the interaction patterns users fall prey to when using your app.

Here's an example of a bad interaction pattern: A user often creates then immediately deletes new projects. Hmm. 

What are they really trying to accomplish? Was there an intermediate feature they wanted to momentarily access? If so, how about extract that feature to stand alone so they don't waste their time with fake project creation?

You've found a hidden obstacle — unnecessary friction that analyzing analytics data alone would have been unlikely to reveal.

More broadly, user recordings reveal where people spend more time than they should. And what causes them to get stuck and give up.

Study as many recordings as you can and write down the obstacles you identify.

Getting more value from user recordings

Removing obstacles is step one. Step two is funneling users into the best experience.

Configure Full Story to track purchase events so you can compare the average onboarding journey of users who purchased versus those who did not. 

Then, optimize your onboarding flow to usher more people toward mimicking the behaviors of those who purchase.

Example: Twitter

Twitter studied their users' onboarding behavior and discovered that if a new user doesn’t Follow at least a handful of Twitter users immediately upon signing up, they’re much less likely to return in the near future.

Twitter therefore redesigned their onboarding to force users to follow a minimum of five people. And they ensure this process is low friction: They show you celebrities you're likely familiar with from topics (e.g. sports, movies) you indicated you liked.

Recordings versus analytics

When identifying obstacles, I recommend watching recordings instead of cherry picking analytics data because analytics lack the context of the user's full experience.

Further, analytics often require the pre-configuration of tracked events a priori. This leads to false conclusions, and it reduces the chance of discovering the unexpected.

So, begin onboarding obstacle identification by watching a significant sample of recordings. Then, refer to analytics to support hunches you acquire from recordings.

Addressing obstacles

We have an idea of the types of obstacles we're looking for. Now let's address them.

As a refresher, our goal is to make onboarding enticing and frictionless. And, optionally, to make it educational and productive when appropriate.

I. Onboarding should be educational

When someone signs up, the in-product experience must immediately answer:

How do I get value out of this product? 

If the answer isn't self-evident, elucidate it with one of these assistive experiences: 

Remember, these assistive onboarding experiences should concisely answer:

How do I get value out of the product?

If your product is dead simple and does not require user education, simply skip the educational experience. Don't waste people's time giving advice they already know.

II. Onboarding should be enticing

Onboarding should never be just educational or enticing. It should be both.

When you educate new users, entice them with value props so they're willing to put up with the education. Tease them with how amazing life will be once they are done.

To do this, start by identifying the magical moments in your product — where the user gets maximum value. Maybe it's receiving a payment, finding a date, or killing an enemy tank. Then, visualize these outcomes during the onboarding experience.

For example, look at how the scammy "dating" site, Ashley Madison, does it. They visualize your end goal (“finding women”) through blurred background photos while asking you to painstakingly enter your profile details. 

They're teasing you with their magical moment while putting you through a slog.

This principle extends to your entire customer journey: Don't ask someone to do something until they’ve previewed the value they’ll get from doing it. For example:

That's how you get people to take action.

Write out your pitch if you have nothing to visualize

If you have nothing enticing to visualize during onboarding, either short-form or bullet point the value they'll get. 

For example, if you’re a service that helps people quit smoking, tell them how many other people successfully quit smoking using the service: 

“About 15% of those who finish registering then complete the course. Don’t be one of the quitters, because if you complete this program, you have a whopping 85% chance of quitting smoking for life. This is your opportunity.”

III. Onboarding should be low friction

The previous two onboarding principles — education and enticement — matter most at the very beginning of your onboarding experience. The next two principles — friction and productivity — matter equally throughout the entire experience.

Let's begin with friction.

During onboarding, it must be clear what the next step is:

Okay, I understand how to use the product. But, what should I do first?

Not only should each next step be obvious, it should appear low-friction: it must be easy and quick to perform.

That's how you build momentum that propels users toward the end of your journey.

Start by listing every action users take toward purchasing. For a website design tool, users might follow this path:

  1. Enter their email and password.
  2. Create a sample site.
  3. Fool around with the design tool.
  4. Invite team members.
  5. Fully build a site.
  6. Enter their credit card. 
  7. Publish the site.

Next, reduce friction at each step by applying these principles:

In summary, remove friction along the path to purchase.

IV. Onboarding should be productive

The lengthier your onboarding, the more important it is that it doesn't leave users empty-handed at the end of it.

For example, if you’re an email app, have your walkthrough guide the user through cleaning up their inbox. This way, users accomplish something they care about while learning to use your app. It leaves them with a small dopamine hit. It delights them.

Another example: If you’re a project management app, have your walkthrough leave users with the scaffolding required to share their project with team members.

Nothing is more validating than receiving a product's value within minutes.

Onboarding examples+

Webflow

When I was VP of Marketing for Webflow, we developed the following onboarding.

First, we learned who the user was so we could customize their experience:

Next:

  • Once a user completed registration, we threw them right into the design tool so they could get their hands dirty experiencing the power of our app. We didn’t want to intercept this experience with a long, boring video or tutorial.
  • But, after 30s, we knew they’d get overwhelmed by the complexity of the interface, so we forced open a menu to automatically reveal our video course. Each video walked you through the steps necessary to becoming a power user. 
  • The very first video was just a 60s timelapse of a site being built in Webflow. This showed the power and simplicity of the app. We were visualizing the end goal. And the video was enticing to watch because you saw a site go from blank canvas to beautiful, full-featured page in just a few steps. 

In other words, we gave them an immediate tease of the product's value. Then we capitalized on their intrigue by introducing education.

UserOnboard.com

For more examples, I suggest going through the onboarding examples at UserOnboard.com. Kick it off with this one. It's one of the best I’ve seen.

Here's the point

Virality

I'll conclude with a brief discussion of virality. Because the triggers for virality surface during the onboarding experience:

It's time for this page to focus on the latter: Consider how, as the person being invited, you didn't come to the product because you necessarily had prior interest in it. Rather, you were prodded by someone into checking it out.

This means you're even more out of touch as to why you should bother using the app. The app's onboarding experience is therefore doubly critical.

Consider how a referred user should not be directed to a signup page that immediately instructs them to claim their referral reward. First, pitch them on why they should even care to use the product and claim the reward. 

Types of virality

To get users to invite others and to compel invited users to purchase, focus your efforts on the most appropriate viral channel for your product.

There are three types of viral channels:

Inherent virality

Inherent virality is the best virality for explosive growth. It's possible when either:

Your product either falls into one of those two categories or it doesn't. If it doesn't, inherent virality is not going to happen.

The power of inherent virality is you don't have to artificially incentivize users to refer — they're going to want to do it on their own.

Word-of-mouth

The second best virality is via word-of-mouth. It's a measure of how good your product experience is. And it's proof that product experience is growth marketing.

Many successful companies have never spent a dime on advertising. Instead, they built something people couldn't stop talking about and sharing.

Word-of-mouth's potential for your business is a function of how large your market is. If you only service 10 customers per month, you are unlikely to see runaway word-of-mouth. The law of large numbers isn't on your side, so you'll need to introduce artificial virality: coerce your few customers into referring at a very high rate.

Regardless of where you fall on this spectrum, every company should be earnestly pursuing word-of-mouth virality.

Artificial virality

Artificial virality boils down to knowing when to prompt for it and knowing how to best reward users.

Here's when a user is most receptive to referring others: when they've experienced a dopamine hit from your product. This happens when they get value from the product.

During these moments, make it frictionless to invite others — and consider providing an incentive.

(A common incentive is the dual-ended reward, in which both the referrer and the referee are rewarded. Usually this takes the form of account credits or a discount.)

Artificial virality has modestly succeeded for some companies. But it's rarely a slam dunk. Because most of your users simply don't care about earning a little bit of cash. They didn't sign up for your app to make a few bucks. Plus, they're jaded by the flood of over-marketed offers they see on the Internet every day.

So, instead of offering cash, I recommend doling out rewards in the form of your product's primary value proposition. Meaning, provide greater access to the product.

For example, when referring someone to Dropbox, you're rewarded with more gigabytes of storage. Storage is a more tangible, more immediate, and more relevant value proposition to you than getting $25 in credits to redeem at a later date.

After all, you signed up for Dropbox because you need to store files. Leverage this user need instead of waving a couple hours' of minimum wage in someone's face.

Even if the cash and the storage are equivalent in value, our minds favor the storage.

However, if you don't have a product that can be doled out in chunks (e.g. GB's, videos hosted, matches swiped), the cash reward needs to be significant

Otherwise, most people aren't strongly motivated by an offer for a bit of cash.

Optimizing virality

When measuring and optimizing virality, focus on three metrics:

If you multiply the last two numbers together, you get your Viral Coefficient. A viral coefficient above 1 is an indicator of viral potential. If you couple this with a short lag time, you experience explosive growth.

And if users are paying you instead of just signing up for free, congrats.

Next page

Ad channels

How to choose the right ad channel (e.g. Instagram, AdWords).

Or, choose a page using the menu at the bottom of your screen.

Next page →