The Best (and Worst) Practices in User Onboarding

Every product demands a certain amount of effort from the user to become successful with it. This path towards success is what you call the “user onboarding process”.

While the best onboarding processes are like smoothly paved trails, for many products they feel more like obstacle runs. Which are cool – as long as that’s what you signed up for. If you just want to get on track with the product, however, you’ll likely drop out of the process.

Data suggests this is happening on a grand scale.

The below graph by Quettra shows the retention curve for Android apps. No less than 77% (!) of users jump off before the fourth day after installation.

And, according to Uber’s growth expert Andrew Chen , same goes for desktop clients and websites. It’s clear: onboarding time is crunch time – you either shine or you mess up – and it represents a huge improvement opportunity.

Best practices in user onboarding

Users drop out of the onboarding process because they can’t imagine reaching their goal with your product, think it takes too much effort, or both.

But what does effort actually mean in an online context? Mouse clicks and typing aren’t exactly tiring. Increased brain activity, as such, isn’t either. Claude Messier of the University of Ottawa looked through a heap of studies and found: specific cognitive tasks result in "only a small" increase in energy uptake by the brain. Furthermore, that uptake is “closely matched” by supply of energy from the body itself.

Eventually, our brains do not significantly tire. Instead, we feel a task is tiring because we expect it to be and because we perceive it as unpleasant. That’s why the much-invoked reduction of user effort can be accomplished in surprising ways.

Keep the eyes on the goal

Users differ in their individual goals. Which makes it hard to estimate their patience, the right moment to remove the support wheels, and the point at which you lose them.

One type of online user follows a rather one-dimensional objective. In the case of Dropbox , for example, it’s a certain amount of online data storage space for their holiday shots. They want the onboarding to help them reach that objective as soon as possible, without compromising decision quality.

As the service provider, this calls for a quick exchange of only the necessary information. This concerns your sales copy as well as the user information you ask for.

Dropbox does an exceptional job at this. They outline the product’s value both visually and verbally, and offer users to sign up immediately.

For more concerned users, they offer expandable information fields. Interestingly, Dropbox trusts their initial value communication so much that they leave open the amount of cloud storage users can expect. They avoid technicalities, and focus only on the benefits.

The other main type of users follows more complex objectives, like understanding a software’s use to later reach a higher tier goal with these skills. Raising conversion on their website, for instance.

According to goal orientation theory and Andrew J. Elliot , learning-approach goals branch out into “striving to develop one’s skills and abilities, advance one’s learning, understand material, or complete or master a task.”

These gradual learning goals tend to require more extensive onboarding processes. The danger here is that the product’s value slips out of sight in the shuffle. So, educational sidesteps are okay as long as users see how each helps them reach primary and secondary goals.

"Now, at every step of the onboarding dialogue, the AI assistant not only explains to the user how something is done, but also why it is beneficial to do so."

Making time fly

A user doesn’t abandon your onboarding because he has crossed the five-minute mark but because he feels like it takes too long. That feeling, as we are going to see, cries out for manipulation.

When people believe that time has passed unexpectedly quickly, they rate tasks as more engaging, noises as less irritating, and songs as more enjoyable.

Aaron M. Sackett

Researchers Sucala, Scheckner and David showed that time appears to go by quicker when we’re not focusing on its passing - when we’re ‘busy’. Also, Aaron M. Sackett and his colleagues confirmed that when time seems to fly, we experience a task as more fun . That’s why your onboarding shouldn’t be a boring jog, but an (easy) obstacle run. Gamification helps. It gives that vitalizing bit of demanding context to unchallenging compulsory tasks.

Account setup is such a task in onboarding. Users have to enter a bunch of credentials before the actual usage of a product -- a bland exercise. But some companies have geniusly deployed gamification to push their users through this otherwise boring phase.

Slack has camouflaged account setup as a conversation. They waived fill-out-forms in favor of a more playful approach that also includes the first use of their software. A chat bot asks users for profile information one by one, giving the step a more natural flow.

LinkedIn uses a popular gamification measure by framing account setup with a progress indicator. The compulsory task gets a whole new purpose: to fill that circle and reach the next label.

This method also has the great benefit of making users less aware of time. According to the Endowed Progress Effect theory , people are more committed to goal completion as they experience progression towards the finish line. More commitment means less focus on time. Time seems to pass quicker, making the task more fun again.

So, when engagement and goal orientation is high, measures to make time fly work a bit like a perpetuum mobile once they’re kicked off.

Keep users going with dopamine

As we saw above, effort is relative. It’s all about how much we deem it worthwhile. Rewarding has a great influence on that. Our body’s reward is the release of dopamine, granting us a feeling of euphoria.

Throughout evolution, the hormone most likely had the function to keep species on the track of success by awarding a right “prediction, choice or action.” Neurologist Judy Willis aptly described dopamine as the brain’s fuel for perseverance.

In onboarding, it can be used just like that. Triggering its rush can help reinforce a desirable user behavior. Like completing all steps of your onboarding process but especially those steps that users might otherwise perceive as unpleasant. In this sense, success experiences keep users going.

Among success moments, it’s the early ones that get them hooked. So, it’s a clever move to provide a little tasks in the very first steps. Preferably, it’s one that’s rather impossible to fail at. Generally though, if you make tasks all too easy, your users can lose interest. As Willis (and our body) sees it, “no pain, no gain.”

Language learning site Duolingo (above) confronts users with some low-risk items for a successful start and adds a progress bar that looks easy to fill. Interestingly, that’s even before account setup.

Obviously, there is no reward for no success at all or success that goes unnoticed. So, ideally, tasks are manageable and success is visible. Popup messages (below), completement bars (above) and, later in the process, welcome emails , show users they’ve accomplished something.

Also, tasks in onboarding should be varying in shape and difficulty since the release of dopamine is lowered by tolerance. To provide novel stimuli, the most addictive sites use something Nir Eyal named “desire engines” .

According to the entrepreneur, desire is created through unpredictability. Which is in line with Judie Willis’ statement that dopamine-reward is effective only “when the outcome is not assured.” Building desire engines can’t start early enough, the best sites embrace it from step one of onboarding on.

Keep the conversation going

But users, too, can be unpredictable. Their needs change and vary all the time. To improve your onboarding process constantly, you have to know how they change. Onboarding emails help you with that. Groove HQ’s CEO Alex Turnbull identified three major accomplishments of the onboarding email :

1. “It establishes a relationship between the customer and me (the CEO).”
2. “It helps us identify any unique needs that the user may have.”
3. “It sets the stage for what’s coming.”

Turnbull named Groove’s HQ 28% opening rate for onboarding emails “terrible.” But testing hundreds of onboarding emails, and reshaping them to their customers’ needs, for them, was “absolutely transformative.”

Above visualization shows the process of onboarding optimization I derived from Groove’s experiences. It starts with assessing your best or most active users and learning about their “decision-triggers”, what got them going. With this feedback you refine the onboarding process to make new users your future most active users. Then you start again.

Worst practices in user onboarding

If reducing unnecessary effort is at the heart of any good practice, too much of it is why a practice fails. Here are the worst practices in user onboarding.

Long signup and registration process

All users favor a lean registration and signup process. It’s the part of onboarding they have been confronted with most often and probably with least variety.

Briefly, users just want to get over with it. That’s why it’s so hard to make a positive impression here and very easy to make a negative one. Making it too long is just one way to do the latter. UXmovement found no less than “8 reasons users don’t fill out sign up forms”:

• Fear of getting spammed
• Fear that a Facebook/Twitter sign up will spam followers & friends
• No option to delete account
• Feeling insecure with personal information handling
• Too much work to fill out compared to value gained
• Asking for information users don’t think you need
• Asking for their credit card number for a free trial
• Product/service is not clear or appealing

I’d add to that missing form validation, which can create maximum friction for users when password requirements are quite specific or generally if many fields are to be filled out.

Untailored communication

In customer communication, too many companies make users feel like a mere source of numbers. Most often, this results of bad timing and ignorance of the customer’s individual state in the process.

For instance, the end of a product trial phase seems like the right moment for email penetration. The user has technically completed all onboarding steps before proceeding to a paid product. Let’s ask them to do so. Right?

Wrong. Because a highly active and a mostly inactive user are at totally different stages in their onboarding process. Active user might be ready to go pro, inactive user might need a jump start. Logically, for an already not very active user, an upgrade offer will sound like a joke.

Advanced-level product emails to disengaged users is like asking ‘would you like fries with that?’ before a customer even steps up to the counter.

Alex Turnbull, Groove

That’s why customizing your communication is totally worth the extra effort. With further knowledge about your users (evaluate metrics like login frequency, feature use etc.) you will be able to create personas and automate the process.

Confusing pricing

Pricing is a part of your value proposition. Few users will come to you saying “I want this and that, price don’t matter!” This is not to say that you must have the lowest price out there. But it means that your pricing has to be comprehensible: what do you charge for which product with which benefits for the user.

A proven way to mess this step up is to simply have too many plans. Look at Dyn’s former pricing page to see how it’s (not) done. Too many plans, too many options - a classic “go figure” pricing page. Good to see that by now, Dyn has learned from it .

The right time to introduce users to your pricing during onboarding can vary. Freemium model providers have no reason to lay their cards on the table before a user’s sign up and first activity. But if it’s done like in above example, anytime is a bad time.

As a counterexample, check out Mailchimp's pricing page above. It’s interactive, helpful, lean and most of all, clear. Users know what’s right for them before even crunching numbers.

Clippy

As put brilliantly by Robinson Meyer in The Atlantic :

“Clippit, the infamous Microsoft Office assistant, lived like a firework, or perhaps like a low-flying helicopter: bright, in-your-face, a streak across the sky, and unbelievably annoying.”

Let’s all learn from Microsoft’s doomed creation and never again dump our user onboarding with similarly impertinent virtual assistants.