Deep dive into how Airtable built their new onboarding

Onboarding is something that always comes up on growth teams and product teams because it's your first touch with a customer, and it's your opportunity to teach them how to use the product, show them value, and get them to an aha moment as quickly as possible. They then ultimately retain long-term and build a habit with you.

Onboarding needs the right love and care to be a growth engine for your business. It's not a surface area to just experiment with a couple of times. If you decide onboarding is something you want to invest in, you should really go for it.

In general, the way that I would determine if and how you make an investment here is to start with a data exercise, and talk to your customers.

On the data side, identify and study the key moments where your users achieve value or reach that first ‘aha moment’. How many of your new customers are not yet achieving those milestones? Users are likely to retain long-term if they unlock value quickly.

Then I would talk to your customers. Observe their experience when they get started on the product. What are they feeling? Do they feel tired? Do they feel energized? Do they feel like they experienced the magic of your product right away?

Onboarding needs the right love and care to be a growth engine for your business. It's not a surface area to just experiment with a couple of times. If you decide onboarding is something you want to invest in, you should really go for it. Make a concerted investment over 6 months or more that's going to make the experience a lot better for your customers, and give you sufficient return on customer retention and revenue outcomes.

Activation (may also be known as onboarding) is a common investment area on Product Growth teams, and has always been a part of mine.

Roughly, the charter of this team is to help individuals and teams using the product become long-term successful in their initial first couple of weeks, first 30 days. This is two-fold: help users taking actions that enable them to find value and retain better, and to become more sophisticated in service of ultimately building better habits.

If you're thinking about building a team in this space, I would create a charter that's bigger than onboarding and more focused on activation at large. You'll think more out of the box, be more creative, and come up with different ideas for how to do right by your users. These typically span beyond the confines of what you might think of as onboarding today.

Activation is typically an evergreen investment area. It doesn't always sit in the growth org. In my experience, when you bring it into a growth environment, staff it with a bunch of experts, put ammunition behind it to really get it right, that unlocks real impact.

One note is that activation and onboarding are not quite the same. Onboarding is roughly providing a user with the tutorial for how to use core functionality of your product, whereas activation is in broader service of visitation and longer-term retention.

I say this because, if you're thinking about building a team in this space, I would create a charter that's bigger than onboarding and more focused on activation at large. You'll think more out of the box, be more creative, and come up with different ideas for how to do right by your users. These typically span beyond the confines of what you might think of as onboarding today.

Data is so important when you're trying to understand the baseline of an area that you're investing in.

I would start your data exercise around activation by going broad, studying many correlative metrics and understanding the typical behaviors of retained or activated users. In general, I would think of the two axes of interesting metrics to study as (1) visitation or retention, and then (2) sophistication.

In the case of onboarding or activation, you might want to look at onboarding completion rates or retention rates at day two, week one, week two, week four, month six.

You may also want to look at visitation: how often somebody comes back to your product in the first 14 days, how often they come in the next three months based on how often they were there in the first 14 days. Any sort of correlation to retention and sophistication where if somebody does a couple of things in the first week, they're more likely to be more or less successful long term.

I would start your data exercise around activation by going broad, studying many correlative metrics and understanding the typical behaviors of retained or activated users. In general, I would think of the two axes of interesting metrics to study as (1) visitation or retention, and then (2) sophistication.

If you happen to have a product that's multiplayer, like a productivity tool, then there's an element of individual retention and sophistication and also team or account retention. Understand if retained users collaborate more than others, and if one person drives that collaboration or many interact over time. Going deep on these dynamics gives you a foundational understanding of how individuals and teams get started today, and where they encounter hurdles.

One note here is that it always feels great when your users are more sophisticated, but it's also important to meet your users where they are. I would leverage the data you have to understand if a more sophisticated user retains better, or not. You may be surprised! It's okay if users are operating at the beginner levels of your product for a little while, if that’s their comfort zone. Ongoing or progressive education is always a great option for them.

We went into our investments in activation believing that there was an opportunity there because of initial data analysis. This helped us narrow down the mechanisms, patterns, surfaces, and moments where we think we can improve the experience for users getting started.

But that didn’t tell us the qualitative ‘why’ behind users’ struggles with the product. We had to spend a lot of time with customers. In general with onboarding and activation, data can't tell the whole story. This is fundamentally an exercise in helping people learn.

Spend time with customers and understand the why behind what they need and what kind of support is going to work best for them. This is critical in shaping an effective product implementation.

Our PM Bryan Jadot and designer Yana Gevorgyan were the design heroes! They drove this effort, spending many hours watching users get started on Airtable in an unsupervised manner with no coaching or instruction. They learned what they experienced, and how they felt as they got started, seeing what they clicked on and hearing them think out loud to respond to what they were experiencing in production.

We had to spend a lot of time with customers. In general with onboarding and activation, data can't tell the whole story. This is fundamentally an exercise in helping people learn.

That then informed some amazing prototypes and mocks that the team could take back to a new set of customers. This was the iterative customer centric process that we took.

That phase actually required very little data. It was about making sure customers’ needs were very precisely addressed, and that they were really able to learn more effectively. We were meeting them where they were.

One of the things that we learned almost immediately is that using the right patterns and mechanisms to teach users effectively can make a huge difference. Many teams over-rely on the tool tip because it is frequently used by many products during onboarding, but there are much more effective ways to teach a user and help them learn and find value! Often a tool tip can feel sort of like a patch on top of the product.

It’s important to keep in mind that a new user actually is pretty tired, trying to keep track of everything that you're throwing at them. So, if you can be more native and more tailored to what a user needs, there will be compounding effects on your activation rate and also generally positive impact on your customer's happiness and experience with the product.

This is a very good question. We spent a lot of time thinking through how our users wanted to learn and build with databases.

In general, what we experienced is that customers’ familiarity with databases, learning styles, and building preferences were meaningful indicators of the kind of support that they needed.

We tried to be as specific as possible in giving folks a choice of using an example app provided by Airtable versus building on top of a database from scratch.

And that bifurcation helped us bucket folks up front into two camps of learning style or building preference, where one would be much more data forward and custom and the other would be much more of a quick tailoring to get it done and personalize it fast.

Figuring out the best design patterns to help users learn is, I would say, emerging in the world of Growth Design. There are some expert growth designers out there — several of whom I’ve been lucky enough to work with!! — who are starting to really crack what works. You’re going to keep seeing new patterns pop up in new products.

Many teams over-rely on the tool tip because it is frequently used by many products during onboarding, but there are much more effective ways to teach a user and help them learn and find value!

I invested in Dopt because I want to contribute to this ecosystem of new patterns for onboarding and activation. I feel that my growth teams over the past few years have started to just scratch the surface of what’s possible for our industry.

One pattern that Airtable uses today is a guided onboarding set-up wizard:

This is a great example of a directive experience, which a large majority of customers experience after they sign up for Airtable.

The wizard is an education moment where we're comfortable being directive because it's something that the large majority, say 90% plus of customers, will benefit from. In contrast, more optional encouragement to try intermediate or advanced features is not something that's a need-to-have for the large percentage of the base.

The wizard is followed by a ‘mole’, which is a popup from the bottom of the screen, that happens to look like a mole popping out of the ground. It helps with more intermediate and advanced onboarding activities when users are ready to try them:

These patterns work really well because they strike the right balance of being very directive, when everybody should be doing something, and being more passive when someone may optionally choose to level up what they're building.

Yes, it’s a great example. We tried to consolidate all of our old tooltips into one place where you could reliably reference and understand progress, and go back and learn more later.

We felt like the cognitive load for customers of looking everywhere and seeing tips in all different places was actually pretty high, and that we could do the work for them and make their experience cleaner and more enjoyable.

And that's ultimately a win-win win because you help more people succeed and then the customers who need that suggestion are delighted by it. And the customers who don't need it never got confused and are able to just carry on with their work. So in service of giving the customer the absolute best experience, a passive approach can sometimes be very powerful.

This is so important. I see so many onboarding flows that jam everything you could possibly know about the product into the very first session. As a result, the customer’s experience is overwhelming.

We can do the work as product teams to introduce something when it is contextual, when it will be most relevant to you and what you're doing in the product. And sometimes being patient is preferred because you can meet the customer where they are and when they need that help.

There's only so much energy in the user's battery. When they decide, “Hey, I'm gonna give this product a try”, you want to honor and preserve that energy, maybe even give them more energy back, rather than saying, “here are 85 things you need to do right now, and even if it's not relevant to you today, someday it will be.”

We can do the work as product teams to introduce something when it is contextual, when it will be most relevant to you and what you're doing in the product.

And sometimes being patient is preferred because you can meet the customer where they are and when they need that help.

To me and my teams, this means that we see “getting started” as a multi-week, multi-month journey. Most of the education happens in the first 30 days, but many of the tips and suggestions happened beyond and after that first session or first initial setup.

That's really what continuous education is all about, investing in the long-term success of your customer in their journey with you.

As you reflect on your own onboarding experience of the product that you're working on, I would think through which tooltips you push to a user today, in product, in that first session that you think they might not remember very well a month or two months later. Remember that your customers are learning how to ride a bike, and you’ll probably need to coach them through hours of practice over a series of sessions to perfect their skills.

That's really what continuous education is all about, investing in the long-term success of your customer in their journey with you.

Thank you!