Going self-service

At Arcade, our mission is to empower the product-led growth economy. Going “self-service” was not a question of if but when. We will only be leaders in the category if we show a great self-service experience.

In this post, I share why we decided to go self-service, how we knew the timing was right, and specific user research and data interpretation strategies that we did to make the investment. It seems like a lot of companies struggle with this question, so I hope this adds value to your decision-making.

Before I continue on, one quick note to say that people interchangeably use “product-led” and “self-service.” All product-led companies have self-service, but not all self-service companies are product-led. I define product-led as the product and business having a Go-To-Market (GTM) and business model that is driven by quick access to product value and virality. I define self-service as a way to sign up for the product online without needing to go through a human.

‍

‍

Why self-service?

We decided to go self-service for three reasons: it was part of our mission, it aligned with our growth strategy, and we had validated the market’s pain and (part of) the product solution.

I have hinted already about how our mission is to support the product-led economy. We gain a lot of credibility with customers once we can point to our product as a world-class experience for self-service.

Second, it was an extension of our values and growth strategy. Early on, we decided to re-use a page out of Atlassian’s growth playbook — to focus on the end user as opposed to the admin at a large company. We felt that by focusing on the end user will in turn increase our chances of larger deal sizes down the line. We have a large enough market that we can optimize for volume.

Third, we were not keeping up with leads. For more context on our product and business model — to date, 75% of inbounds are driven by organic discovery. We know this because when prospects fill out the waitlist, we ask, “how did you discover Arcade?” 75% of them mentioned seeing an external Arcade or hearing about it through a friend or colleague.

By June 2022, I was still manually emailing new leads with the Arcade link to get started, and it was clearly a friction point for growth. But I was struggling to find time between customer calls and candidate hiring. During our New York City offsite as a team, we decided that it was time to invest in it.

‍

‍

Stages of waitlist - when is the right time?

Almost every beta product or startup starts off with a waitlist. Here’s a vintage shot of our first landing page:

Why did we start with a waitlist? We wanted to have a close relationship with our customers. By forcing them to get on the phone to talk to us about their pain points, as well as look closely at our beta product, we were able to see what the friction points were. We got to know our first one hundred customers very well.

I will argue that there are stages of the waitlist.

The first stage for us was validation. The goal here is to understand and validate if there is a pain point and if the MVP solves this issue. This lasted from June 2021 to November 2021.

Then the second stage is refinement. By now, there was a deeper understanding of why customers use the tool. We added a way to log into the product. Calls became quicker, as the leads knew what the product was. It became more about specific feature requests. I just sent them a link after the call to send them the product. This was from Nov 2021 to Feb 2022.

The third stage is optimization. We decided that we needed to focus on funnel metrics. We stopped taking phone calls and sent customers a link to get started. Then we paid close attention to activation product metrics and retention. We started doing more sophisticated analysis around cohort data and patterns in activation. Here, we did batch processing with cohorts - every two weeks, we sent new cohort signup links after investments in product and growth hack techniques. I go deeper into this approach in this article. This was from Feb 2022 to June 2022.

After a certain point, there needs to be an informed decision by the company between staying with a waitlist strategy vs. a sales-driven GTM (requesting information about the prospect in order for them to access the product).

Around June 2022, we decided that we knew enough about the customer and GTM strategy to make the investment in self-service. We also realized that we wanted to do a larger launch, and it was only going to help our growth if we made the investment now.

‍

‍

Shifting gears to self-service

Once we made that decision, we realized that we needed to do a few things:

• Do user interviews to test how usable our product was without any guidance or context. This was a separate initiative from our power user interviews.
• Capture information in product that we weren’t capturing in our waitlist signup form.
• Transfer information that customers were getting from email, Intercom, and Zoom into the product.

‍

For the user interviews, they were quite straightforward — we segmented our waitlist by persona. We reached out to customers in each persona type and explained that we wanted to give them the product and watch them use it. We then offered them swag if they got on Zoom with us. Most people accepted (as much as I want to think they were thrilled to influence the roadmap, the swag was the carrot). Each one was an hour. The structure was:

• We asked them about their background and use case.
• We then explained that we were going to watch in silence, but they should talk out loud about their thought process and confusion. We asked permission to record the interview.
• We watched them (painfully, at times) struggle through some key questions about how to get started.

‍

Through this process, with dozens of interviews, we discovered a lot of papercuts (such as random buttons that really didn’t belong) and key questions that we needed to address. We included papercuts and bugs within the scope of “self-service” as that polish really up-leveled the product.

‍

There was a lot that we discovered, but some key discoveries were:

• Users were confused about the first action that they needed to take inside our product.
• There were some bugs around navigating the edit page. We did not realize that users were taking certain actions in the product.
• We needed to explain best practices for building Arcades that we were talking through in Zoom calls.

‍

After building a deeper understanding of our first time customer experience, we decided then to build out solutions.

‍

‍

What we did to enhance the product experience

First, we designed a new onboarding flow where we got critical information about who our customers were. This was going to replace our Typeform. We also decided to add intelligent onboarding — changing their first impression of Arcade based on what they indicated their use case was.

Second, we created content in the form of both Arcades and documentation that walked users through how to build a great Arcade experience.

‍

‍

Third, we integrated these within the Getting Started journey.

Fourth, we mapped out a better strategy for what to get users to do with our product and when. We did this primarily through email. This ended up requiring a lot of data research about what actions correlated to active engagement and when we should ping our customers. It ended up being quite complex:

Finally, we thought about moments of delight that would translate to an overall amazing experience. One of my favorite ones is that when you first log in, you can hover over the “Getting Started” card and see your name pop up. It adds a nice touch of personalization.

‍

‍

‍

The impact

Overall, it has proven to be the right investment for us. We’ve seen 10x growth in new inbounds, and improvements in cohorts taking the right actions.

One thing to keep in mind is that we also experienced a higher rate of inbound questions through Intercom. It has definitely put more of a load on our engineering team to ensure that we keep up world class services and response times. Self-service means that we have to be 10x more receptive to customer questions through all channels.

We still have a lot of work to do, but we learned a lot throughout this journey. Going self-service is not a simple switch to turn on — it requires a lot of research, data interpretation, design and product investments. We hope this was useful for you to read through, and feel free to contact us anytime about how to think about your journey, and how Arcade can help!

A quick note to acknowledge the team: Aleks led the design, Nigel did a lot of the data analysis and thinking behind when to reach out to customers and when, and Rich, Charlie and Ben built a lot of the components and features.

‍