Increasing Wayflyer's conversion rate by 25% with redesigned onboarding

Wayflyer's onboarding flow was a strict on-rails experience where steps could only be completed in sequence to access the app. This was a simple and efficient solution given that almost all of our prospects came from outbound sales contact and there was little to explore other than the main goal of getting funded. These customers had someone on hand to guide them through the process, so the UI was purely functional and built very much for their target customer in the early days. In order to scale, Wayflyer wanted to serve inbound leads more effectively. With a multi-product strategy on the horizon and expansion into other markets, the journey needed reimagining to ensure it could cope with the scale up.

Discovering problems to solve

To make sure we were working on the right problems, we had a period of discovery research which drew from multiple sources, both qualitative and quantitative. The product manager and I listened to sales and onboarding phone calls, reviewed data analytics from Mixpanel and Hotjar, reviewed support chat transcripts and interviewed sales staff about the problems encountered during onboarding.

We found:

• Customers got ‘stuck’ on steps in the onboarding flow when they didn’t have access to connect something or operated in ways the app didn't cater for
• A step early in the flow asked customers how much funding they were looking for, but they didn’t always know the answer to this and would rather know how much they could be eligible for
• If a customer had more than one store, they had to create multiple accounts to connect them all because we could only support one connection at a time. This caused operational and user headaches
• Accounting system connections couldn't be done in the Wayflyer app, leading to customer suspicion and staff effort
• We asked customers for a lot of upfront effort before they could get an idea of their eligibility for funding
• Parts of the onboarding flow were not mobile friendly and about 25-30% of signups were from mobile
• Customers didn’t always trust Wayflyer, the business model or connecting data sources. This increased the need for Wayflyer staff intervention
• Manual processes like emailing bank statements took up more of the customer's and Wayflyer’s time
• Lack of clarity on process and progress. Customers were sometimes unsure whose court the ball was in and how long things should take

Solving the onboarding challenges comprised many projects that culminated in the overall improvement in metrics.

• Fixing UI bugs and improving the mobile experience with updated components, such as a full screen scrollable modal
• Introducing accounting platform connection as a step in the flow, along with an eligibility calculator
• Updated bank connection flow for international expansion, covering more banks
• New multiple connection UI pattern that allowed us to scale onboarding for new market segments
• Introduced onboarding checklist for flexible step completion and clarity of progress

Getting started

With the problems clearly defined and some ideas on how to solve them established, the first step was prioritising them which we achieved with impact/effort matrices for each step of the onboarding flow.

To help formulate the vision, I learned about best practice onboarding by completing the Growth Design Masterclass and thoroughly evaluated both competitor onboarding journeys and those from all industries concerned with multiple product offerings. I distilled my findings into some key concepts that I could come back to for inspiration throughout the design process.

Outlining the vision

I created a lo-fi Figma prototype of the entire flow as a first draft vision of how we could improve the experience, based on our research. This was really helpful in discussions with engineering and key stakeholders to show the rough direction we wanted to head in.

‍

Improving key steps in the flow

Bank connections

Bank connections were the preferable method of obtaining banking data for the Wayflyer team. Emailing bank statements resulted in higher operational costs and a more time consuming job for the customer. The existing design did not successfully communicate this preference or give the user any clues as to which countries connection was suitable for.

Prior to starting this work, we only had one connector (Plaid) which did not serve all the countries that we operated it, so a project ran in parallel to add more connectors (Basiq, TrueLayer) in order to increase the percentage of customers connecting.

There were two main problems to solve:

• Deprioritising the email option while keeping it accessible for those who can't connect
• Routing users to the correct bank connector for their country once the expanded set became available

How might we provide an obvious route to emailing bank statements for customers who cannot connect, without promoting it as an equal choice for those who can?

Because only 20% of customers were connecting bank accounts via the existing bank connection screen, the business challenged the onboarding team to make a quick improvement to this before tackling the more complex issues of an expanded range of connectors in order to reduce the burden of statements being emailed.

This meant redesigning the landing page to deprioritise emailing statements and clearly displaying up front which countries were available for connection, while also enabling access to an 'email statements' option. After a few iterations based on design critique feedback and a successful usability test, we released the first version of the new bank connection landing page and increased the ratio of connections vs emailed statements.

This was a quick fix, so we swiftly moved on to the next iteration that would incorporate an expanded set of connectors, also expanding the amount of customers that could onboard without Wayflyer assistance.

‍

Given a future of multiple connectors, how might we route the customer to the correct one based on their country, without overcomplicating the connection UI?

Since the new connectors would cover the vast majority of countries and banks, it was now safe to further deprioritise the email statements option in the UI and we didn't need to list out countries as help text anymore as there would be too many. Instead, we could rely on the user selecting the country where they opened their business bank account in order for us to display the relevant bank connector (e.g. Plaid, TrueLayer)

By adding a step to capture the country of incorporation at the start of onboarding, we were able to streamline the banking connection flow by pre-selecting a sensible default, eliminating the need for users to choose their country again in the majority of cases. On selection of a country, the relevant connector and a consent statement would be displayed.

Flexible completion with the onboarding checklist

The existing onboarding flow was inflexible and inefficient because if a user got stuck on a step, they couldn't explore any of the other steps which held up their overall onboarding journey. They might have had access to analytics for example, but needed to invite someone else to their application in order to connect bank accounts. This was annoying for users but also came with risk of abandonment and staff intervention.

To improve the efficiency and flexibility of the onboarding process, I designed an onboarding checklist that we routed users to after the pre-qualification steps of country of incorporation, monthly revenue and funding indication. This aligned with the business goal of supporting a multi-product strategy where once into their dashboard, users could explore other products like TikTok Ads funding or Insights.

By the time users landed on the checklist, they would have completed their first item already (company details) which leveraged the goal-gradient effect, an intentional design choice to encourage users to complete their onboarding.

The checklist also gave us a way to play back entered data to the user in a succinct format and give them the option to make more connections to enhance their application.

‍

The onboarding project was transformative for the business in its scale up following Series B investment. It not only improved conversion rate and user experience but also smoothed out many operational inefficiencies, eliminating several manual processes and allowing many more users to be funneled efficiently through the onboarding journey.

Onboarding was also the perfect testbed for many of our evolving design and research practices such as unmoderated usability testing, Hotjar analysis, prototyping and design critique. In this way, it offered a lot of personal development opportunities for not just myself but the wider design team. I was able to be a key contributor to the emerging design system and guide many decisions around components and design patterns that were needed in the product.

Many of the core concepts from this work are still in tact today, suggesting that they have continued to play a key role in supporting Wayflyer to scale. By the time I Ieft the project, we had eliminated the majority of onboarding related support chat interventions and were moving on to new work around trust and expanding the number of available connections.