As technologists and entrepreneurs, we want to build something new, create from scratch, and bring our creative ideas to the world. As a veteran of two decades in the startup world, I can personally attest to this rinse-and-repeat formula. Rarely do we think about how to help an existing business evolve when the product is either mature or stagnant. Such a task may seem less exciting.
But continually innovating with a well-established brand once you’ve achieved market fit can be just as exciting and rewarding as starting a new business.
The Challenge of Innovation
In August of 2022, after two decades of ownership under media conglomerates, Evite was extracted by private investors and was undergoing a corporate reinvention just as the world was reopening from the pandemic shutdown. The company had committed to a shift in strategy, planning to move away from an ad-driven business model to embrace its premium services and commerce opportunities.
With this mission, the team was swimming upstream against 20 years of historical current. Every system and process, every user experience and touchpoint up to that time had been built to gain page impressions in an ad-driven world, where pageview counts mattered more than a holistic, user-centric design. In addition, Evite was a collection of disparate user experiences across every device, since consistency was not historically a priority as the business expanded and launched new features and new platforms.
Because of this, from desktop to mobile and from Android to iOS, the experience was variable. We had a dozen of permutations of the same product and an even greater number of incongruent individual features running through the system. We were challenged to re-engineer a consistent product across all platforms so that customers could seamlessly move from device to device without interruption.
The Danger of Closed-Loop Data
Our product was so complex and intricate that it prevented true innovation. Over the past 20 years, we had accumulated mountains of both product debt and tech debt. After the strategy shifted away from advertising, we tested often that following year. We ran A/B tests ad nauseam but rarely found any real winners. We were running in place, micro-optimizing within an overly complicated product that we couldn’t move forward in meaningful ways.
Yet, while internal testing data suggested that the product we had was optimal, anecdotal data suggested otherwise. From casual users to power users, from employees to friends and family, feedback all told us that we could do much better in serving the needs of hosts and planners.
But how could we extricate ourselves from the interdependencies and constraints of the product and start anew? Overhauling systems is never easy, and it’s even harder when we must do it while supporting an existing user base of millions of people a day.
Back to Basics: Focusing on the User Journey
With the shift in company strategy, teams were restructured to align with the new mission. Fueled by the energy of new team members and grounded in the domain knowledge of existing teams, we embarked on what would be an eight-month initiative to unify the user experience across all Evite platforms.
6 Steps to Overhaul UX Without Breaking the Platform
- Conduct user research.
- Focus on mobile-first design.
- Test, test, validate, test again.
- Do multi-round user testing.
- Pre-A/B test to validate key changes.
- A/B test the entire experience.
Conduct User Research
We started unencumbered by historical data. Putting aside the current product experience, the team interviewed our core users: the parents that host birthday parties, the families that plan graduation celebrations, and the power users that organize company events. We wanted to understand the end-to-end journey that a host goes through when planning a get-together. This includes the thought process before coming to Evite, the actual experience of designing a digital invitation, and the post-invitation management of the event up until its occurrence.
The results were enlightening when we viewed them against the existing product. Our current offerings seemed to meet some needs, some others less so, and still others weren’t even considered in the current feature sets. The research provided insightful data, particularly by giving a more comprehensive and holistic view of the host journey than we had historically used. For instance, we discovered that, if we wanted to build a good online digital invitation product, that product also needed to help hosts offline as well.
Mobile-First UX Design
Armed with a better understanding of the full user journey, the team went on to conceptualize a brand-new user experience. This model included how to navigate, browse and find the right invitation template quickly; how best to customize a given template; how to easily step through the entire creation funnel to invite guests; and what tools to provide the host once invitations have been sent out.
We deliberately conceptualized everything on a mobile-first basis. Designing for a smaller form factor meant that we had to cut out the clutter and focus on the things that truly mattered. And extrapolating from mobile to desktop, rather than the reverse, also maximized the consistency between each experience.
Test, Test, Validate, Test Again
Once the team felt we had a compelling new experience, we set forth to test and develop the new concepts and designs.
Multi-Round User Testing
To validate our thinking on the new user experience, we conducted several rounds of user testing. Initial tests employed simple mock prototypes. Later rounds involved live, functional prototypes. As we iteratively went through development, the team made tweaks to incorporate key user testing feedback in parallel. Though some feedback was timely, some meant we needed to recode parts of the experience along the way. But by incorporating real-time, field-tested feedback, the result was much higher confidence in the product that we were about to launch.
Pre-A/B Test to Validate Key Changes
Seeing that this was a huge undertaking across the entire company, we also identified some key areas where changes might have a large impact on business-critical metrics like revenue. For these factors, the team explored ways to individually A/B test similar changes in the existing product experience which was currently serving millions of users a day. Recognizing that these learnings were not completely comparable in the new experience, we felt nonetheless that they were adequate proxies for the changes to get an early, directional read on impact. This exercise, like in the case of user testing, built additional confidence in our eventual project launch.
A/B Test the Entire Experience
After three months of concept, design, and user testing — and an additional five months of development — we finally launched our overhaul of the entire core Evite user experience across the web, mobile web, iOS app and Android app. The entire end-to-end overhaul was wrapped in an intricate A/B testing split, where users in the test segment experienced an entirely different, holistic product throughout their user journey. While we monitored the successful performance of the test segment, we maintained the existing experience for the larger control segment of users. In this manner, the business operation was unaffected as we experimented with the rather significant change to the product.
Early data shows impressive performance, and the Evite team is excited to see where we can go next. There is still much more to do to complete the remaining user journey. A unified experience is only the beginning. It is the first step in our new mission of being truly helpful to the host.
Break the Cycle, Not the Product
There may be times when in order to break out of a circular loop or leapfrog the current state of things — especially when we cannot seem to improve on the outcome through endless testing — we must suspend the closed-loop data analysis and introduce some intuition to look at the bigger picture. If nothing else, this frees our mind from known constraints, allows us to ask broader questions, and encourages us to innovate beyond what we have. We did this by taking a holistic view of the user journey and analyzing how helpful we are to the user in real life. This allowed the team to glean key insights on how to build a better product, unencumbered by the existing constraints. But we continued to use data to keep us safe, to validate concepts and confirm beliefs along the way, and finally to measure real impact to the end users.