Insight/Action/Outcome: The Jupiter team noticed a 14–15% drop off in their onboarding journey when new users had to enter specific information, and the median conversion time was almost two minutes. Focusing on data insights from Amplitude, the team made design changes and introduced autofill options, which led to increased conversions and an additional 300 new savings accounts each day. They also reduced the time for users to complete onboarding by an average of 80%.
The modern world is filled with countless choices. From music playlists to credit cards to insurers, there are a growing number of products and services closely tailored to individual uses and desires.
While an abundance of choices promises to make consumers’ lives easier, it makes personalization more challenging because the cost to switch services is getting cheaper. It’s becoming harder to connect loyalty to a company, and consumers are constantly on the hunt for vendors and service providers who can better suit their tastes.
It’s a tough problem to solve. At , we’ve been paying close attention to these trends, especially as they pertain to the Indian financial services (finserv) market. The more we understand our users, the more we can increase personalization to cater to our customer base.
How Jupiter Money leans into personalization
Jupiter is an Indian financial service startup dedicated to helping customers develop a holistic financial wellness plan. Consumers use our app to access banking-related services, including deposits, credit cards, lending, and investments. The goal is to help our users achieve financial well-being, which allows them to make smarter financial decisions, and build an airtight plan to reach their goals.
As an analytics and AI applications specialist who handles data science at Jupiter, I’ve had the privilege of watching the company grow. One of the biggest contributors to this rapid growth has been our approach to personalization. Finance tends to be very personal by nature, and we’re finding new ways to lean into our personalization strategies, such as:
- Help users establish a goal. We want to help our users take baby steps to build better financial habits. In one example, we encouraged users to grow their savings by increasing access to micro-saving and investing opportunities. We prompted them to save or invest as little as 10 rupees a day or a small amount every week, which changed how people thought about traditional savings methods.
- Nudge users toward self-established goals. Transparency and visibility are key for finance, and we want to show our customers the trajectory of their overall financial situation and where they are on that trajectory. Based on a customer’s current activity and financial history, we can offer specific advice: Maybe this is the right time to start saving for a major purchase, or maybe they have some idle cash they can put to work elsewhere. We can use their information to help brainstorm opportunities to tackle their next financial milestone, whatever that might be.
Our growth relies on personalization and first-party data, and that’s why we rely on .
Getting data in order with Amplitude
A huge part of developing personalization is getting everyone on the same page. Understanding best practices can help your company move forward with fewer headaches. And democratizing data allows us all to speak the same language.
Using behavioral cohorts within Amplitude lets us understand how users experience our product and build specific personas based on that data.
The best way to start is to get your data in order, and Amplitude helps us do that in a few ways:
- Centralize tooling. Standardize your data warehouse and select a platform that provides a single source of truth. Amplitude makes it easy to manage data, which increases efficiency and improves .
- Start with cohorts and personas. Identifying cohorts and micro-cohorts is an ever-evolving process at Jupiter—partly because we’re adding more data and partly because customers constantly change their relationship with our product. Beyond factors like age and location, behavioral cohorts within Amplitude let us understand how users experience our product so we can build specific personas based on that data. We even name our cohorts to make them more meaningful. The closer we feel to our customers, the easier it is to remember that we’re building for them. And when we know who they are, we can do a better job.
- Understand, measure, and democratize data. Understanding data isn’t the same as synthesizing it. We’ve all heard the old adage, “garbage in, garbage out.” If you don’t effectively manage data and standardize metrics at the start, it will be much harder to gather meaningful information and find the stories behind the data. When we started our Amplitude journey, we listed every possible user action and touchpoint and these actions. Then, we shared them with different teams. Our Amplitude dashboards are easy to see and understand and provide internal teams with a holistic view of our users so we can make better decisions.
- Transform knowledge into a cohesive strategy. Now comes the hard part: putting all the pieces together. Making mistakes at this stage can be expensive and stressful. That’s why Jupiter has established two major objectives to guide our strategic efforts:
- First, we will always root for the customer, which means every personalization activity must provide meaningful value.
- Second, we prioritize consistency to ensure the customer isn’t confused at various touchpoints. We consider these guiding principles when determining which elements to introduce and where to personalize further.
With these actions under our belt, we were free to turn our attention to our new user experience.
Data insights improves onboarding and credit card business
Our team observed a significant drop-off in our new user onboarding journey. Looking at the data in Amplitude, we could pinpoint the two most problematic issues:
- Around 14–15% of our users dropped when they had to enter personal information because they didn’t have these documents readily available. Similar technical or user-driven drops were observed at various steps in the journey
- The median conversion time was almost two minutes because of backend identity verification checks, which resulted in a poor user experience.
We also conducted a detailed analysis of every technical issue that occurred during the onboarding process. Focusing on data insights, we took our observations and learnings and engaged in a backward design process, starting with what we wanted the user experience to be.
Our efforts led to great results:
- We made design changes to address the reasons for organic drops and introduced autofill options at appropriate junctions. These changes led to increased organic conversions and an additional 300 savings accounts each day.
- Backend configuration changes helped us reduce the time it takes users to complete the onboarding journey by an average of 80%.
- Bug fixes helped us to improve Jupiter onboarding conversion by five percentage points (pp).
Similar data deep dives led to a boost in our credit card business. We introduced design and copy changes to set better expectations around supplying that information, which led to an increase in credit card approval rates.
Through targeted improvements in onboarding design and backend processes, our team successfully enhanced user experience and conversion rates by taking a data-informed approach to personalization.
Preparing for pitfalls
Personalization is rarely straightforward. Even if you do everything right, failures are part of the process. At Jupiter, we have learned a few things about rolling with the punches:
- Recognize failures early. Learn from them, pick them apart, and then move on gracefully. Failure happens—the quicker you get your bearings, the quicker you can learn from them and start working on bigger things.
- Personalization doesn’t always start or end with data. Seeing the depth of data in Amplitude means we don’t have to rely on gut feelings alone. Still, there’s something to be said for human intuition. The powerful combination of data and empathy can positively influence product development.
- Great results won’t last forever. Users evolve, sentiments change, and personalization isn’t a one-and-done project. Gathering insights and adding personalization will be a continuous journey for any organization.
Users evolve, sentiments change, and personalization isn't a one-and-done project. Gathering insights and adding personalization will be a continuous journey for any organization.
Personalization is easier than you think
Personalization can be a powerful tool, although it can also be daunting for companies that haven’t done it before. Despite the pace of digital adoption, it’s never too late to start, and it’s not as hard as it looks when working with Amplitude.
Everyone has customers, which means everyone has data. Take your data for a test drive and pick out the relevant pieces as you go. Have the discipline and curiosity to learn, and your users will tell you everything they want—so you can deliver the personalized experiences they crave.