Three Years of Building Investing Products — Lessons Learned

Photo by Christian Erfurt on Unsplash

An entrepreneur’s perspective

My journey in Fintech started in 2016 when I had an idea to make investing less daunting and more relatable but didn’t know how to get started.

After a few days of non-stop hunting for information, I realized how close and opaque things are in the investing industry. I would come across terms like Broker-dealer, clearing, settlement, custodian, an investment advisor with minimal information to learn more about them.

Coming from the tech world, where we share everything from building more fault-tolerant systems to scaling databases, it’s surprising how little is shared in financial services. I tried books, Investopedia, Google and they don’t come close in unveiling how the current infrastructure works.

As a founder wanting to build our MVP, I wasn’t really after understanding all the plumbing. What I really after was a simple, powerful APIs that will get the job done. To my surprise, there weren’t many good options.

I spoke to a few folks who have some experience in this space; they referred to some of the tools they were using but those tools weren’t suitable for our use case. We were building a complete digital investing experience — No paperwork — No manual reviews.

Later, I started looking at the disclosures documents of some of the robo-advisor apps in the market and came across a solution that consistently powered most of them. I quickly visited the website and started exploring. Unfortunately, I didn’t find the information to be useful, mostly vague, and then hit another wall.

Show me the APIs…

Through this experience of research, I knew there is so much I needed to learn — APIs, regulatory requirements and decided to become an insider by joining a major financial services company.

An insider perspective

To expedite my learnings, I decided to become an insider and joined John Hancock, where I built and led the engineering team for the Digital Advice business unit.

It turns out, most of the IT at incumbent companies are still running on mainframes that are not interoperable and flexible. Thus, the majority of the financial services giants need a modern solution to build next-gen apps as well. It’s an unsolved problem and there is a universal need to re-platform and hire tech talent in the financial services industry. But, the regulatory hurdles and upfront capital requirements have detracted innovators from entering the space.

Investing is also a highly fragmented industry — often, a product is built by integrating multiple third-party solutions that add to the speed and agility of the business. There is no one complete toolkit for investing compared to “Stripe for payments” or “Twilio for communications.”

Working at an incumbent company, our choices to build modern investing experiences were limited. We couldn’t use any of the internal tools and had to look outside to build a fully automated investing app.

Experience building with legacy Providers

1. Sub-optimal user experiences

Apps today must have instant gratification, and delivering the “aha” moment as quickly as possible is critical. We struggled with this due to the constraints of the legacy APIs.

For instance, we would lose roughly half of our customers that get stuck in the onboarding process with our provider’s KYC system, which added higher support volume. Manual reviews and the multi-day process resulted in a significant drop in account conversion. The other half that did get auto-approved, it will take us 4–5 business days from creating their brokerage account to fully investing their money. Not to mention, we couldn’t open brokerage accounts in non-business hours.

It wasn’t just the onboarding, but several ongoing issues with overdraft fees, 7–10 business days to withdraw money that negatively impacted our app ratings and NPS. It suffices to say, all such events made our users unhappy and kept our engineering and support on their toes, the time that could have better spent on improving product and delighting customers.

2. Poor business and developer experience

Not only the APIs are hard to use, but they are also very fragile and often break. This added additional overhead for our engineering team to patch the systems with a lot of defensible programming. Random downtimes during business hours and extended planned downtimes on weekends are common, and they continue to wake our engineers in the middle of the night.

3. Inflexible and less expressive APIs

Legacy APIs are also very inflexible, often what seemed like a boolean feature in the back-office end up becoming a month-long feature because we had to work around their systems and do the most of the heavy lifting. Thus, further impacting our velocity to iterate and experiment with new features.

The ongoing high operational cost, reduced developer productivity created a low NPS for developers and operations team.


Today, with the proliferation of cloud computing and opensource, launching a web business takes just a few days. Accepting payments takes hours. But starting an investing business takes months, millions of dollars, and regulatory approvals.

Having experienced the pain point first as an entrepreneur and later as an operator building on top of legacy tech, I embarked on a personal journey to help build a modern back-office that enables developers to delight users, increases business’s speed to market, and reduces total cost of ownership.

Whether you are an incumbent looking to modernize your back-office or a fintech startup looking to add investing feature in your app, I’m happy to help. If that’s you, you can find my contact info here.