How AI-powered mobile push notifications accelerate Qure's growth

Retention +26%
Сonversions +46%

About Qure

Qure.Finance is a revolutionary trading app offering social-media style investment insights from global finfluencers to young investors. Qure helps a new generation of investors earn money by following the insights of top-rated financial influencers, allowing them to learn and draw inspiration from their trading styles.

+44%
Conversion growth
+26%
Daily Retention
200%
CTR Uplift

Background

Fintech mobile apps industry has been growing rapidly over the past 3 years (the number of attributions has gone up by 35% since the beginning of the pandemic in 2020), bringing many new innovative products, like Qure, to the market.

However, the strong spike in fintech popularity among mobile users has also made the average cost of User Acquisition nearly 4 times higher over the same period of time. In this new reality, marketers have to put more focus on attracting high-value users to their apps, and make sure they efficiently retain those users who have already installed them.

Taking a new innovative product to a highly competitive US fintech market in 2021 (and being up against Robinhood & Public), meant that the team at Qure had to find a path to fast & sustainable app growth. The team had already established an efficient user acquisition strategy, which provided a reliable flow of new app users every month. This left two other growth variables to be discovered: maximising user retention and user reactivation rates.

Challenges

One of the main goals for the team at Qure was to increase the frequency and length of the app sessions for their users: more time users spend in the app means more chances that they would subscribe, post trading insights, or interact with the content.

Prior to implementing ngrow, Qure were mainly focused on user acquisition during the initial growth stages. On the push notifications side, the team used a few different manually configured push scenarios and targeted broadly segmented user groups with several versions of push message content.

Prior to nGrow, Qure had manual push notifications, which were time-consuming, lacking personalisation and flexibility; delivery time was not adjusted, push texts were not A/B tested.

This approach worked fine for a smaller audience at the start, but as the app had reached a new MAU milestone, it needed to become more robust and personalised to effectively return new users and support further app growth.

In order to drive retention and user engagement, Qure has decided to look at push notitifcations solutions. Being a fast-growing startup in a highly competitive market, it was important to launch the push notifications channel fast, but with the minimum distraction for the team. nGrow was able to meet these requirements with its 2 week no-code integration and full ML mobile push channel automation.

Mobile push notifications are very dynamic and require a lot of manual hard work: you need to test hypotheses, generate new push scenarios, and replace older, "burned out" content. In this sense, machine learning algorithms take on the role of a marketer, automating push marketing and maximising its impact on the business.
Alex Sergeev
CEO ngrow

Solution

One of the biggest problems is developing a complex solution that matches all our needs. Building it from scratch. Finding and hiring the right talent with ML expertise, etc. This was our main difficulty.
Stan Hoody
CEO Qure

The team at Qure reached out to ngrow in the middle of Q3’2021. The overall integration process from start to finish took 3 weeks — including initial meetings to collect the requirements, paperwork, and technical implementation.

ngrow is a no-code solution and it doesn’t require an SDK installation, which simplified the technical part of the process. The team at Qure only needed to provide access to their Amplitude platform via API, so it could start sharing data with nGrow. The data was then analysed with machine learning algorithms and users were distributed into smart segments, based on their in-app behaviour (i.e. likely to buy, likely to leave, etc.).

13 days after the actual project start, Qure were ready for their first test launch. Normally, we also parallel tech integration with content generation. Creating push messages usually takes up to a week on average (including localisation). Running both processes simultaneously helps shorten the time to launch.

Once integrated, nGrow started sending out push notifications to test content effectiveness for each user group. nGrow uses online multivariate testing, also known as the multi-armed bandit algorithm. Leveraging this algorithm for push notifications allows us to achieve up to 20x better results than the traditional A/B test approach. The process is fully automated — AI dynamically shifts the focus on the most efficient texts; content with higher CTR rates is sent to more users, and vice versa.

AI engine demonstrates improvement even from the first push delivery, but the best results are usually achieved in the 2nd or 3rd week when more statistical data is collected. In this case, ngrow reached the “production” efficiency level in 5 weeks.

Outcomes

"Thanks nGrow for automating our push notifications. We see good consistent results from the start, and there is no more need to manually manage our push marketing"

In just 3 weeks since the first notifications were sent, Qure’s push notifications click-through rate has gone up by 2x times, and Daily Retention increased by 26%. On the back of a better user engagement, the number of app conversions has also accelerated by over 44%.

One of the challenges for Qure was push delivery time: US users are based in 6 different time zones. ngrow helped automate push campaigns by sending notifications at the best time of the day for each user, contributing to a better CTR.

nGrow has also helped to streamline push content A/B tests, which maximised the marketing push channel efficiency and allowed to run dozens of tests automatically.

+44%
Conversion growth
+26%
Daily Retention
200%
CTR Uplift