How the use of AI and ML led to become a 100% digital platform, with 3X improvement in customer acquisition

In this Research NXT Interview, Iti Mehrotra, the Head of Marketing and Strategic Alliances at 5paisa Capital Ltd (, discusses how they used AI and ML to become a 100% digital platform with no branch or RM network, and achieved a massive scale in such a short period. 

The consumer is more open or rather expecting different offerings, relevant communication, and a proactive approach.

How has been your experience as a marketing leader in the financial services space, and what’s your observation w.r.t the use of technology by marketers in the financial sector, which is known to be a digital pioneer?

Iti Mehrotra: I have been associated with the BFSI sector for my entire career span of 11 years, and I agree that I have always seen the Financial Services sector at the forefront of using technology. Fortunately, I have been part of the teams that were leading these changes and got the opportunity to closely see the shifts from welcome call to onboarding journeys; from SMS banking to app-based banking; from win-back campaigns to predictive modeling; from one fits all to n=1 approach; from Sales team excels to Sales Tabs and now from centralized call centers to decentralized set-up at every employee’s home with better productivity and tighter security. 

Of course, a decade back, a handful of organizations used to lead the change on the marketing side as making the user adopt or accept change used to be a challenge too. When SMS Banking was launched, it was difficult to convince people to check balances on mobile and not call customer care. At ICICI Bank, I think in 2014 or 2015, we had personalized the website’s homepage with the customer’s name, among other things. However, this was a real concern for a set of customers as to how they were identified without logging in. At that time, we had to stop showing names and opt for subtle personalization. 

This has now changed post JAM and the wallet wave. The consumer is more open or rather expecting different offerings, relevant communication, and a proactive approach. And I feel that the sector is very much aligned or rather ahead of these expectations.

The use of Martech does not limit itself to engagement only; we are also using it to identify the right moments to take user response, thereby generating continuous feedback for future enhancements. is now amongst the Top 3 Investment apps in the country. Additionally, your app is currently available in 4 different languages. How do you manage visitor engagement on the platform with relevant content/products across the customer’s journey?

Iti Mehrotra: 5paisa Capital Ltd is one of the fastest-growing Discount Brokers in India. Since its inception in 2016, our focus has been to enable a seamless DIY investing experience for the customers. We have a ‘look beyond Tier 1’ approach, and mobile is the key to scale. Initially, the focus was on building a unified and better investment experience. Today 5paisa is a Finance Super App with a wide range of products- equity, currency, commodity, mutual funds, digital gold, insurance, loans. Then as you said, we recently launched our app in Hindi, Gujarati, and Marathi, helping us connect with the customer better, especially beginners in the investment space.

Of course, with a rapidly growing base, there is a need to educate, engage, and cross-sell to ensure ROI on the acquisition costs and to get optimum CLTV. We are utilizing the power of ML and AI to personalize the customer experience using real-time behavioral data. We are able to make use of timely, relevant engagement opportunities based on each user’s exact needs. 

From the time user installs the apps, our campaigns kick-off and then keep on modifying based on customer behavior or life stage. Whether it’s the account opening journey or whether drop out during purchase, there are journeys build in to help him continue the course. Touchpoints and behavior patterns are identified to decide whether the user needs to be engaged with a notification or whether he needs a hand holding via call. 

As we all understand, engaging with the customer cannot entirely be the play of AI/ML. To set up campaigns, marketers need to understand the customer and the underlying enablers for the demonstrated behavior. A team of data scientists helps us see the trends and behavior patterns to develop our own insights into the user behavior, a UI/UX team helps us redesign a simple screen removing the need to do a campaign in the first place, product manager plugs in a solution that reduces a five-step process to two. The communication specialist speaks the customer language to improve response rates. 

The use of Martech does not limit itself to engagement only; we are also using it to identify the right moments to take user response, thereby generating continuous feedback for future enhancements. has an extensive subscriber base of 6 lakh retail investors and high volume traders, and all of that on the online platform. How do you segment your customers, and how are you ensuring personalization at this scale?

Iti Mehrotra: We are a 100% digital platform with no branch or RM network, and this has been truly instrumental in achieving this scale in such a short span of time. More than 90% of customers prefer trading on the app.

For us, personalization begins at the lead stage itself, where we identify the user’s state, and the prospect is mapped to the agent who is likely to speak his language in case of a drop-out.

Using a combination of factors like life stage, product stage, vintage, networth, and so on, the personalization happens in various forms:

  • Offerings: Base and Premium Pricing Plans are available for customer segments, and features are bundled keeping in mind the need of the customers in the respective segment
  • Look and feel: Customers with upgraded plans have different color schemes to give a differentiated and premium experience
  • Communication: Propensity and affinity models are in place to suggest the right product at the right stages
  • Offers: The offers for up-sell and cross-sell are extended basis users past interaction with the product and the propensity to buy 
  • Language: As mentioned earlier, the 5paisa app is now available in 4 languages, which helps us speak customers’ languages during his entire interaction. This information now can further be moved out of just UI/UX experience and extended to the campaign communication, thereby improving the response multi-fold.
  • CTAs in Campaign: Young and higher vintage customers are likely to stick to self- help channels and prefer no calls. Alternatively, beginners or clients in older age groups prefer call support. These considerations are taken into account while building journeys

We are not limiting the use of AI/ML to just customer acquisition or engagement. Customer service is another aspect that we feel can be a game-changer in the long run.

What kind of AI enablement are you leveraging in your overall marketing strategy? 

Iti Mehrotra: We use ML, and AI capabilities across our legs of performance marketing, be it search or display campaigns or for fraud detection or lead assignments. On the search side, ML had resulted in 2.7X growth in leads at 42% lower cost. This, coupled with the investments on the LMS side, led to 3X improvement in acquisitions. 

On the customer engagement side, as I mentioned earlier, we have a stack of automated affinity and inaction campaigns set up. So we are able to retain a customer if he drops out from a purchase or propose to him Product A as we know that other customers like him prefer this same product. A combination of such campaigns has helped us improve client engagement leading to ~40% contribution to overall cross-sell revenue and an 8% improvement in client retention. 

However, we are not limiting the use of AI/ML to just customer acquisition or engagement. Customer service is another aspect that we feel can be a game-changer in the long run.

Understanding the queries: Sentiment analysis is something that most of the brands have been doing. However, we feel that it is essential to have a real-time understanding of customer queries and complaints, especially from play store and app stores, with an index to indicate severity. Such real-time analysis can help us flag issues in time, understand customer exact pain points, and quantify them. We are doing a POC in this area to have actionable insights for the product and customer service teams. 

Solving the queries: As a discount broker, the focus is to enhance the DIY eco-system. While self-help channels are available, there can be a lot of queries that do not require human intervention and be responded to immediately. We have partnered with Haptik and launched a chatbot for our clients with a vision that users can access support instantly while on our platforms. 

What kind of themes have you observed w.r.t the change in customer behavior in this volatile market, and how are you supporting your customers to make the best decisions?

Iti Mehrotra: Traditionally, retail investors tend to buy at highs and panic and sell at low levels. However, in the last few months, retail investors have not lost their appetite for buying stocks at lower levels. This behavior is a combination of many factors like more awareness and access to information, millennials having higher risk appetite, and the fact that the customers are at home with slightly more time at hand to research the right stocks and opportunities. 

When the investor is showing drive and passion for investing, the onus is also on the brokers to provide a favorable eco-system. At 5paisa there is a range of features available to support the customers:

  • Online and free account opening: To encourage retail investors, we have ensured that the account opening is a 10 min process an absolutely free
  • Stable systems: The rising customer and trading volumes have created pressure on online broker systems. At 5paisa Capital, we have invested in scalable backend so that users are not left stranded on high volatility days
  • Economy Updates: We understand that users need some insights into the economy, sectors, and markets and we have started covering these in our weekly content doses
  • Stock and MF recommendations: Regular recommendations on stocks or mutual funds are generated to help users make investment decisions. We also offer smallcases on our platform which are basically theme-based stock baskets
  • Research tools: For intermediate and advanced investors, we have in-built research tools that can be used to evaluate short and long term investment ideas
  • Live Webinars: We conduct a weekly webinar with experts on a range of topics from investment climate to sector performances to algo trading to tax filing.