Female hands using mobile phone and holding credit card

How to serve millennials and Gen Xers with chatbots – Spoiler: It’s hard work

Creating a responsive chatbot is hard work, but if financial institutions want to attract and retain younger clients, it is an absolute must, Mobiquity chief creative officer Mike Welsh said.

According to Mobiquity, only 18% of banks and credit unions have invested in chatbot capability. That means many institutions will play catch as millennials and Gen Z become more significant market forces.

What millennials and Gen Xers want and why

Why must they invest in the technology? Because that’s how those cohorts live, Welsh explained. Do you want to sell them products and services? Meet them on TikTok, What’s App, and Instagram. That’s where they spend their time.

That also creates an expectation of a service level financial services firms will struggle to match.

“Trying to figure out how millennials and xennials are going to save money isn’t as important as how they’re going to interact with the thing,” Welsh said. “Most of these businesses in regulated spaces are terrified of how to do that. Most of them do these weird chatbot things that are garbage. Most people close them.”

Your target customers know when they’re speaking to a bot, so no matter what name you give it, you have to offer more than a smiling face in the pop-up window. They’re comfortable speaking to them if they receive value in return.

Chatbots must be intelligent and transmitted differently

The key to that value is providing more intelligence than many versions on the market, Welsh explained. They need to be intuitive and be available to hand the conversation to an actual person at the right time.

That handoff will come with the suggestion to contact them at a number. The customer puts it in their What’s App so they can talk.

“The next app you download for banking isn’t going to be an app,” Welsh said. “It’s going to be a phone number. It’s going to be inside of the spaces you already occupy. You’re on Signal; you’re on Telegram; you’re in a place where you can interact. 

Mike Welsh headshot
Mobiquity CCO Mike Welsh said understanding how and where younger generations communicate online is a key aspect of a successful chatbot strategy.

“It’s encrypted. It’s more secure than the actual bank experiences most of the time, and you control it.”

Looking behind IVR-based chatbots

Few get it right, Welsh added. Those who do, conduct their business in private spaces, so customers can safely disclose personal information, conduct wire transfers, and even make a down payment on a mortgage.

Neobanks and fintech startups are better positioned to create a positive chatbot experience for a few reasons. They don’t have to deal with legacy infrastructure. Their founders are likely from the generations they’re targeting. Their habits are the same. 

Mobiquity is helping a client create a neobank from scratch. Completely digital, it won’t have branches or tellers. It will have a 25-person call center that is front-ended by an intelligent, warm handing off to an actual human, not more clunky technology.

“IVRs are terrible,” Welsh said. “And chatbots are just text-based IVRs, and everybody knows where that goes. You have sentiment analysis that you can do when people start having interactions.”

Value exchanges are crucial and hard to provide

Mobiquity also works with healthcare clients. Welsh likened the financial industry’s search for a quality chatbot to an intelligent AI system backed by actual doctors. Call a number, explain the symptoms, maybe answer a few questions for clarification, and soon receive value.

That system has successfully personalized your experience, Welsh said. That is stored for the next interaction, making the system more intuitive and predictive.

Banking systems need fair value exchanges to be successful. To create that from a programming perspective takes millions of rows of utterances, responses, and prompts.

Voice and bots share that problem, he added. Creating basic smart speaker interactions require tremendous levels of effort

“I don’t think many banks, many businesses have contemplated the expense, the cost, the maintenance, then factoring and refactoring the regulatory hurdles,” Welsh said. “The blowback on this is if you use it once and it’s not a good experience, you’ll never use it again. You’re not returning for versions 1.2 and 2.5; you’re just never returning. 

“Why? Because Capital One’s got a great app, right? You go in your face ID and see your bills and monthly loan payments, giving you a free credit score. So there’s a fair barrier to entry for anything that isn’t going to be more intelligent, more intuitive, and more built around utterances and read prompts.

And one size doesn’t fit all, Welsh cautioned. If an institution has a Middle Eastern presence, for example, they have to factor in cultural specifics like a preference for carrying large amounts of cash. Does it respond effectively in local languages? Can it verify identities if people are wearing niqabs?

Those few successful systems have a human nearby. The combination of AI and a human is what makes them effective.

How B2B chatbots deliver value

There are B2B chatbot versions, Welsh said. Imagine a small company considering an expansion. The founder tells the bot, who asks her if she wants to buy the building. The bot sees the company payroll, so it knows its size. It reads the lease because it’s paid from the account. It can research ownership, taxes, and price ranges with the address.

“I put that together in an AI, and I say, ‘you know what, what if you bought the building, and then you start to lease your space back to yourself?’ Then you get the opportunity to move to the next floor when that lease is up. “(Perhaps you) use it as collateral to buy the building across the street. 

“These are the things (where) a bot is a great way to get a little bit of info to start; it’s not a way you’re going to manage every day.”

How younger generations are different

Because we’re talking about money, users will naturally have a different level of interaction. Prompts are effective if they take you to the source of an overdraft fee notice.

That’s especially true for younger generations who don’t have the same grounding with money that their parents had. Many are unbanked and operate exclusively through alternative methods like Pay Pal, Venmo, and Zelle. They didn’t have to go with Mom to open a bank account; they didn’t know how the system works because they had never learned.

What they have learned is the interactive system within the popular apps. That forms their expectations. Design your systems with that frame of reference.

The future is coming a little at a time

Imagine someone asking their voice assistant about a forthcoming concert, Welsh suggested. They decide they want to go, so they ask a bot if they have enough money in their account. If so, the bot buys the tickets.

The bot can identify the cheapest site for online shopping and ask if the owner wants to buy from that company.

“That transfer of money happens without you thinking about it, and then boxes will show up,” Welsh said. “It’s going to be where the path of least resistance is for that generation.”

Welsh doesn’t expect high levels of intuition to arrive during his career. He sees places where models can be linked to cases that provide more on-point banking interactions.

“How do you take your gig money and put it in? How do you save for retirement? How do you build a business? How do you grow a business? There are lots of cases where there’s a nice interaction that works for those kinds of businesses.”

  • Tony Zerucha

    Tony is a long-time contributor in the fintech and alt-fi spaces. A two-time LendIt Journalist of the Year nominee and winner in 2018, Tony has written more than 2,000 original articles on the blockchain, peer-to-peer lending, crowdfunding, and emerging technologies over the past seven years. He has hosted panels at LendIt, the CfPA Summit, and DECENT's Unchained, a blockchain exposition in Hong Kong. Email Tony here.