In 2017, we saw large institutions such as Ally Bank, U.S. Bank, and Capital One begin to engage consumers through Amazon Alexa devices, empowering their respective consumers to get fast answers about their finances, receive affordability insights, and pay bills, all via voice. Yet, just because these large institutions made the strategic investment in voice channels, should all financial organizations?
What’s the relevant business case for voice channels at your institution? More sales? Lower operating costs? Improved experience? Increased retention? This is a great question, but what about the value of simply learning about new voice channels? Waiting to learn about, and experiment with, these channels until the market has fully adopted them may put you too far behind to catch up with the pace of voice adoption by consumers.
Voice-first digital interactions are rapidly gaining in popularity. For instance, according to , voice is already used in 20% of all searches, and that market share is expected to rise to by 2020. As voice recognition technology continues to improve in accuracy, and services continue to be added, more consumers will shift to voice engagement.
When institutions focus only on voice technology’s impact to the bottom line when making a decision to experiment, they may forgo an opportunity to learn from these increasing voice interactions, which could in turn influence a decision for future investment, based on real data from users. This focus on learning directly from real voice interactions are why some institutions have decided to get started with voice channels … now.
Town & Country Federal Credit Union was the first credit union to launch an Alexa skill empowering their members to get fast answers about their account balances. “A major driver for us was the learning opportunity to understand voice services for the future.” said David Libby, President of Town & Country Federal Credit Union (TCFCU), “We are always in a mindset of how we can improve the member experience.”
So, given the fast approaching future of voice, the right question to ask is what’s the cost of learning compared to the risk of doing nothing?
Regardless of the size of your financial institution, you can take steps to begin learning from voice channels to intelligently plan your future digital strategy based on real interactions and data, as we discuss in our eBook, ‘‘. You first need to decide which voice channels are most likely to be adopted by your customers and members. According to research by , the voice-enabled smart speaker market is mainly controlled by Google and Amazon, with Amazon controlling a 71% share of the U.S. market. Given their popularity, Amazon Alexa and Google Home are the best places to start.
Next, you want to deploy a low cost, low risk solution that allows you to observe real interactions and gain insight into adoption and engagement. By observing usage over time, you can forecast future usage and cost savings, and make a data-driven decision to phase in more advanced voice capabilities over time. This approach limits the risks and costs of deploying large-scale technology innovations all at once.
This phasing in of capabilities over time based on adoption and engagement allows for a ‘Crawl, Walk, Run’ approach as you ease into voice channels.
When you’re starting out, you want to manage two key obstacles; 1) the perceived security concerns that can hinder adoption and 2) the need to integrate with any internal or external systems.
The “Crawl” step eases consumers into voice technology by providing information that’s publicly available and doesn’t require any authentication steps to prove who they are. It’s important to recognize that anyone could get this information from a website or app or at a branch, but at times simply asking for the information via voice is more convenient.
The information you decide to make available via your new voice channel can be broken down into the following categories:
- Support: this category supplies information that helps people solve a problem such as how to reset their password.
- Origination: this category supplies information that helps people get started with a new product such as a credit card or auto loan.
- Guidance: this category supplies information that helps consumers complete an action such as how to make a loan payment or set up direct deposit with their employer.
- Glossary: this category explains financial terms to consumers in a jargon-free way.
- FAQ: this category supplies consumers with basic information such as a routing number.
Beyond the benefit of learning from real voice interactions, automating responses to the aforementioned categories opens up a new channel of self-service, which can deflect calls from your call center, providing cost savings.
“Our members will be able to get general information from our Amazon Skill, without needing to call or email our employees,” said Trailhead CU’s President/CEO Jim McCarthy. “This enables our Member Service Officers to focus on helping members deal with more in-depth, account-related questions.”
Even a small reduction of inbound calls can have an impact, especially to smaller institutions whose staff must break from regular work to answer any support calls.
Once you’ve introduced a basic voice capability to your consumers, and are seeing promising adoption and engagement, you can begin to add capabilities, such as reviewing recent activity and checking account balances. These are still the most popular financial features of online banking and mobile banking apps.
Thus the “Walk” step entails providing consumers with fast and convenient access to basic transactional information as a logical next step. Naturally, this requires consumers to prove who they are and associate their identify to a voice-enabled smart speaker. This introduces two new obstacles to manage; 1) collaborating with your online banking or mobile banking provider to authenticate consumers through Amazon’s and Google’s account linking workflow, and 2) meeting their security requirements for financial voice apps.
To overcome these obstacles, start by consulting your online banking or mobile provider and determine the following:
- What experience do they have in deploying a voice-enabled solution? If none, determine if they have any relationships with experienced conversational AI vendors.
- What APIs do they provide to authenticate consumers with Google and Amazon’s account linking workflows?
- What APIs are available to empower conversations around financial management? This is important to know in advance if you might support a broader range of financial management tasks in the future.
Assuming you do not use a vendor partner of a current banking provider, once you have determined how to handle customer-verification for security purposes, you’ll need to go through an approval process with Amazon and Google which can take as long as three months or more to successfully complete. They each require you meet strict account-linking security protocols.
For a consumer to request simple transactional information, their bank accounts must be associated with their voice-enabled smart speaker. If not, the consumer will need to be guided through the account linking process, which must be initiated through the web or mobile apps of Amazon Alexa or Google Home.
During the account linking process, additional layers of security should be added. One method used by vendors is requiring the consumer to create a voice pin for their voice-enabled smart speaker that will be associated with their accounts. Once the voice pin is setup, the voice-enabled smart speaker can request this pin after a period of inactivity, usually between thirty seconds and five minutes.
It’s important to remember that voice is simply another channel, and there will be times when using your voice is more convenient in a private home environment.
Once you are observing healthy adoption and sustained usage, you can further expand your voice capabilities. According to the , 71% of financial institutions surveyed considered “transferring money between accounts or making payments” as an “extremely” or “very important” part of their future voice capabilities.
Thus, the “Run” step empowers consumers to transact via voice, such as making P2P payments via Zelle or paying their bills. As said, “I think you ignore this [voice channels] at your peril if you’re a financial institution, because as we get more comfortable and as we have more devices and this becomes habit, it’s clearly what we’re going to want to do a lot of financial transactions on as well.”
This introduces new obstacles that need to be managed, such as additional security controls and deeper system integration.
In addition to voice pins, you’ll want to add another layer of security to protect against unauthorized transactions: multi-factor authentication. For instance, if I want to send $50 to a friend via Google Home, I’ll need to provide my voice pin to initiate the transaction.
An additional layer might require I provide another one-time pin that’s sent to another device I control (like my smartphone) to execute the transaction. You can expect both Amazon and Google to invest heavily in voice identification technologies that make these types of transactions frictionless, but that could still be a ways away to be fully secure.
This type of payment capability involves working with your core or online banking provider. Some core provides such as FIS are making integration easier by providing access through open API programs like Code Connect, ultimately bringing down the cost of getting to market with new technology. However, not all core providers are created equal and you will need to apply pressure from the executive level to drive an integration through.
Overcoming the initial technological hurdles are important, but equally critical is the delivery and support of your new voice capabilities. Below are a few other things you’ll want to consider and control before, during, and after launch.
1. Conversational Utility
One of the most critical aspects to be mindful of is the conversational output of a voice-enabled smart speaker. Take the time to ensure that what you’re communicating reflects your institution’s voice and tone, is actionable, and is free of jargon. One way to ensure this is to consult your call center staff in order to learn their best practices for explaining information to consumers in a way that’s clear and useful.
Additionally, make sure conversational outputs are digestible. Unlike reading from a screen, when using a voice device a consumer can’t control the pace at which they are consuming information. Avoid long and drawn out conversational outputs, as they can feel unnatural and be difficult to remember.
2. Generating Internal Excitement
Getting your staff excited about why you are exploring voice as part of your future digital strategy is a must, as they are a major part the delivery. One unique way to spark this excitement is to give a “taste of the future” and ensure the staff themselves are familiar with voice interactions.
For instance, before any development costs were incurred, the executive leadership at Town & Country Federal Credit Union simply gave away Amazon Echo devices to their senior managers, Board of Directors, and everyone on their digital strategy team so they could experience for themselves the convenience of voice-activated devices.
3. Quality Assurance
Before you roll out new voice capabilities among any group of actual consumers, have your staff test and verify conversational outputs for alignment with your institution’s “voice” and tone of communication. It’s important for you to understand exactly what you’re offering to your consumers, and that it represents you just like a member of your team does.
When it comes time to launch, make sure you have a sound marketing plan that supports both in-branch and digital strategies. One in-branch approach is to place an Amazon Alexa or Google Home device with a placard near tellers who are servicing consumers. It’s a simple way to pique someone’s curiosity about the interaction. For this to be successful, you’ll need to make sure your branch manager can train the staff on how to answer expected questions and clearly communicate the benefits.
5. Post Launch
Once you launch, you’ll want to study conversational logs! Some metrics to measure include:
- Accuracy – how often are you identifying the desired action correctly vs incorrectly
- Adoption – how many people are using your Amazon Skill or Google Home Action?
- Engagement – how many unique people are using your Amazon Skill or Google Home Action? And how often?
As we saw in the past, when a device (like the iPhone) presents a way to access information that’s different, but convenient, consumers will adopt it. With the coming ubiquity of voice-enabled devices, your institution has an , to stay competitive in that market, and most importantly, to make a decision from an informed position about what part voice should play in your digital strategy.