Messaging, Bots and Conversational Apps
Chatbots, Alexa Skills and Messaging, Why Now?
Today its about messaging platforms. Messaging and conversational interfaces within well known messaging platforms such as Amazon Alexa, Facebook Messenger, Slack, Skype, WeChat, Kik and Telegram allow companies to chat with their users. These conversations allow the customer to achieve her goals in a format that was previously not available. These dialogues, powered by bots or conversational apps, let users share intent, provide context, and complete tasks all within the messaging platform. For example, as a user chats with another user on Messenger, once a location or address is shared, Messenger will offer a link for a ride (starting a flow to hail an Uber or Lyft for example). The user could then request an Uber directly from the messaging app without leaving the conversation.
In the ’90s, the web emerged as a new channel for companies to connect to, do business with and create value for their customers. It would be difficult to find a company that does not have a website or maximizes the use of the web to interact with their users and customers. An online presence is a must.
In 2007, Apple introduced the iPhone and subsequently opened up the App Store to distribute new functionalities via a new channel around mobility and mobile applications. This new era opened up huge opportunities for companies to create easier and more context aware interactions. Today, we don’t leave our home without our mobile phones as we use them to communicate, transact, work and play. While there is still plenty to be done in mobile to enhance user experiences, messaging platforms have emerged that alter the way businesses can communicate with their customers.
The mobile home screen has limited real estate and competing with other mobile apps to gain the user’s attention is quite difficult. Users think twice and assess the service value before downloading another mobile app. The app install and usage fatigue that some companies are experiencing, especially with younger demographics, has triggered the opportunity to engage them in this new conversational manner.
Additionally, there is research that shows users are spending more time in messaging applications compared to other traditional social networks. No matter what demographic group you belong to, probably not a day goes by without you texting or interacting with someone via messages. Individuals will start to expect a similar experience when chatting with companies and brands and will invite these brands into their personal space, inside their chat app, where most of their activity happens. This trusted space is precious.
Position yourself 5 to 10 years from now, as a user you will most likely want to interact with trusted brands through natural language interactions, so that they can help you fulfill your needs. Companies need to strengthen these natural language capabilities in their products, apps, and bots to allow their users to communicate with them with ease. There are multiple touch points and considerations worth exploring. Let’s take a deeper look…
Key Channels for Messaging
Facebook acquiring WhatsApp for approximately $19 billion signaled a strong strategic direction around messaging. When reaching users, a brand needs to understand where individuals hang out. The platform choice is an important early decision. This is similar to how brands and engineering teams initially opted to launch their products on iOS, then Android and other mobile OS platforms. In order to reach user conversations today, brands will need to decide which platforms to target and build on.
During the F8 conference in San Francisco in 2016, Facebook showcased the 1-800-Flowers experience launched on the Messenger platform. The brand decided to launch on Messenger given the platform’s reach (900 billion users) and strong support from Facebook. 1-800-Flowers is very focused on customer support and maintaining a relationship with their customers, so they jumped at the opportunity to be one of the first in the space. In terms of app discovery and approval, Facebook requires an app submission where they will go through a review process. While its undeniable the Facebook has worked to diligently to bring about the rise of chatbots for their Messenger platform, adoption has been hindered. In some cases by the chatbots themselves. But in other cases by challenges not within the control of botmakers. In many ways those challenges have given rise or opportunity for other platforms.
Amazon Echo and Alexa Skills and Google Home with its Google Actions have penetrated the home market. Today users can use the most natural form of communication their voices to connect with services. And while these ecosystems are young. Its apparent that their ease of use will will encourage more adoption and more Skill and Action development.
Enterprise apps might be better served by a team collaboration platform such as Slack, who also opened up APIs to create bots and applications that run on its messaging platform. Slack bots are typically geared towards team collaboration use cases and business / enterprise needs, that can be tackled by the users working in Slack continuously throughout their day. Slack bots can be created and used inside a team, but require Slack’s review and approval prior to being posted in their app directory.
WeChat is the best example of a more mature platform. Strong in China, where users have been chatting with companies and brands for some time now. Today users go to WeChat first to find a business or engage in a transaction (these business accounts are called Official Accounts). The platform reduces friction for them handling users’ identity and payment credentials. WeChat has not aggressively pushed their platform in the west, but the adoption and usage is a guidepost of what is to come.
There are also other tools, that we categorize as connectors, that allow the publisher to push functionality into multiple channels including SMS. As a decision maker, you need to remember each platform is different. For example, Messenger is able to handle cards, buttons and images, while Slack can handle slash commands. To maximize the user experience on a specific channel, companies should get to know the platform’s capabilities well to be able to extract the most value out of it.
Top Use Cases for Chatbots and Messaging
We can think of messaging platforms as the inbox and notification systems for our lives. As further functionality gets channeled there, the more these platforms will become our go to place to communicate, sort out errands, perform and track tasks, and much more.
The notification use case is a clear one to start. For example, my teenage kids open their email every few days, but they are constantly chatting inside their messaging apps. They, along with their generation, are great candidates for receiving delivery notifications, acceptances, requests or other notifications in their messaging app. Similarly, other types of triggers could be channeled this way for them.
Any process that can combine human and machine intelligence, is a ripe early candidate for messaging platforms and bots. For instance, imagine a travel-booking transaction flow, the user engages the bot within the platform to determine travel alternatives, but gets to a point that merits additional human intervention, lets say for handling special baggage or travel credits. However, once that is completed, the human agent pushes the user back to the bot to close the process with an automated payment work flow or special deals. Starts with a bot, hands off to a human, returns to bot to complete. This is being done today.
Brief human interaction replacement processes also fit this model well: customer service, placing a reservation, troubleshooting, quick guidance question, and so on.
Form replacement also works when the user needs guidance for completing fields in a form. We are used to completing forms online, and putting information again and again in different sites. The more you use a messaging platform, the more info it will have on the user and therefore the faster it can share data with the service you want to interact with. For example, in sending a gift to your home for your spouse, the only thing the user will need to do, for now, is choosing the gift. Spouse’s name, address and payment info, all would be handled by the platform. This represents a dramatically different interaction relative to buying online, over the phone with a human, or through a classic mobile app.
Coaching is another area where we see good bot cases (e.g. fitness/health industry). Bots can provide personal coaching as, unlike a human coach, they are always with the user -day and night. A user could rely on a personalized coach to follow a fitness plan, through regular exercise or a dieting schedule. The coach could be a bot that is always on, on their toes, keeping tabs on next steps and tracking activity.
As with the form completion example, structured onboarding processes are popping up as an appropriate use case for conversational interfaces. Onboarding to a new system, or to a new community, or to a new company, could be handled by a conversation with a bot that will capture the information it needs, auto-filling the rest based on historical understanding, and pushing the process forward, diligently.
Food ordering and delivery works well as a replacement to placing an order on the phone or going through a website or mobile app to get the order online. As you can see, the list can go on and on.
Bot Design – Starting Points
We need to remember that an automated conversation or bot is a new way to interact with users. As humans, we are used to communicating with other humans. We also interact with systems and software and these interactions have been limited to the command line (programmers and systems admins), point and click (web environments) and touch interfaces (mobile). We are now connecting with systems in natural language, both in text and with speech (think Siri, Echo, your vehicle) and sharing our intent and context. Keeping the user in mind is of utmost importance and these tips should be remembered when designing a conversational interface.
- Pick an initial domain that is narrow enough to provide value to the user quickly
- Compare alternative ways to fulfill a similar request to validate your approach
- Define whether the process can be completed with a bot or additional human-assisted efforts are required (human/bot handoffs)
- Come up with the speediest happy path in the customer journey towards completing the request
- Include common guidelines and tasks that start to appear across multiple bots such as: start, help, stop, new, goodbye, etc
- Maximize specific platform capabilities for the user flow (i.e. Messenger has cards and buttons, while other platforms have slash commands like /start for Telegram)
- Consider cases for a conversation recovery and flow recapture
- Treat the relationship with precious care, spamming will definitely not work when a user allows a brand to enter their personal messaging environment
Conversational messaging is different than more structured website or mobile app experiences. In a messaging environment, a user can explicitly share with the company what they want to achieve, what item they want to buy, and why they want to do it. This provides brands a unique moment in time to interact with a user in need. The potential unstructured nature of that conversation can be used for the brand’s advantage.
Of course, when developing for voice you can throw all of the above out the window. Check out this story to get a better understanding on our thinking for building Alexa Skills and Google Home Actions.
As someone who guides your company’s initiatives and helps with brand creation and positioning, defining these brand and user interactions is extremely important. In the same way prior to publishing a mobile application to any app store, you look for internal consensus on approach, brand voice, user interface and goals, a similar approach needs to be taken when building bots. The bot character, persona and voice needs to be designed to match the brand, further strengthening the relationship that you are creating among users of your product or service.
We recommend tackling a bot project in a similar way as you would when launching any other communications channel, campaign or software product at your organization, with added focus around bot UX design, multi-department team collaboration and app testing. The early initiatives will be learning experiences that will allow your company to continue to mature your conversational capabilities and refine use cases and connect with your users the way they want to interact with you.
Setting up a panel to test the bot is key so that a wide range of people can give it a test drive and see where the conversation takes them. All data points are useful in these early phases of bot design and development.
We walked through the capabilities of new conversational platforms, reasons why users reside on those platforms today, design ideas, use cases and considerations. Bot design and development will continue to evolve, with innovative platforms, channels and tools that will help companies get their message out. We believe these new channels will grow to be the key way users interact with software and enjoy the services they want. We invite you to connect with us to learn how you can design and implement your messaging app strategy.