—TOMER SHARON, WEWORK
LET’S EXPLORE SOME OF the use cases where bots can be utilized. This is not an exhaustive list by any means, but in order to start thinking about design aspects of your bot you will need to get a taste of what is out there.
From buying on Amazon to ordering a ride, conversational bots can facilitate commerce in our lives. When done right, conversational commerce can be more intuitive and engaging than traditional commerce. No more shopping lists on the fridge—you just say “Alexa, add sugar to the shopping list.” Travel bots can replace travel apps and websites, providing everything from booking to alerts of flight times, and customer service as well.
Users no longer need to install an app to get a ride—they can just ask their favorite @Uber or @Lyft bots for a ride in their chat app. This means that, for the first time since the mobile revolution, there is a clear separation between intent and installing an app. Assuming discovery is done right, this will also mean that the user acquisition cost for commerce will go down, possibly providing a more cost-effective means to reach your users.
A very interesting bot in this space is Kip (Figure 4-1), a shopping bot for teams. From office supplies to snacks, Kip handles the complex coordination of getting everyone in the team to add to the group order.
The interesting thing in this example is that the bot is introducing a new ecommerce concept called the “team cart,” in which multiple members in the chat can add to the cart and the admin can pay for it. In this way Kip is enabling a way of shopping that was not available until now.
This is where chat platforms for work, such as Slack, focus their efforts. Here we can find bots for HR, legal, sales, marketing, facilities, product, engineering, and other departments. GitHub has coined a term for its way of managing DevOps through chat: chatOps. Most startups connect their customer relationship management (CRM) systems to get notifications for new clients. Entire business operations can become more productive with personal assistants that help us do our work better.
Figure 4-2 shows an interesting legal bot use case.
There is a strong incentive to use bots for business workflows. Most corporate workflows are cumbersome, require logging into legacy systems, and are often time-consuming. Using bots to facilitate short, contextual, and actionable tasks can greatly improve the productivity of a team.
Personal and professional productivity is a growing market. Here we can find bots focused on reminders and to-do lists, personal and team task management and completion. While these seem like simple use cases, they are very popular in the mobile app world and appear to have high engagement and install rates in the bot stores.
I also see many use cases for personal bot coaches that help users with weight loss, finances, parenting, sports, and more. It seems like the nature of the medium—having the bot talk to you in your chat app—makes the interaction more effective and engaging.
A good example of a coach bot is Lark (Figure 4-3).
One of the incentives for using bots in these types of use cases is user compliance—bots can provide a more personal experience that is harder to ignore, compared to mobile apps for example, and users are often more willing to provide information to a bot than to fill forms in an app.
Bots for this set of use cases are starting to replace email or in-app notifications. These could be news bots, price watch bots, analytics report bots, or bots that notify you when your kids get home. There are a couple of differences between bot notifications and traditional mobile notifications:
Notifications sent to a group/team/channel chat are more collaborative, and we see teams collaborating and taking action faster and more productively than when emails are sent to a group.
While traditional notifications take you back to an app or a website to take action, many chat platforms provide you with a set of controls, such as buttons, that you can use to take action inline.
Given the right use cases, notification can very quickly turn into taking productive action (Figure 4-4).
These micro workflows can happen in consumer use cases, such as a discount alert with an action button to buy, as well as business use cases, such as for actionable reports or approval processes.
The incentive here is that using bots for reports and alerts improves the actionability, transparency, and context. One alert in the #DevOps channel is worth a thousand emails.
This is an interesting set of use cases where the service is actually provided by humans, but the bot acts as a router/operator and connects the user with the human service provider. In the same way that Lyft and Uber connect you to a human driver (at least, at the time this book is being written), a bot can connect you to another human who then facilitates anything from IT support to songwriting.
A good example of an operator bot is Sensay. The Sensay bot (Figure 4-5) lets you instantly connect with a real human whenever you need advice or inspiration. It works across platforms and across devices.
While some of the services provided by the humans Sensay connects users to would be hard to replicate with a bot, the actual act of connecting people is mundane and can easily be executed by a bot.
The incentive here is to provide a more friendly and useful version of the common interactive voice response (IVR) systems we all love when calling our service providers. The hope is that text-based bots that are delightful, personalized, and actually get us to the right person can change our negative perception of most common answering machine–like IVR systems.
This is one of the most common use cases for bots. Here, the bot serves as the first line of support, for internal employees or external customers. For internal use cases, the bot can answer questions like “What is our vacation policy?” An external consumer brand bot can answer questions like “What are your business opening hours?” Support and FAQs are an easy use case because they usually follow a pattern of a single request/response, and the questions are usually repeated and easily trainable.
There is a strong incentive to use bots in customer support use cases—this is because bots are typically much more cost-effective (and in many cases faster) than humans at performing simple repetitive tasks. These business use cases are very popular for bots on Facebook Messenger and Slack. According to initial experiments, bots can easily cover approximately 40% of internal and external support tickets.
At the time of writing of this book, the most requested bot by Slack users is one that integrates Salesforce CRM with Slack. Business users keep telling us that they want to be able to run account lookups from within Slack while talking about clients.
CRM is by no means the only system integration requested, though. From Google Analytics to Merkato, WorkDay to Concur, users crave simple integrations that will save them time and make them more productive.
Statsbot (Figure 4-6) is a great example of an integration bot. It pulls information from Google Analytics, Mixpanel, and other marketing systems and integrates the insights from these systems into Slack.
The core incentive here is that users do not want to context-switch between apps to get the information they need or run their workflows. They want to converse with the tools and services they use for work inside their chat apps.
These use cases are exploring ways to entertain and delight the user. Entertainment is a big part of our lives, and bots can be a part of that—from full-featured games on Kik to Alexa telling my kids fun facts and knock-knock jokes.
These bots are taking a different path from our traditional concepts of entertainment: neither very passive, like a TV, nor very rich and engaging, like a game console. The bots are trying to turn a conversation into a fun activity—go figure! Who would have thought conversation could be fun?
Fun fact: Early experiments done with kids’ movies showed that users were able to have long conversations with bots about their favorite characters and movies—sometimes even longer than the movies themselves.
An example of a social entertainment bot is the Swelly bot (Figure 4-7). Swelly lets you pick between two options and shares the voting results of all users. It is a delightful experience to casually vote on foods, fashion, vacation spots, and more.
Swelly also has a (non-bot) mobile app for both Android and iOS, but the team reports strong engagement on the bot user interface. Bots can reengage with users in an easy way and drive them back to the conversation or game. One of the core incentives here is that bots can reengage with the users and encourage them back to the service in a less intrusive and more customizable and friendly way than app notifications, for example.
In this set of use cases, bots try to use the chat medium to create brand awareness and engagement. As bots become more popular and gaining traction with apps becomes more and more expensive and difficult, marketing managers are seeking ways to build bots for their brands.
There are some interesting use cases around notifications of new products or discounts by top brands, and a lot of other experiments. Bot builders are still trying to figure out what a valuable and engaging brand bot looks like over this new conversation interface.
Remember that bots are only as good as the services they expose, and bots for brands are no different. The Whole Foods Market bot in Figure 4-8 is a good example of a bot that not only provides access to the brand, but also adds value for the user.
Remember that bots are only as good as the services they expose, and bots for brands are no different.
The core incentive here is app fatigue—users are tired of installing specific brand apps. Bots provide brands a new and fresh way to engage with their users in a useful way.
As the market matures, we will see more and more use cases emerge. I talked to one of the CEOs of a leading mobile platform, and he said that no one could have guessed that asking strangers to come pick you up from your home, with their private cars, would become one of the most profitable use cases of the mobile world. We still do not know which bot will make it big, but history has taught us that a few will change our world.
In the next chapters, we will dive into the anatomy of the bot, exploring different elements that compose a bot and how they manifest themselves in different platforms. We will take concrete examples and learn from their builders what worked and what didn’t.