If you ask anyone who has lived through the 60’s about their vision for 2021, you’d probably get an answer that resembles life with the Jetsons. While we haven’t quite mastered the creation of our very own Rosey the Robot yet, we have the next best alternative — chatbots.
So what are chatbots?
Chatbots are tools created to better and rapidly connect with a consumer’s needs conversationally. Created as a lower-cost alternative to live support teams, chatbots also have the added advantage of immediate response times, cutting the wait time usually expected by a live person.
Before we had Alexa and Siri, there was Eliza — the first ever version of a chatbot created in 1966 in an MIT Artificial Intelligence Lab by Joseph Weizenbaum. Eliza was able to react to specific keywords based on a series of defined rules and respond as needed. Decades later, chatbot functionalities have heavily evolved, but the general gist remains the same — each of these programs reacts to keywords that trigger pre-selected responses or actions based on predefined rules or artificial intelligence/machine learning analysis.
Chatbots can be housed on company websites or directly on messaging platforms, allowing marketers flexibility when it comes to planning for different strategies. Plenty of organizations take advantage of both options and include chatbots on messaging apps (Facebook Messenger, for example, when someone lands on a company page), as well as their websites for users who land on their homepage.
Chatbots aren’t a one-and-done deal; there are multiple uses for these programs that can help achieve a myriad of different goals. Below, we highlight some of the different types of chatbots and when to use them.
Types of bots
There are so many different types of chatbots currently available that we could write a whole book on the subject. However, for the sake of this article, we’ll cover the two most relevant versions for marketers.
Informational chatbots are bots specifically designed to relay any specific data information to the end user. These programs pull information directly from a company database or outside data source to help answer a consumer’s needs.
An example of an interaction with informational chatbots includes users looking to see what available books a library owns. That chatbot can then pull from the library’s database and showcase what’s available and what’s currently being rented out. Another example could include going on a health website and asking the chatbot about updates on COVID vaccine percentages in your area.
Utility chatbots act to serve consumers with specific actions, including initiating a payment, generating custom reports, placing an order, or booking an appointment. Utility bots will generally connect with other software in order to make their existing actions available to the user.
An example of an interaction with a utility chatbot can include going on a food delivery website and making an order directly using the chatbot, or going on a travel website and entering your desired vacation location and dates to receive a customized vacation plan.
Chatbots can also be broken down into technology types based on the bot’s capabilities. Flow-based chatbots, for example, only utilize pre-defined conversational flows that are already mapped out by a marketer. This prevents a user from being able to ask anything outside of the bot’s current capabilities and allows them to follow a direct and clear to navigate choose your adventure series that includes multiple user paths.
AI chatbots, on the other hand, allow a user to input any keywords they desire, with the bot using machine learning and artificial intelligence to respond in the most appropriate way possible.
Conversational marketing and the role of chatbots
Now that we’ve defined a few different types of chatbots, let’s dive into why chatbots could be a worthwhile investment for your marketing efforts.
Meeting customers where they live
- There are over 2 billion active users on WhatsApp and over a billion users on Facebook Messenger. Executing chatbots on messaging platforms allows you to engage with your audience in their active social habitat (this audience will be more likely to open a messenger message instead of being taken to another platform) while giving you a large pool of individuals to tap into.
- A well-planned chatbot can serve as an engagement driver, and generate new leads for your business in a matter of seconds.
Personalization and authenticity
- With chatbots, you can build individualized conversations at scale, directly addressing a consumer’s needs while adding an additional layer of personalization to the conversation. This creates authentic relationships with your audience that leads to action and acts as the direct line between the consumer problem and the solution.
- Live chat supports are great to add a personal contact with your consumer using a live person instead of machine learning technology; however, live support can present its own set of issues. Having a full live chat support team can be very costly to maintain and will never guarantee automatic responses to a consumer’s request the same way a chatbot would, which would ultimately drive more people to become disinterested and leave your product page.
Do’s and don’ts
If chatbots now seem like a worthwhile investment for your product, it’s important to make sure you get it right and avoid some of the most common mistakes associated with this technology. Below are a few tips on best practices and things to avoid when initiating a chatbot.
- Make it evident that your chatbot is not a real person. Your consumers will always pick up the difference between a live support member and a bot, so it’s always wise to be upfront about it.
- Create a conversation flow chart in order to set the chatbot’s purpose, tone, and reactions to different prompts from your customers. Make sure to plan ahead to properly draft your bot’s script to account for different user scenarios and always account for language to cover any errors or lack of understanding if a user asks something that might stray from the chatbot’s main purpose. Focus on the chatbot’s main goals, what functionality it needs to achieve, and how you can realistically integrate that task into the existing chatbot feature (for example making an appointment, checking flight status, lead generation).
- Once your conversation flow is created and the chatbot is ready to be utilized, conduct a thorough test with several people to make sure the chatbot functions the way it’s supposed to and doesn’t require any further tinkering or tweaks.
- Be useful, quickly. While we want our chatbots to be personal and conversational, we also want them to get to the user’s intended action quickly. Try to avoid a long conversation flow chart that takes multiple steps to get to the audience’s intended action for better results.
- Give the user exactly what they want, nothing more. It’s easy to get distracted by the many functionalities a chatbot can perform, but doing too much can distract from your main goals and lead to less action. Limit the different actions a bot can make and help the audience stay focused (chatbots that give users the option of what to respond to instead of having them type something out is a good example of this).
- Make sure you’re really thinking about what the right action you want your bot to perform is; provide a service that your audience actually wants to use.
- Start with one leading question and continue with follow-ups to gather more detail; that way your user knows exactly what the main purpose of your bot is.
- Personalization is key. Certain AI technologies can gather data about your audience and personalize the messaging to match their needs. This is just as important as dynamic creative used in paid advertising or email marketing.
- Keep improving and refining your bot. The more data you get from your users using your bot, the more equipped you’ll be with the necessary tweaks to improve your product.
- Always be on the lookout for new and improved technologies that can help better improve your chatbot’s technology.
- Don’t try to do it all. Your chatbot isn’t intended to be a virtual assistant like Siri and Alexa. Decide on your bot’s main goal and assign activities to it that will achieve those goals. Don’t get too caught up in other shiny features that your bot can include just because they’re possible.
- Avoid sending mass messages. No one wants to receive an unprompted chat from a bot, especially on platforms that are used for more personal communication like FB Messenger.
- Don’t overwhelm a user with a series of questions that don’t bring them closer to achieving the main goal. Always make sure you’re only asking for the necessary information to shorten the full user journey as much as possible.
- Chatbots shouldn’t be included on pages that are focused on getting the user to read more about the organization or download a specific resource. This will distract from the page’s goal.
- Never assume that every person wants to chat, or that each user is the same. Try to curate your chatbot to tend to each user’s unique persona and where they currently are in your funnel.
- Try not to make your audience ever leave your current messaging platform. Embed all of the available functionalities needed into your chatbot to avoid any drop-offs.
WeChat, China’s most successful messaging platform, is one of the first early adopters of chatbots. WeChat introduced the chatbot functionality to its application back in 2013, and it has since exploded with thousands of different bots. Brands flock to WeChat to deliver automated services for consumers on the platform, which already houses over a billion active users. These services include using bots to automatically pay your bills without leaving the app, placing a food delivery order, or even merging a chatbot with AR technology to try on brand new makeup products.
H&M was also able to take early advantage of chatbot functionality when they launched their very own bot on the Kik messaging app platform in 2016. What differentiated H&M’s chatbot from other traditional e-commerce brands is the functionality of the bot acting as your personal assistant, instead of just selling you a product. The bot would ask the consumer a series of questions regarding their style, and start to deliver different outfit ideas that could be refreshed and edited until the user would find the perfect match. Once a user would find their dream outfit, they could immediately add the entire look to their cart right away.
Chatbots don’t always have to be external tools, and this was evident with Adobe’s internal employee bot that was created during the pandemic. In order to find a way to tend to thousands of employees’ support tickets and issues, Adobe created a chatbot directly integrated with Slack to minimize the weight of email support requests for its support teams, and to be able to better route employees directly to their right division. The bot was also able to route people directly to FAQ pages in situations where an existing solution was already present online, instead of needing to file a ticket with a live representative.
Speaking of the Covid-19 pandemic, the World Health Organization launched their very own chatbot on WhatsApp and Messenger called the WHO Health Alert, which informed people about breaking Covid news and resources. The bot included answers to several frequently asked questions regarding the pandemic and offered any new travel and health breaking updates.
Domino’s Pizza took a stab at their very own chatbot called “Dom,” which was able to directly place an order, review past orders, and give status updates on any active purchases. Domino’s took advantage of multiple active messenger devices, including their chatbots on apps like Messenger, Slack, and Apple Watch.
Chatbots are an incredibly worthwhile investment if the functionality makes sense with your product, and when optimized well. There’s a big misconception that bots are solely used to drive sales, and you’ll see from the above examples that this is not necessarily the case. Chatbots can help solve countless goals for whatever industry you’re in.
Ready to invest in a chatbot for your business? Look into existing chatbot builders to see what makes the most sense with your goals, if your bot requires slightly more complex features, look into building one yourself. Remember: testing is everything, even if you’re seeing great results in your initial chatbot set-up, never forget to scale and make adjustments to ensure you’re always optimizing.