An overview of conversational marketing chatbots

What is the definition of conversational marketing?

Conversational marketing combines human conversations, which may occur via voice channels, live chat, or messaging apps, with interactions orchestrated by AI, which may take the form of chatbot interactions or triggered messages that provide customers with the information they seek, immediately and in the appropriate context.

Indeed, one of the major characteristics of conversational marketing is contextually appropriate communication. With the purpose of minimising friction from the customer journey, it allows brands to always give the correct information, at the right moment, while taking into consideration all prior interactions and data obtained to eliminate repetition and make the consumer feel understood and valued.

Conversational marketing has emerged in recent years as one of the most effective methods of creating customer relationships and allowing consumers to explore your products and services at their own pace, regardless of where they are in the customer journey.

It shouldn’t matter what channel the conversation takes place on if the right conversational technology is used – just like chatting with a friend on voice, email, messaging app, or social media, they will always remember your last interaction and you can pick up where you left off without missing a beat.

Conversational marketing is so popular because a company can do this with all of its customers, all of the time, across all channels.

Chatbots’ Role in Conversational Marketing

One of the primary drivers of the expansion of conversational marketing has been the development of easily available chatbot building technology. Chatbots are ideal for a wide range of applications that rely on the immediacy and 24-hour availability that chatbots give.

They can now be distributed on any digital channel, including the web, WhatsApp, Messenger, and others. Instead of seeking for a phone number, many clients opt for a chatbot connection so that they may get their question answered right away without having to browse an IVR menu or wait in queue.

People intuitively grasp the 80/20 rule and know that for the majority of their enquiries, they can get a speedy response from a self-service chatbot that has been programmed to offer answers to the company’s most commonly asked questions, whether they realise it or not. For the remaining 20%, they will usually be content to wait for a human agent. To be honest, with the quick progress of chatbots and tools such as ChatGPT, we may soon be referring to it as the 95/5 rule!

Meanwhile, let’s take a look at some of the ways you can start conversations with potential customers, as well as some of the use cases where chatbots excel.

What are the rules for starting chatbot conversations?

You should be aware that in most countries, there are restrictions on how chats on digital messaging channels can be launched.

While a web chatbot can gently inquire if it can assist someone who is exploring your website, it cannot initiate a new discussion with a person on messaging channels. Companies such as Meta and Viber have rigorous regulations in place, so it is a good idea to become acquainted with them before you begin developing your conversational marketing use cases. The regulations for WhatsApp, for example, are detailed here.

Developing access points for chatbot marketing

• If you have paid for adverts on the web and social platforms, include buttons that initiate a dialogue. Meta makes this incredibly easy by connecting Facebook and Instagram ads with WhatsApp. 
• If you have physical media such as billboards, point of sale displays, or advertisements on street furniture and public transportation, you can include QR codes that will initiate a conversation when consumers scan them with their smartphones.
• QR codes can also be included on product packaging. Someone may be looking at one of your products in-store and have a question about it, perhaps about similar styles or colours, which a chatbot might easily answer.
• Incorporate chatbot connections into your contact information, whether it’s online, in business directories, or on physical signs. People can message organisations with a Google Business profile directly from Google search results.
• All types of disruptive firms have employed guerilla marketing strategies to drive engagement, and chatbots can play a significant part. One of our customers, Unilever, executed a highly effective campaign to raise awareness about a new product line. They put up eye-catching and intriguing posters in city centre areas, along with a WhatsApp number that visitors could message to begin a dialogue to learn more.

Conversational marketing chatbot use cases

Chatbots are being used in a variety of creative and efficient ways by brands in their marketing. This is driven by two developments. First, chatbot technology is rapidly evolving, making them smarter and more versatile than ever before. Second, people are becoming more accustomed to interacting with chatbots and increasingly expect a nice experience.

Product and service research

When someone has a need but isn’t sure how to meet it, a chatbot can be a wonderful tool to help them investigate the topic and find the finest goods to help.

Assume I’m an aspiring YouTuber looking to enhance my sound recording equipment. When I enter into a technology store, I am confronted with a wall of microphones and audio devices, and I have no idea which one is appropriate for my setup and budget. In-store, I can ask a shop assistant for assistance, but making the same purchase online would be difficult unless there was a virtual assistant that could help me narrow down the best mic after a few basic inquiries.

Chatbots are ideal for this use case because they can analyse massive volumes of information in milliseconds to generate a shortlist, which they can then show in an engaging and dynamic manner, such as with an interactive image carousel.

Furthermore, store opening hours aren’t an issue for chatbots, and they may continue to drive purchases late at night and on weekends – when aspiring YouTubers are most active.

Assistance in getting started

Customers who sign up for a service online may take several paths depending on their own circumstances and information. Rather than presenting a massively extensive web form that covers every possible scenario, employ a conversational chatbot to assist users through the process, asking for tiny amounts of information at a time and answering queries when anything is unclear.

Chatbots, when incorporated into a customer interaction platform, can also be used to reduce drop off rates and re-engage with consumers who have abandoned the process. The ride-sharing firm Bolt implemented this for its driver sign-up process, raising conversion rates by 40%.

It’s one thing to attract clients to your service business, but if they don’t get up and running easily and start obtaining value from your service, they’ll go.

Chatbots are wonderful at assisting customers as they explore your platform and begin to get the feel of it. Nobody reads manuals anymore, do they? To avoid overwhelming new subscribers, a chatbot can drip feed the appropriate information at the appropriate moment in the process.

Engagement

This is the sweet spot for conversational chatbots; people enjoy interacting with them when they are well-designed and provide something new and amusing. The possibilities are almost infinite, and organisations ranging from government agencies to sports clubs are making use of them. In fact, Arsenal FC was among the first to use a chatbot to connect with fans. Robot Pires (named after Arsenal’s iconic midfielder Robert Pires) is available on Messenger, Skype, Telegram, and Slack and provides the latest club news, fixtures, and is an expert on player data and past outcomes. Do you need to resolve a pub argument on a major result from the 2015 season? The robot will be aware!

Other chatbot interaction use cases you might wish to consider:

• Competitions and quizzes can be used to promote new products.
• Use a chatbot to keep loyalty club members up to date on benefits and special offers.
• Keep website visitors on your site longer by directing them to pages that they might be interested in.
• Public relations stunts aren’t for every firm, but chatbots are a terrific way to pull them off, like Domino’s Pizza did with their Dom Juan chatbot on Tinder.

Upsell and cross-sell

This is a significant use case because components of it may be included into all chatbots that a company deploys, even those with transactional or informational purposes.

AI product recommendations, for example, can be included into chatbot conversation logic to help clients locate things that they might be interested in that could help them solve issues that they are reporting.

However, care must be made to ensure that the offers are appropriate and targeted to each customer and do not detract from the chatbot’s core goal. If the person is not interested, they should be able to decline the offer and opt out of receiving similar ones in the future.

How to Create a Conversational Marketing Chatbot

We described the distinction between rule-based chatbots and intent-based chatbots in a recent blog on the newest advancements in chatbot technology. In summary, a rule-based chatbot helps individuals find information by offering a cascading set of possibilities to choose from. Although this is not precisely a “conversation,” it is a useful tool for marketing use cases in which a person wants to learn about the products and services that you offer, or just find out where their local store is.

In contrast, intent-based or conversational chatbots are intended to simulate a two-way discussion with another human. Indeed, with certain well trained conversational chatbots, you may wonder if you are dealing with a human rather than AI.

Natural language processing (NLP) is used by conversational chatbots to analyse textual inputs from people interacting with them in order to determine what their objective is and offer appropriate and factually correct answers.

Conversational chatbots are more difficult and time intensive to develop since every possible branch of the conversation flow must be programmed to offer acceptable responses. This is known as conversation design.

What exactly are intentions and how do they work?

A conversational chatbot’s capacity to effectively recognise human intent is critical to its effectiveness. This in turn influences the ‘next steps’ that the chatbot takes, such as initiating a specific conversation flow or performing a desired task.

An example can readily demonstrate the concept of intent.

Assume a person wants to know how many miles they have in their frequent flyer account.

• They initiate a conversation with the airline’s chatbot and, after verification, ask, “Please tell me what my free mileage balance is.”
• The chatbot reads the text and begins by removing words it recognises as irrelevant before determining that the primary intent is ‘mileage account balance,’ which it then triggers. If the intent is unclear, it can prompt the user to clarify by presenting them with options to choose from.
• The account balance workflow connects with the airline’s loyalty system to retrieve the current balance and then formulates it into an appropriate response, taking into account any regional or customer specific variables, such as number format – “You have earned a total of 30,000 free miles.”

This is a fairly straightforward instance, but the technique is the same for any use case. The only variation is the total number of sentences that must be trained into the chatbot for each intent. It is fairly uncommon for a single intent to include over 400 confusing statements.

5 steps to creating a chatbot for conversational marketing

1. Determine the chatbot’s purpose.

In other words, for what purpose is the chatbot being developed? Be explicit about what the chatbot’s capabilities should be, and equally important, state what the chatbot will not accomplish. This prevents scope creep, which slows down projects and causes them to deviate from their original objective.

2. Develop a chatbot persona.

Conversational marketing chatbots are all about generating interaction and attention, but that doesn’t mean you can abandon your brand identity and tone of voice completely. You want people to trust your chatbot, therefore it should be aligned with your brand’s values and your customers’ opinion of you. This will imply:

• The language employed
• Interaction tone – whether the dialogue is formal or informal
• Whether you can incorporate humour or a sense of playfulness

You can go to Google’s great advice on creating a persona. We also believe it is a good idea to give your chatbot a memorable name and an avatar that expresses its personality.

3. Create a high-level dialogue flowchart.

A conversational marketing chatbot script should resemble a family tree, with numerous generations and many branches, rather than a straight history. This allows the discourse to be more spontaneous and less structured. The individual interacting with the chatbot should have many options at each branch in the flow and can thus choose whatever route they like to accomplish the desired conclusion.

Always return to the initial aim to ensure that you don’t waste time over-engineering low-value side quests that players may try and investigate.

4. Write the script for your chatbot.

Now is the time to create all of the words and responses that the chatbot will use to reply to the intent of the individuals interacting with it. If you’ve done a good job of describing the chatbot’s persona and mapping out the conversation flow, this should be much easier.

Here are a few pointers:

• Always remember to write in a natural and conversational tone, and avoid using terminology that you would not use in a chat with a friend. For example, you would never say, “Before cancelling our agreed-upon plans, please refer back to the terms and conditions of our friendship.”
• Ensure that each branch of the conversation flow finishes naturally and does not offer a dead end that leaves the person wondering what to do next. You won’t get it right the first time, but comprehensive testing should clean out any unhandled instances.
• Keep it brief and to the point. Long, difficult-to-read blocks of text should be avoided.
• If an intent is unclear or ambiguous, request additional information before proceeding to the next stage.
• If the chatbot was unable to assist the customer in achieving their goal, provide them alternatives. Provide the choice to divert to a human agent, link them to another resource that can assist them, or collect their information so that they can be contacted later. The final sign off should always represent what was accomplished during the discussion.

5. Test, test, and test again

The testing step is not a checkbox exercise in which you can cut corners. You must test every conceivable branch of your conversation flow to ensure that it makes sense and adds to the chatbot’s purpose.

There are specialised tools that can assist you automate portions of your testing when you have a big number of intents and sophisticated logic. However, nothing beats using real people, preferably from diverse backgrounds and ages. These could be trained testers, a sample of your target population, or even members of your team who were not involved in the chatbot’s design and development.

Thorough testing will discover flows that may need to be modified as well as extra phrases and terms that must be covered.