In just a few years, the messaging market has changed dramatically. We have progressed from a dominant SMS channel to a very large number of messaging apps that offer complex conversations managed by chatbots, which can interact naturally with people. Today, understanding interactive conversational marketing is a priority.
As a result, communication methods for both individuals and companies have changed significantly. By offering greater functionality and simple interfaces, messaging applications have succeeded in capturing a large part of the mobile messaging market between individuals.
With a 394% increase in the time spent on messaging apps between 2016 and 2017, apps have brought the messaging market back into the spotlight. Here are the three services that stand out:
- RCS or Rich Communication Services. Presented as the evolution of the SMS channel and driven jointly by Google and telecom operators, RCS has a very good chance of being THE messaging service of the future
- Facebook Messenger, which opened up its API in 2016 and is now a pioneering, stable and efficient channel
- WhatsApp, whose API is gradually opening up to brands, will enable the first interconnections to be launched in 2019
However, it is not a “stable” market and many players coexist in this area:
- Business Chat, Apple’s answer to these messaging apps, but currently complex to deploy
- WeChat, a very large player but mainly for the Chinese market
With this rapid growth, a new way of managing customer relations has emerged: chatbots and conversational interfaces. But if the emergence of these conversational agents has been so rapid, it is also thanks to a more than favourable economic situation.
New mobile usages
Over the years, the smartphone has established itself as the leading display screen (ahead of the computer and the tablet). This profound change began in 2007 when smartphones saw their autonomy increase and their screens change. Larger and more tactile, their uses are evolving. The number of sessions per day is exploding, about 150, but attention spans are decreasing. With an average of 70 seconds per session, it is now essential to capture the attention of users from the very first moment.
And in this context, one format stands out from all others: the chatbot.
In existence almost since the beginning of the IT era, they have distinguished themselves and become a key channel for marketing with messaging applications. Fast and available 24 hours a day, they allow you to create engaging and personalised conversations. An average conversation lasts about 120 seconds, during which an average of 20 messages are exchanged. No other format can boast such a degree of engagement.
Users also say they are socially ready to converse with a chatbot if it improves the quality of service. The figures show this with a completion rate of 86% for started conversations. And technologies are developing… It is now possible to connect these conversational tools to your business information (IS) system to automate recurring and easily processable requests.
Comprehension of human language is improving
A machine can talk to another machine, that is a matter of record. But with a human being, it is immediately much more complicated. There are an infinite number of standards that make the humanisation of conversational agents very complex: Tonality, language used, spelling mistakes, familiarity/formality of address, the context of the last discussion, digressions and so on.
However, in recent years, artificial intelligence has progressed considerably. NLU, NLP, ASR, irrespective of their names, these technologies aim to reduce this gap. But today, these technologies, with their immense potential, are still in their infancy. They represent a genuine differentiating factor in the face of competition as well as also exorbitant investments to deliver a satisfying experience.
The right moment, the right technologies, everything seems to be in place to revolutionise customer relations.
But you still have to look at the solution from the right end!
Don’t try to move too fast too soon
When you launch a chatbot, you are starting a project that can last for years. First, because the scope is unclear, and it is difficult to assign it definitively to a single corporate department. The chatbot is the third distribution channel after points of sale and websites, and, like their predecessors, they affect all the company’s services, from CRM to e-commerce and, of course, essential business information services (IS).
Then because the uses cases are numerous and that seems logical given the scope they have. Chatbots can assist after-sales service as well as help to highlight a high point, so it is difficult for business managers to know where to start.
Obviously, all this has a cost, sometimes phenomenal, and with an uncertain expectation of return on investment. Because if the theory suggests that chatbots will be the new Eldorado, the reality on the ground can sometimes be quite different. Like any new tool, the conversational interaction still needs to prove its worth. For this reason, it is essential to select the right Use Case.
It will also be a question of finding a Use Case which can be deployed quickly, the ideal first step being to avoid having to enlist input from IT, thus controlling time spent on the project and therefore the budget as much as possible. It will then be a matter of choosing relevant KPIs to promote this channel internally and justify future investments.
The autonomous chatbot is, therefore, the perfect solution, it is based only on decision trees, much less complex and expensive than semantic recognition technologies (NLP/NLU) and can operate without any connection to IS systems. Its objective is to promote a brand highlight during a given period, during sales, for example. Budgets and project times are reduced and better controlled, and it becomes easier to prove the channel’s effectiveness.
But for an autonomous chatbot to be a success, it is still essential to ensure proper media coverage.
Ensuring strong media coverage generates traffic
If you have followed the previous recommendations, you may now be wondering how to bring traffic back to your chatbot. Here we will mainly address the case of Messenger; RCS and WhatsApp only offer limited features in terms of media coverage for the moment.
With Messenger, the possibilities are numerous: first of all, it is possible to relay your chatbot to your customer base. SMS gets the best results with an average of 10% conversion, where dedicated email is at 1.5%.
You can also add two plug-ins onto your website; the first one will allow the user to interact with your chatbot through a pop-up placed on your site. The second is designed to send a message to the user after an action on the site (an order for example).
Finally, the Facebook ecosystem remains an excellent tool for generating traffic in addition to the organic posts on your Facebook page and native ads, inserted in the news feed; you can also use Instagram, its stories and sponsored posts to ensure the success of your chatbot.
The success of your chatbot does not, therefore, depend only on the Use Case set up, its media coverage is just as essential. Without that, the best chatbot in the world will have little chance of proving its worth.
At this point, you have everything you need to start your first chatbot, and it will be interesting to conduct several iterations to learn more about the behaviour of your users on these new messaging services. Future steps will involve the progressive integration of business information systems into your chatbot (CRM, product flow, payment, reservation management, etc.) to bring ever higher quality to the service delivered. NLP/NLU technologies, which while complex and expensive today, remain technologies of the future, and not forgetting the human being who, when the context requires it, adds inestimable value in customer relations.
Article originally published on the myElefant blog site – July 2019