Ever since the Facebook has opened up its Messenger platform (FBM) to ‘bot’ developers at Facebook developer conference (F8) in April, bot designers have been busy in swiftly developing chatbots. These bots are smart script with the layer of Artificial Intelligence (AI), Natural Language Processing (NLP) and Machine Learning (ML).
Guess, What’s the mainstream bot which is AI-powered out in the market? We all know about it. It’s……Google. It’s a bot because it is based on NLP of the search queries of the users. It does its job pretty well. It provides most accurate answers based on the user’s search queries and it does it even without the need for the user to additionally click on the results.
Similar to Google, if we closely observe, we are using largely the invisible bots on day to day basis. These are nothing but the various algorithms powering the most widely used apps people are using daily. Whether it’s a conversation or not, you might be interacting with a bot whenever any structured output is given to you based on your deliberate input(s)/or behavior. For example a question answer type of format is one of the ways a conversation is initiated. It’s nothing but a bot which is trying to mimic a human like conversation.
Over the time, bots have become smarter. This has happened as they are interacting with us, the humans. They are learning from us, gradually understanding our needs and deliver the results in an automated way which is relevant to the query of the customer/user.
However, still the messaging/chatting alone is not the best way to interact with the bots. For instance, Facebook Messenger has close to one billion plus users. Every single company would like to introduce bots for the messenger app to engage with its customers. We are still exploring other possibilities of interacting with the bots beyond chatting/messaging or SMS.
There are two two classes of bots getting into mainstream. One which is used for messaging/chatting which are scripted and other which is AI and NLP powered bots. Most of the brands are using the scripted bots on major platforms like Facebook Messenger and Kik. When it comes to AI-powered bot, it poses number of risks when It comes to tricking the system, due to the sophistication involved in its design and development.
Users are not interested in what is happening at the backstage. They are only interested in the play which is happening in front of their own eyes. The inputs and outputs are related. The information which is given as input impacts the information which is coming out. It can be sometimes an unexpected outcome. The brands find the scripted bots most feasible and easy to use strategy to test the new technological era of the bot world.
Brand marketing will be managed by Chatbots/Scripted bots to a large extent in the future. They provide an extension to the brand’s existing marketing strategy by making it easy, natural and fun to interact for the users who are engaging with the brands and the best part is that it doesn’t appear as an advertisement or a promotion or just any other marketing gimmick.
The scripted bots are slowly becoming the integral part of brands’ mobile engagement strategies. But the chatbot strategy should not be secluded one. It should carry on along with the brands’ other marketing and promotion strategies well. The brands should know how to layer the chatbot experience along with other engagement experiences. For example, Distributors want to promote a new film. One of the most effective approach could be the chatbot. It takes the customer engagement to a whole different level. Promotional campaigns like trailers and GIFs can be delivered over mobile when customer ask for it; which could possibly provide the personal experience to customers like location of nearby theater, show time and ticket availability, ultimately influencing bot users to purchasing the ticket.
The bots alone cannot sustain interest and engagement; they must be complimented with other marketing engagement campaigns that drive awareness and interest. Remember to create the marketing channel of driving traffic to the bot before before investing in chatbot project. Once the two strategies are in place, bots will create a full circle branding strategy to keep customers engaged and entice them to promote and share the bot with their friends.
The performance of the campaign matters. The brands which are implementing the scripted chatbot strategy, must have access to analytics to measure how well or bad the bot experience is. They should make sure they have a constant update on things like the read rate of the bot messages, the timestamp on the average back and forth conversation that’s happening with each user in the bot, how many people went through the funnel to reach the KPIs (clinking a link or making a purchase), and an understanding of the sentiment (was it a positive or negative experience overall).
Lets discuss the strategy followed by a retail brand American Eagle Outfitters (AEO) as how the brand implemented the AI powered chatbot system into its current operations. During the the holiday season last year, the retailer tested new and effective ways of communicating with its customers via chatbots. AEO has successfully introduced its chatbot – the AEO Holiday Gift Guide Bot on Black Friday. This chatbot was released on both Kik and Facebook Messenger. It prompts user s to take a short quiz and provides them in return gifting ideas based on the answers given by them.
It is possible for the customers to send any message to the chatbot technically however, AEO monitor the conversations. This is done in order to review the performance of the bot that whether the bot is answering to the queries of the customers appropriately as well as to monitor the customers behavior with chatbots. Creating a generic bot is in its early days, and the engagement span with customer is shot in digital age. If bot is not covering the desired queries, customer engagement will be a challenge. Hence, AOE has decided to create a scripted bot to guide the customer as well as the bot on messaging journey.
For instance, a message sent by the user,” I’d like to shop for bras” would be directed towards the bra shop. On the other hand, if the customer sends a random message will receive a message from the bot as “I didn’t understand that, but tap ‘Home’ to see what we can do.”
In order to measure the success of the bots, the retailer is benchmarking their performance against social metrics. “The premise is that people are on these messenger apps more than they are on social networks,” Bell explains. It is remarkable to know that since the introduction of the chatbots, the brand has acquired more customers than what it did via all social channels combined within two weeks of launch of chatbots.
It has become increasingly popular nowadays for brands to explore and implement the bot strategy into the mobile engagement world. This is going to run through into the foreseeable future. And there is no doubt about effectiveness of bots.
Are you are a brand which is also exploring a chatbot based strategy to compliment your existing marketing/promotional strategy to increase user engagement, Rasbor is creating customer chatbots for enterprise, which can be served on Facebook Messenger, Telegram or can be integrated with Android and iOS applications.
Do share your experiences on working with Chatbots. We are building a platform where you can create a custom chatbot and publish it on Facebook, Telegram, Andriod and iOS. Kindly share this article and follow me on Twitter.