The Multichat Blog

Guest post: Chatbot Evolution

by Multichat Team AI, AI Chatbots, Chatbots Machine learning, Machine learning Technology,

Tailored-messaging based on demographics, user data, and AI integration.

Written by Doron Schwartzblat
Head of Platforms at Straker Translations

Chatbots are becoming more widely used across various industries, and leading companies have already started implementing Chatbots as a part of their customer-service workflow*.

*Chatbots will power 85% of all customer service interactions by the year 2020.

The increasing usage rate of Chatbots does not only present businesses with new challenges, but also forces the industry to come up with creative solutions, which will in turn, help reshape this field and determine new trends. This made a very interesting discussion between and Straker Translations.

Clear communications have always been a barrier when dealing with cross-border commerce and/or customer support, and over the past 11 years, Straker helped thousands of clients to translate the content of different subject matters, and for various target audiences. The key factor in delivering a high-quality translation is to make it sound like it was actually written in that language – that is only achieved when applying the adequate tone of voice and adjusting the style to the target audience.

  • From a language perspective, marketing managers agree that tailoring the messaging to a respective target audience would deliver higher conversations overall. For example, German shoppers between the age ranges of 40-60 would almost always prefer a formal tone and style, whereas younger age groups react better to informal speech and colloquial verbiage.
  • From a business perspective, has shown a significant increase in lead-conversion statistics and lower cart-abandonment rates whenever a Chatbot is properly used and customized based on industry-specific matrix. delivers a great experience around lead qualification, customer support, chat broadcasting, registration, surveys and answering FAQs.

Having Chatbots supporting the most common chatting Apps such as Messenger and Viber, delivers a fantastic engagement result from users, as they are not required to install any additional apps. The response time is instant, the solution is available 24/7 so users are never left without a reply. The audience can be segmented and the messages can be targeted accordingly, in other words, the system can be fully tailored and personalized to each user’s preference.

Taking the above aspects into mind, we can clearly see there is room and demand for a more sophisticated user-interaction, and this is exactly where language localization would play a significant role.


A shared effort


Combining the expertise and technology of both companies, and Straker have started to work on a bespoke, highly-customized Chatbot; it will have pre-determined responses which would vary according to the user’s demographics. To provide a usage example from the Travel industry; the user will have to input their order number or email as a part of the support process, and once verified and linked to an account with demographics data, the relevant script will be allocated to that chat window. In other words, the Chatbot will crosscheck the user’s details and will assign a tailored, and pre-determined answer script to that chat window. This should create a pleasant chat experience for users who engage with the bot, and it is expected to deliver a much higher conversion rate and customer satisfaction from the service offered.

Fusing this custom Chatbot functionality with AI, translation memory, passive data collection and machine learning, will enable companies to save costs on support personnel, and deliver outstanding results within a shorter timeframe.

The AI would work in the background to collect & analyze chat data, and implement better answer prediction for various industries, all based on translation samples Straker would accumulate in the system’s memory. Statistics from on user engagement and satisfaction results will also be fed into the Straker’s Ai powered engine, to analyze data around the efficiency of the different translated scripts.