The recent surge in Ai-powered chatbot adoption by education institutions and edutech schools has provided students and prospects with around-the-clock channels to engage with their institutions. Some bot are more successful than others in tapping the best of NLP to achieve their stated purpose of providing near human support. At askadmisions.ai we have benefited from the learnings of implementing 30 + automated channels mostly on WhatsApp.
Here we are listing down 5 objective metrics that may help a decision-maker evaluate the success of a chatbot
A good admission chatbot establishes its scope and attempts to determine prospects’ context at the very beginning of the conversation. With interaction buttons now available in session responses on WhatsApp, the bot may prompt a user to choose between a set of predetermined flows before proceeding to actual Ai-powered free-flowing conversations. Defining the flow narrowly may help achieve a high percentage of positive resolutions, but will only serve a small percentage of incoming queries. A well-designed flow tree should target addressing 70 - 80% of user queries. This is challenging in scenarios like edutech operators who mount bots not just on website but also on other entry points like FB, offline promo - using QR code and google results
Sorting context should also account for Authentication in case of existing student queries as well as dealing with the increasing problem of spam messages
While super bots like Alexa and Siri are designed to understand and react to almost anything under the sun, domain-specific chatbots like admission bots either on the web or WhatsApp must be designed to resolve and disambiguate intent and its expression in various common utterances. In an education chatbot queries about procedures, course names expressed in acronyms and affiliations would need some nuanced handling
Interaction depth is a function of relevance and rich content. For a chabot to go beyond the first relevant response and engage users for a few minutes deeper integration with the content management system.
Chatbots only serve content that is available for them. On Whatsapp where the platform offers rich content delivery options ( image, sticker, video, and audio) the bot should be able to fetch content from the institution's content repository with ease.
At askadmissions.ai we provide a content dashboard that helps create, sort, and store rich content in course-specific formats as well as topical classification of content. PDF, documents and language-specific video content may be tagged for location too. This allows the bot to receive user location using WhatsApp location extension then respond with content in the relevant language
In addition to answering queries, chatbots should also strive to keep the conversation going beyond the first query and response. This is especially critical for chatbots that are deployed at the point of student acquisition with marketers preferring longer first interaction that could lead to a better top-of-the-mind recall. Bots on WhatsApp have ready interaction components that could come in handy. Some options are
Using quick replies to nudge students to explore more about the responses provided, especially for course-specific queries
Using the list as a fallback to allow students to select from predefined faqs, is particularly useful in dealing with error response handling
List-based selection prompting, WhatsApp lists may have up to 10 options, to refine queries as well as lead students to delve deeper into responses
For all its benefits WhatsApp cannot replace the reliability and empathy of a counselor and it is no wonder students prefer to iron out their doubts with an admission counselor in case of an institution or a course consultant in case of edutech offerings
After initial qualification bot should reveal the identity and provide one-click links to connect with counselors
Finally, institutions and edutech decision-makers must also account for resources that could monitor and put the bot on the path of continuous enhancement. At askadmisisons.ai we provide granular controls and error visibility that allows administrators to view and handle errors while easily enhancing topical coverage
For more information on how WhatsApp chatbots could help automate your admission inquiry process, please connect with us on +91 9886399991 (Whatsapp). Wishing you all an excellent admission season.
Omni-channel tech stack for admission management and direct to student audience building. Omni-channel tech stack for admission management and direct to student audience building