top of page

SiplyBot

A chatbot interface to provide users a humane experience to solve their repetitive queries.

My Role

  • User Research

  • Content Strategy

  • Conversational Design

Worked with

  • Product Managers

  • Customer Support Executives

  • Development Team

Tools used

  • Miro

  • Notion

  • Adobe XD

What was the problem?

1

Increasing number of repetative queries from customer

2

User prefering humans (customer support) to explain them the solution

3

High-cost of hiring new customer support members

Identify repetitive queries
Identifying and shortlisting the repetitive conversations.

What is the solution?

Design automated interactions

Converting the repetitive interactions into an automated conversation flow.

Make the conversation human

Structuring and crafting the voice and tone of the conversation to sound like a human.

Users are very emotionally invested 

When the users spend money or want more information about the benefits, they are highly motivated to get answers instantly.

Users prefer human to solve their query
Even though the product flow and FAQ answer the query, most of our users prefer a human to help them through the journey.

Insights from initial research

71% of the users choose local languages to communicate
When explaining the queries to the customer executive, they explain their issues in local language.

Ideating to find the suitable solution

Categorising the conversations

We built a notion sheet with all the repetitive flow

Identifying the available solutions

Conducted competitive research the available approaches

Choosing the right approach

Discussed and finalised for Rule-Based chatbot

Categorising the conversations

How are companies using
Chatbots?

Your paragraph text (4).jpg
1. Be Friendly. Be Human.

Write as if they are talking to a human.

Defining the content strategy

2. Be Concise.

Simplicity cuts through clutter.

3. First prevent. Then Fix.

Be proactive when providing a solution. 

4. Guide them step by step.

Use bullet point to make the content more actionable.

Your paragraph text (5).jpg
Chatbot Final Final.jpg
Your paragraph text (7)_edited.jpg

Result

After releasing the first version, the total customer calls were reduced by 19%.

Challenges

  • Understanding the variable ways different users can frame their questions.

  • Localising the content into 10 other languages because of technical constraints.

  • Couldn't incorporate media to help the user with more personalised solutions.

Learnings

  • Learning the concepts of NLP and how "intent" works to help the user get their desired solution.

  • Learned to plan and work around looped conversations to complete the communication circle.

bottom of page