Here is my take on the problem statements which were of particular interest to me. For ease of scanning, these have been structured as follows:
Metrics
Approach
Recommendations
Artefacts
Reducing drop-offs and improving the first-time user experience for learners.
peakperformer is a digital coaching platform that provides remote 1:1 coaching at all career levels with a aim to transform employees into global leaders.
Metrics
Improving drop off rate by xx%
Decreasing the median time to finish form from start to end
Improving D30 Retention
Read More
Approach
For learners, the onboarding experience was overwhelming. I tried to identify the condensed set of issues faced by them.
Condensed set of issues:
Users perceives the effort to be on the higher side.
Users don’t see the value of providing additional information.
No attempts to winback dropped off customers.
For coaches, I decided to target the existing coaches of peakperformer and what causes the coaches to churn.
Mismatch of expectations
Misaligned incentives
Recommendations
Improving awareness among learners
Adding Intro videos Helps in guiding users about the benefits they might not be aware of Introduce the users to the core objective or functionality of the peak performer.
Option to schedule onboarding call This will help in knowing learners well and help learners in understanding how peakperformer can help them to achieve their goals.
Asking relevant information
Asking information that is required and deliver value and removing fields which can be asked at later stage.
Removed fields like city, company name since they don’t fit in context and doesn’t justify why the ask was being made.
Recommendations
Adding progress bar
In an attempt to get more users to complete the steps, keep them motivated and reduce churn rate, we can add a visible progress indicator. Humans are biased towards completing tasks they are given. A progress bar reminds learners that they’ve been given a task, and they feel as though they should complete it.
Increase recurring demand
by onboarding new users and motivating existing learners to schedule more sessions.
Recommendations
Introduce loyalty campaigns where they would be rewarded for staying with peakperformer.
Proactive communication
Proactively communicate with them about the problems they are facing, content they are working on etc.
Help IELTSmaterial to increase the number of signed up users on the platform
Metrics
Median time spent by IELTS test aspirant on the portal
Read More
Approach
Understanding Users
Possible pain points
Can’t communicate with employers due to language barriers.
Not confident enough about his IELTS test eligibility, course contents, time taken for preparation, etc.
Growth Equation
IELTSmaterial Growth Model = Number of Sign up x Number of candidates who are able to maximise their IELTS score x Increase in the number of engaged users.
Recommendations
Increase the number of sign ups
Using Content Marketing
Conducting webinars, workshops and seminars in different colleges for final year graduate students
Creating a mobile platform
Conduct free Know your score test on the platform and provide IELTS Test Fee Scholarship
Recommendations
Focusing on communicating the magic moment by increasing engagement
Sending Push Notifications
Show them skill gaps where they are lacking and a plan to bridge those skill gaps
Proactive Communication
Recommendations
Create a reward based referral structure for existing users.
By providing reward points which they can use in future for availing mock tests. For example for referring to a 1 candidate, a user can get some X points (e.g. 250 points). On accumulating Y points (e.g. 5000 points) users will get a free mock test with a detailed blind spots report.
Incentive based referral programme where user gets incentive based on referrals. Every existing user of IELTSmaterial will have their referral code using which if any candidate joins then they get $200 in their account.
Identify friction points along within the product and provide solutions for the same.
Read More
Approach
Friction points within the product
Product demo videos not available for users across channels. Indic Crawler, tracks online mentions in the Hinglish language only, limiting the tracking to English and Hindi language only. Only Text based social listening Current AI based chatbot provided as a part of OneDirect messaging cloud is fluent in english only.
Recommendations
Create a series of Demo videos for OneDirect Product suite
Product demo videos not available for users across channels. According to HubSpot, 73% more visitors will purchase after watching a great product video. Create a series of Demo videos for OneDirect Product suite which explains how a OneDirect product suite works. This will help to build brand familiarity.
Recommendations
Collecting data in vernacular languages with ability to get reliable sentiment analysis.
Indic Crawler, tracks online mentions in the Hinglish language, with the ability to capture certain Hindi words .But that limits the tracking to English and Hindi language only. According to the Google KPMG report, 70% Indians find local language digital content more reliable. Furthermore, 88% Indian language internet users are more likely to respond in their local language as compared to English. We can improve that by collecting data in vernacular languages with the ability to get reliable sentiment analysis. For example, if our client is Domino’s then tracking in vernacular languages (here gujarati) should be available.
Recommendations
Offering Image Analytics for social listening
Social Listening of what is being said online about a brand or a product is done by text analytics only.Over 350 million new images are uploaded on Facebook and 85 Million images uploaded on Instagram every day. Along with text analytics offering Image Analytics for social listening as a feature as a part of Onedirect Suite to understand sentiment and brand perception can help to improve the Customer Experience.
Recommendations
Enhancing chatbots
Social Listening of what is being said online about a brand or a product is done by text analytics only.Over 350 million new images are uploaded on Facebook and 85 Million images uploaded on Instagram every day. Along with text analytics offering Image Analytics for social listening as a feature as a part of Onedirect Suite to understand sentiment and brand perception can help to improve the Customer Experience.
To improve the customer support service provided by BLU (Bajaj finance chatbot)
Metrics
Improving bot in different scenarios
Read More
Approach
Test the bot across following scenario
Functionality - What function does this bot perform for Bajaj Finserv? Does the bot handle use cases for this function?
Usability - How good is the bot’s conversational flow? General usability?
NLP - How well does the bot handle the user’s input language?
Business Recognition - What KPIs do businesses in this industry track? Doesthe bot’s conversational flow provide a way for the business to track these KPIs?
Design - UI/UX (including conversational flow)
Bot Persona - Is the bot’s conversational flow consistent at all times? How welldoes it handle small talk?
Recommendations
Emotionally intelligent
For the scenario stated to the right if response is given based on tone of the user with added smileys then it can make the bot a lot conversant and empathetic to the user Example response: I'm Good :) Thanks for asking. Hope you are doing well"
Accuracy of results
Image to the right shows the case of inaccuracy in delivering the content or responding with inacurrate message. This can be improved by optimizing the bot iteratively using conversational analytics and by adding NLP capability.
Recommendations
Message personalization
In a scenario when user is coming back to the website and is logged in then if some personalized experience by making the conversation friendly which makes him feel more understood and connected. Example response: Hey Davis, Your Application for loan is under review. Is there anything else I can help you with?
Recommendations
Offering Image Analytics for social listening
In BLU conversation flow works well for happy path scenarios but needs to avoid dead ends in negative path scenarios. This can be improved by 1. Redirecting users to different services Example: Your FD X1234AX's maturity date is 30/11/2020. a. Renew it b. Avail Loan against FD
Here in above example the bot is pitching Loan against FD to the customer who already has FD with Bajaj Finserv