How BeatO Overcame User Drop-Offs by Optimizing Blood Glucose Level Tracking for Diabetic Patients
About BeatO
BeatO is India’s healthcare App to help, monitor and smartly manage diabetes. It currently serves to 100K+ people all over India.
My Role
I was the lead UX designer. Let me explain my role in two parts:
Part 1 — Understanding requirements, creating hypotheses by acknowledging user needs, business goals, and validating the same through research.
Part 2 — Creating various help documents like task flow, scenario mapping, customer journey mapping, mind maps using data and insights to collaborate with the entire team.
Duration — Oct 2016 to Jan 2017
Team — 1 Designer (Me), 2 front end developer, 1 back end developer and 3 internal stakeholders (CEO, COO and CTO)
Objective
To reduce the user's drop-off rate and enhance the flow of the BGLT (Blood Glucose Level Tracking) feature.
Brief of BGLT (Blood Glucose Level Tracking) feature.
BGLT is the key feature for BeatO users. As a diabetic, the one has to keep track of their BGL(Blood Glucose Level) multiple times every day depending upon the severity of their condition.
For using BGLT (Blood Glucose Level Tracking) feature, BeatO app offers a physical device known as BeatO Glucometer. It is an external device developed by BeatO to measure BGL (Blood Glucose Level) of an individual.
Value of BGLT feature
This is not only a key feature but also the most valuable feature for the business. After a lot of efforts, BeatO team manages to build an entire product service which is benefiting the user and generating value to the business.
Discovery Phase
To discover and understand the problems of BGLT (Blood Glucose Level Tracking) feature flow, I started with the support tickets. We have the support tickets of users from two departments:
1. Customer Support
2. Diabetes Educators
Observations
After analysing the support tickets and identifying the problem, I decided to do Behaviour Research. I did my field research with 10 users by visiting their homes.
Findings from observations
So during my research, it’s evident to me that the users are having a significant problem while performing the first 3 steps of BGLT (Blood Glucose Level Tracking) feature flow.
Step 1 — Insert BeatO Glucometer into the phone
- Inserting glucometer in the reverse direction.
- Facing technical issues.
Step 2 — Insert the strip into the glucometer.
- Inserting the strip in the reverse direction into the glucometer and waiting for the further instructions.
Step 3— Prick & insert blood in the strip.
- Not inserting the right amount of blood into the strip.
- Inserting blood into the strip in the wrong manner.
Comparison with the Traditional Glucometer
When we shipped this feature earlier in Oct 2016, we had an informed assumption that most of the diabetic people are already using a glucometer device and they will be familiar with the process of using it. So we decided to ship the BGLT (Blood Glucose Level Tracking) feature with simple images and help text, but that didn't work well.
So this time, I study the function of Traditional Glucometer device to understand the key difference by comparing it with BeatO Glucometer.
Findings
From the above comparison, it's evident that the process of using the BeatO Glucometer is quite different from the Traditional Glucometer on specific steps.
Conclusion
Form the above observations and findings, I came to a conclusion that, due to the unique nature and different usage behaviour of the BeatO Glucometer, it's evident that the users need a robust learning graph.
The app should be able to help the users to perform the initial three steps of BGLT (Blood Glucose Level Tracking) feature.
Hypothesis (1st Draft)
As per the above conclusion and understanding of the problem, I tried to focus on the user learning curve, so I decided to create and add a few new elements in the existing screens.
1. Graphical presentation for initial 3 steps. Tried to keep the graphics accessible and straightforward.
2. Beato Glucometer Help Guide — In the help guide, I showed insightful guidance for using Beato Glucometer through graphical illustrations and also provided a “How to use” video option.
Hypothesis Flow (1st Draft)
Testing the Hypothesis (1st Draft)
After completing 1st Draft, the next step was to test it, and for that I showcased it to various internal stakeholders from different departments like diabetes educators, a few of customer support employees, CEO and development lead, to gather their feedback.
Accumulated Feedback
After the showcase and a few discussions, the overall cumulative feedback was positive in terms of the affordance and simplicity of the solution, but there are few things which can be added to enhance the solution.
1. Adding personalisation.
2. Solving the problem of unconscious incompetence of user behaviour.
Hypothesis (2nd Draft)
In the 2nd draft, I tried to incorporate all the above feedback.
1. The first thing I added is the personalisation impression through conversational UX.
I replaced "Help Guide" text with more personalised and relevant text in the first 3 steps.
2. To solve the problem of unconscious incompetence of user behaviour, I updated the graphical presentation with two states, right and wrong for the initial 3 steps.
The users who are unconsciously making some mistakes in the usage of BGLT feature now have more meaningful insights.
Hypothesis Flow (2nd Draft)
Testing the Hypothesis (2nd Draft)
Now we have the refined hypothesis ready, it's time to test it with external stakeholders (The Users). I wanted this to be validated by the users on a substantial level before we can launch it, so I tried to figure out how it will be done conveniently considering our constraints.
The first thing was to identify the other channel of communication that we use with our users besides the App, and there are a few mediums:
1. WhatsApp
2. Emailers
WhatsApp — Our customer support team is using the WhatsApp to communicate with the users. They usually address the queries and problem of the users, which also involved the usage problem of the BGLT (Blood Glucose Level Tracking) feature.
I created the Beato Glucometer Help Guide WhatsApp post and asked the customer support team to share it with the users who are facing problem using the BGLT (Blood Glucose Level Tracking) feature.
Emailers — The Emailers have been used in the User acquisition process and post-sale process of BeatO Glucometer.
I updated the post-sale Emailer with Beato Glucometer Help Guide.
User Response
On WhatsApp and Emailer, the support team received positive feedback from the users on "BeatO Glucometer Help Guide." Some of the users also mentioned that the help guide was beneficial to them in understanding the BGLT (Blood Glucose Level Tracking) usage process.
The final shipment of the BGLT (Blood Glucose Level Tracking) feature
Form the above repose of all the stakeholders. We decided to ship the BGLT feature and to measure the impact we used analytical tool CleverTap.
Impact
Measuring the impact is one of the most essential parts as here you will get to know the result of your hypothesis.
Data Insight (Oct 2016) of BGLT feature flow before implementing the changes.
As per the above graph, there was an 88% users drop off rate in completing the BGLT (Blood Glucose Level Tracking) flow.
Data Insight (Jan 2017) of BGLT feature flow before implementing the changes.
As per the above graph, now the 88% users drop off rate reduced to 34%
This shows the substantial success of enhanced BGLT (Blood Glucose Level Tracking) feature flow, although there is still 34% of users drop off rate, which needs to be reduced.
Improving a product according to the user needs and business requirements is a never-ending process.
Conclusion
Understanding of the problem in depth through user research and trusted collaboration with multiple teams and stakeholders helped in executing a robust user experience for BGLT (Blood Glucose Level Tracking) feature which not only reduced the user drop off rate to a substantial level but also helped to increase the User Retainment Rate and the DAU (Daily Active Users) rate over the platform.
