Helping patients and doctors in a broken healthcare system

Slingshot Health is a healthcare marketplace where patients have the power to name their price and receive healthcare on a time that fits their life and providers can fill their open slots and last minute cancellations with new patients.

Slingshot Health conversion funnel had not have any significant improvement in conversion rate for the past 1.5 years and was believed to be in need of a redesign to update the funnel to fit best standards and generate more trust in new users.

My Role:

Entire product design from research to conception, visualization, testing and hand-off to engineers.

Understanding the problem

Thanks to our customer service department helping me dive through the past year of customer feedback and with the help of product, a list of user pain-points and product ideas was created.

Top user pain-points

Once I gather enough information, I proceeded in collaboration with our product manager to prioritize the more pressuring pain points for the user:

I started the ideation process by creating the following use cases and wireframing possible solutions for the detected pain points.

USE CASES

1. User knows exactly what service he/she needs: I want to get a dental checkup and cleaning

2. User knows what specialist to see: I want to see a Dentist

3. User is experiencing pain (symptom) and know they need to be seen by a doctor: have back pain, I need to see someone about it

I knew I needed to solve for three things in the same flow, symptom, specialist and service. Service was pretty straight forward, specialist a bit more complex if the user wants to pick something different to a general consult. Symptom was the most complex as they can be linked to different illnesses and treated by different specialists. 

The Aha! Moment

In the ideation stage, while I was thinking through the flow, I realized something... There was a pattern: users wanted more information and clarity, users were in need of more guidance!

The truth is healthcare is too complex, how do I as a patient know what doctor to see? where to go? what’s the fair price to pay? what is going to happen in that visit, what’s included and what are you going to do with my information?. 

For example, If you have toothache or your knee is hurting, you are aware of the symptom. If you know or suspect the cause of the symptom, you might be able to pick among different specialists but maybe you don’t know what specialist to see because you simple don’t know what is the name of the doctor that treats the cause of the problem. If we educate the user, and empower them with knowledge, we would increase the probabilities of them seeing the right specialist without us playing doctor.

Once the wireframes were done and share with Product, I proceeded to share them too with a healthcare provider to gain also insights from the doctor’s perspective.

I proceeded to create low fidelity designs and a prototype to help us test the “education” hypothesis and a type-ahead as a better way to simplify the search/selection of services.

Participants were given several different health scenarios and asked how they would find the service they need.

This prototype needed to allow the user to actually search and type information so that we could observe their behavior, so I use a new tool called proto.io that made that possible.

Success!

The results of the user testing confirmed we were in the right track, participants seemed to utilize the flexibility of the typeahead. Different people search for each prompt differently, showing that there was value in keeping the search flexible for different search types.

And the most exciting result: “Education” was of great value for the user.

Room for improvement

1. Participants seem confused entering the flow at different stages: the flow felt different depending on their search (symptom, specialist or service).

2. Many of the participants were caught off-guard by the number of sessions question.

3. When selecting general consult or searching for a specialist without entering a symptom, participants were looking for the option to enter more information about their condition they could tell a doctor but couldn’t.

4. Even though we were not testing the service concept this time around, participants were still confused by the “name your price” language and mentioned wanting to see a list of doctors with reviews. It will be important to always keep in mind that Slingshot’s service model is different than what most users are used to, so the value proposition needs to be very clear and upfront to prevent confusion and drop-off.

Divide and conquer

After gathering all the feedback and have confirmed that an educational approach was of value for our users (and actually can differentiate us from competitors) I proceeded to iterate and created high fidelity designs for the first steps in our funnel (selecting a symptom, specialist or service and narrowing it down to an specific service).

As we work in Agile, we have broken down the redesign of the conversion funnel in steps (search, availability, price, signup, patient information, payment) which will allow us to ship faster and perform A/B testing on specifics steps so we can refine or iterate on them to improve the experience and increase our conversion rate which is our main goal.

Early results

These project is work in progress. We have released a new “Sign up” form and “Payment” Step and observed the following results:

1. Requests placed went from 10% to 17% = 7% increase.

2. Accounts created went from 10% to 36% = 26% increase.