Sib Raza

Background:

Brightstar operates in the device insurance space. It partners with more than 200 wireless carriers in over a 100 countries. Brightstar enables customers to file insurance claims for lost, stolen, and damaged devices through an omni-channel approach. Annually, 27 million subscribers elect to file device protection insurance claims on the online claims web site (i.e. Subscriber Online Claim). As convenient as it is for many customers, the online site can be susceptible to many different types of fraudulent attacks such as account take over, synthetic identity, automated attacks and scripting, etc.

Problem:

Currently, Brightstar investigates these sophisticated fraudulent attacks through a solution that requires manual intervention. This type of approach is difficult to scale when the number of filed claims increases. Additionally, it fails to properly assess fraudulent behavior due to its reliance on manual intervention.

Research:

I began my research by parterning with data analytics. It was determined that the number of filed claims increases by “x” % YoY and an increase of file claims meant an increase of fraudulent claims. Traditionally, the solution would be to invest more in Asset Protection labor. In fact, Brightstar’s cost of labor increased by “x” % compared to the previous year. Ultimately, it can be inferred that this type of linear relationship would limit our ability to scale quickly and properly assess fraudulent behavior.

Solution:

As a result, I presented to the executive team a solution that involved partnering with vendor "X". vendor "x" provided an automated solution that could be integrated with our application. Once the solution was approved, I worked closely with several different stakeholders on how to implement and execute this solution.I also worked closley with data analytics to ensure that the results from the solution could be quantified correctly.

Solution Benefits Included:

• Redirection of current fraud review spend would fully funds “x” cost of service.
• payback period for this service is roughly 4 months.
• Solution provides access to Vendor's consortium network
• Over "x" billion data points that Brightstar can see regarding Fraudulent Activity
• Less Friction in application since "bad" customers are filtered out quickly.

Results:

$6.0 million in Cost Savings

Time to Process Claims Reduced by 2 Days

8% Increase in Net Promoter Scores

Reduced Churn Rate by 7%

Problem:

Customers prefer to file claims online. However, the online website had several webpages in the customer journey that was causing unnecessary friction to the customer.

Some of these challenges were:

• Vague or not enough instructions
• Poor navigation from one page to another
• Certain elements are not consistent on website

Research:

I worked with UX/UI to establish personas for each client. We listed several possible challenges that each client group could face as the demographic and education level for each client group was different. We conducted user testing and collected feedback from the interviews. I also worked with sales and customer service to gather any feedback that customers might have provided to them.

I also partnered with data analytics to quantify the problem. We determined that certain webpages had high bounce rates. In addition, visitors spent a lot more time on specific webpages compared to our normal time. Lastly, we also determined how many calls were made to the call center regarding an online site issue. After identifying the pain points that our customers and internal stakeholders were facing, UX/UI and I created several mockups and implemented an A/B test. The results from the A/B test were deployed to production once alignment was achieved with IT.

Results:

$1.2 million in Cost Savings: Lowering Call Volumes

Bounce Rate Decreased by 55%

Reduced Time on Site by 9.5 minutes

Increased Net Promoter Score by 6%