How I Leveraged AI to Drive $1m in Revenue and Improve the Shopper Journey for OtterBox Customers Online
TimeLine
January 2018 – depature.
Role
UX Lead
Technology
iAdvize, Demandware, Azure
Background:
OtterBox (the industry leader in protective mobile cases and accessories) faced a common problem in e-commerce: shoppers often struggled to find the right product due to the large number of options available and a lack of understanding of the make and model of their device. To address this issue, I identified and implemented iAdvize, a customer engagement platform that uses AI and brand ambassadors to provide pre-sales customer support on e-commerce sites. By using AI, we could help answer shopper product questions quickly through automation and provide an easy way for human assistance if requested, without taking away from internal customer service resources. This improved experience would reduce customer friction by providing means of self-service, increase product find-ability through recommendations and reducing returns, helping to reduce the e-commerce carbon impact.
Planning:
To ensure success, here are the primary things I did to prepare and organize the project:
Conducted extensive research to understand and identify the most common types of product questions asked by customers.
Aggregated review and FAQ data from the e-commerce store with data from our CS platform and IVR.
Worked with iAdvize Customer Success manager and the OtterBox Brand team to source Brand ambassadors. Ambassadors were asked to share their product questions and what information would be help them make a purchase decision while shopping on the site.
Engaged an internal team of Customer Service, Product Owners, Engineering, Creative and Product Marketing to help rewrite responses, and to ensure we had the best product information possible.
Defined some new and unestablished KPIs such as cost per chat, AI / Human chat utilization, resolution time, chat to conversion rate, and average response time.
Mapped out the user journey, making sure to consider and include non-sales related inquires to prevent the user from getting stuck and provide them with a solution to their needs.
Process:
With the necessary project plan and data established, we were able to move forward with developing the AI and implementing the solution for testing. This is what the process looked like in practice.
Worked with my customer success manager to use iAdvize to build and train the AI, using NLP (Natural Language Processing) and ML (Machine Learning) algorithms on the data collected.
Created training guides for brand ambassadors, ensuring both the AI and human representatives had and training on the same details.
Continuously trained the bot to become more proficient at answering consumer questions.
Designed experiments to continually learn how best to optimize this strategy, what questions users are asking and optimize them through NLP.
Optimized the UI to improve performance, which included designing interaction tests looking at placement, event triggers, and styling.
Bi-weekly meetings were set up to review performance, identify and review test insights, and touch base on upcoming product launches and create a process that would ensure the bot and ambassadors would have the details they need in time.
Additionally, monthly meetings were set up to review performance, identify and review test insights, and touch base on upcoming product launches and create a process that would ensure the bot and ambassadors would have the details they need in time.
Outcome:
The outcome of this project was a significant improvement in the customer experience by reducing friction in the shopping process and increasing conversions. By using AI and human assistance, we were able to answer shopper product questions quickly and provide an easy way for questions to be answered if human help was required, without taking away from internal resources. This approach helped to remove unconscious customer bias, and it became a valuable feedback loop between the customer and the business.
Chat to Conversion Rate: 10%
Average NPS: 68
Call Deflection Rate: 49%
AI / Human Chat Utilization: 85% AI / 15% Ambassador
Average Response Time: < 2 minutes
Resolution Time: < 6 minutes
These metrics allowed us to track the progress and success of the project, and make data-driven decisions on how to optimize the chatbot and improve the overall customer experience. It was a true testament to the power of AI and human collaboration in providing a seamless and satisfying shopping experience for our customers and is still implemented on the site today.
Through my leadership, I presented a compelling business case for using AI and brand ambassadors to provide pre-sales customer support on the e-commerce site, gaining buy-in from key stakeholders and leading a cross-functional team towards a common goal of enhancing customer experience, driving sales, and reducing customer service inquiries and returns.
Accomplishments:
Throughout the process I showcased strong problem-solving, analytical, and project management abilities. I identified customer pain points, led research efforts, and developed a solution that effectively optimized AI and human interaction to drive conversion and achieve the best ROI. I also guided the team to continually improve the strategy through ongoing analysis of user questions and NLP optimization. iteratively refining designs through user feedback, and quantitatively measured success using metrics such as Net Promoter Score, conversion rate, and cost-effectiveness, to effectively demonstrate the value for both the business and the customer.
I developed and applied knowledge and expertise in Machine Learning, Natural Language Processing (NLP) and e-commerce best practices to create a seamless and efficient experience for customers. Through this, I effectively strengthened the relationship between e-commerce and customer service, driving internal initiatives to collaborate and improve the customer experience, as well as streamlining technology roadmaps.
I learned that not everyone will like the new product, especially when it comes to chat. I received a lot of feedback internally from colleagues and from customers through iAdvize CSAT scores, website and customer NPS data all letting me know how much they hated it. This reinforced the importance of empowering the user with the ability to customize their experience by creating user options, especially when it comes to functionality that might be perceived as annoying and divisive as chat UI can be to some. This is an important aspect to consider for accessibility. Site on site customization options will continue to evolve and become the new normal.
Through this project, I was able to drive a paradigm shift within the organization, aligning customer service and e-commerce teams to work towards a common goal of creating a seamless, connected customer experience. Utilizing emerging technologies, specifically AI, enabled me to lead cross-functional teams in the development and implementation of a solution that not only improved the customer journey, but also fostered a more engaged and collaborative environment, seeing first hand the importance of holistic, user-centric design and the power of technology in driving organizational success.
Project 1
How I Strategically Optimized the OtterBox Purchase Funnel to Increase E-Commerce Sales by 66.7%
Project 3
How I Strategically Helped OtterBox Migrate from Demandware to Salesforce Commerce Cloud
Project 4
How User Research Helped Transform a Legacy Vendor System Through Customer Insights
Project 5
Streamlining Parking Permits: How Four Web Applications Simplified a Complex Workflow