Case Study: Reducing customer support tickets through enhancements

This Event Ticketing program is my bread and butter at Cadillac Fairview. My role in this case study has been a broad range of Product Management.

I established the data analytics framework that captures insights for both ticket sales platform and the customer experience and support process. The analytics have enabled the cross-functional teams to make data-driven decisions on enhancements and correct on-site processes in a timely manner.

After the event season, I also led workshops and enabled the cross-functional team to pinpoint the root causes of complex issues, then ideate for solutions together. We learned each problem from multiple factes, ideated in silence, then dot voted. This process has multiple purpose:

1. More brains equals more brain power. Everyone can contribute to what they observed and what they think the problem is.
2. As we ideated silently, we are not primed by other people's ideas, which could lead to untapped potential of other perspectives.
3. Good solutions may not be the first few ideas we come up with. To catch a big fish,we catch a lot of fish and throw back the little ones.
4. Stakeholders are on board with solutions from the start. No need for lengthy buy-in process.

‍

Executive summary

Context

Cadillac Fairview’s traditional event Visits with Santa welcomes families to come take pictures with Santa in 5-minute time slots.

Problem Statement | Pain points

In 2022 Holiday season, customers experienced long wait time for support due to unexpected number of requests.

2023 pain point coming soon.

Objectives

1. To reduce the number of customer inquiries 
2. To enhance the ticket booking experience
3. To augment the on-site visit experience
4. To bring customers information they need as quickly as possible

Constraints

1. Vendor-based service - leverage out-of-the-box features
2. Budget constraints on the ticket sales platform enhancements.
3. Leveraging existing customer service capabilities.

Results

57% reduction in number of support tickets

81% reduction in confusion about how many tickets a family need to book

40%+ less inquiries due to customer “not receiving confirmation emails” 

Faster ticket resolution time. Happier customers.

‍

Trade-offs

Average booking time increased from 1m22s to 1m45s 

Slightly increased efforts in support model

Context and Pain Points

Cadillac Fairview organizes Visits with Santa at its shopping malls every Holiday season. It’s a paid ticketed event with 5-minute time slot for each group to take pictures with Santa. Every season, they sell 45,000+ tickets across Canada.

In 2022 Holiday season, the number of customer emails exceeded what we expected. This resulted in long wait time for a customer ticket to get resolved. Some customers would not hear back until after their appointment.

2,680 support tickets created per 1,851 unique customer requests

105 Average emails per day. Peaked at 457 emails / day

10 days average resolution time. Some tickets aged 20-40 days without a response.

Objectives

Improve the overall customer experience, from before booking to post-visit including preparing for their appointments and photo experience.

1. To reduce the number of customer inquiries 
2. To enhance the ticket booking experience
3. To augment the on-site visit experience
4. To bring customers information they need as quickly as possible

Constraints

Main focus is the in-person experience. The Ticket Sales experience and support process should be improved at minimal cost.

1. Vendor-based service with both the Ticket Sales platform and Photo Booth provided by vendors
2. Budget constraints on the ticket sales platform enhancements. Make the most use of Out-of-the-box features with flexibility for essential enhancements.
3. Leveraging existing customer service team and platform capabilities.

Pinpoint the problem

We gathered quantitative and qualitative data from web traffic, customer emails and survey.

1. 2,143 customer emails received (4.2% of total reservations including cancelled ones). 2,680 support tickets created.
2. Lots of issues with the photography process and the experience to get the photos.
3. Many tickets with avoidable inquiries / issues such as asking to confirm the appointment time (not getting confirmation email or typo in email), and confusion about how many people is covered under 1 booking. 

Empathize and Ideation

As a team, we picked the more complex problems to have special workshops to empathize and ideate solutions. For more simple problems, simple discussions as lessons learned were facilitated.

1. Align on a problem statement.
2. Keep asking “Why?” to find root causes and break down the problem. Use data to validate hypotheses.
3. Diverging - Silent ideation to generate lots of ideas
4. Converging - Clustering ideas and dot voting to narrow down the most valuable solutions

Solutions

We discussed the proposed solutions with relevant stakeholders to validate the feasibility of the feature or process. Major changes include:

1. Issue: Photo issues

Summary of root causes:

• Vendor product and process left room for human errors.
• The triaging process of having CF to be the 1st point of contact for everything including photos flooded our email inbox.

Solutions:

1. Changed photography vendor
2. Directed customers to contact the photography vendor for any photo-related requests

2. Issue: Not receiving confirmation email & Incorrect email address

Summary of root causes:

• At least 50 typos and email address change requests identified
• Customers accidentally deleted their emails
• Customers booked with their different email addresses then forgot, and checked the wrong inbox

Solutions:

1. Disabled paste command for Confirm Email Address field - customer has to type their email address twice
2. Monitored email bounce rate

Result:

In 2024, we only identified 16 cases of typos, which means a 68% reduction thanks to disabling copy paste command. To achieve this number, whenever my team resolves a ticket categorized as an email issue, I asked them to note down their triage process. Using that information, I then tag each of those tickets with a subcategory including "Typos".

3. Issue: Confusion about how many people one booking covers

Summary of root causes:

• Each time slot was 15 minutes but accommodated 3 groups (each group got 5 minutes). The count beside each time slot said “Spaces: 3”. This was combined with user mental model of booking 1 ticket per child of the previous years.
• Customers finished booking in 1 minute 22 seconds on average and did not read information on the booking site which says 1 ticket covered up to 6 people.

Solutions:

1. Verbiage change from “Spaces: 3” to “Groups: 3”
2. Experiment 5-minute time slots at a lower traffic event

Result:

In 2024, there are only 2 tickets related to the confusion, both of which are not associated with the experimental event. This means there is no significance in using 15-minute time slots or 5-minute time slots.

I attribbute the success to clearer copy and sticking with the structure that the customer already learned last year. Sometimes, having no changes is intuitive.

Overall Results

2022 Stats

2,680 support tickets created per 1,851 unique customer requests

105 average emails per day. Peaked at 457 emails / day

10 days average resolution time. Some tickets aged 20-40 days without a response.

2023 Stats

1,164 support tickets created per 1,107 unique customer requests / feedback

50 average emails per day. Peaked at 178 emails / day

75% tickets resolved in 48 hours or less

2024 Stats (Coming Soon)

Trade-offs

Changes came with some trade-offs that were within our acceptance.

1. Average time to finish a booking went up from 1m22s to 1m45s due to disabling the Paste function in Confirm Email Address field.
2. More efforts to aggregate the support data from the photography vendor with ours.

Next Steps

Keep doing what is working well. Aim to reduce customer requests without compromising business benefits.

1. Review rescheduling and cancellation / refund policy
2. Improve facility processes to ensure stable internet connection for a smoother Photo experience, to reduce photo-related requests.

‍

‍

‍

‍