The Wave
From 40-second times to instant resolution at scale.
The Wave
When dinner-hour call volume spiked 8× and hold times hit 11 minutes, The Wave needed infrastructure that could scale instantly. They deployed CAPSE in 5 days. The results changed how they think about support permanently.
Industry
On-demand delivery
Scale
35M+ orders / month
Use case
Customer & Dasher support
35 million orders a month. And a support operation that couldn't keep up.
The Wave operates at a scale most support teams never have to think about. On a typical weekday, their customer and Dasher support lines handle hundreds of thousands of calls. On a Friday evening between 6 and 9 PM, that number multiplies. Missing orders, late deliveries, payment issues, account disputes all arriving at once, from every time zone they operate in. Their existing support infrastructure a combination of in-house agents and outsourced call center partners was built for predictable volume. Dinner hour peaks weren't predictable. Neither was the quality that came out the other side. "We were hiring ahead of demand and still falling behind," said James Liu, VP Operations. "The math didn't work. More volume meant more headcount, more training, more overhead and we were still seeing CSAT scores that made leadership uncomfortable.
Three problems that compounded into one broken experience.
Volume spikes that broke the system
Dinner-hour call volume peaked at 8× baseline on Friday and Saturday evenings. Human agents couldn't scale to meet it. Wait times during peaks regularly exceeded 11 minutes well above the 3 minute threshold The Wave had set as their internal SLA.
Hold times that lost customers before the call started
A customer calling about a missing $15 order expected a 2-minute resolution. What they got was an 11-minute hold, followed by an agent who needed to re-pull their order details from scratch. By the time the call was answered, the customer had already decided what they thought of The Wave.
Inconsistency across shifts and partners
Between in-house agents, outsourced partners across three time zones, and weekend coverage, DoorDash was effectively running multiple support operations with different training, different scripts, and different escalation thresholds. Customers experienced a different The Wave depending on when they called.
X8
peak volume vs. baseline on Friday evenings
11m
average hold time during peak hours
3.2
CSAT score below acceptable threshold
What About Us
"I was blown away by the storytelling in every frame. The vintage film simulation added such a nostalgic, warm feel to our session, capturing those intimate, fleeting moments we would have otherwise missed."
1/3
Business results, made legible.
60%
↓ from $4.2M to $1.7M / quarter
Reduction in support opex vs. same quarter prior year
4.7
↑ +47% improvement in 90 days
CSAT score up from 3.2 at deployment
94%
Human agents now handle complex cases only
Of calls resolved without human escalation
180ms
↓ 99.97% faster than human hold time
Average first response time down from 11 minutes