Delight Is Not an Accident: Reflections from SendBird Spark on the Future of Customer Experience
The keynote at SendBird Spark on Tuesday, November 18th, opened with a familiar sound.
A robotic voice, the kind that has trained a generation of customers to lower their expectations, looped through the standard script.
“Your call is being transferred. All of our representatives are assisting other customers.”
Everyone in the room at The Foundry knew how that story ends.
You are not a person. You are a ticket.
Then John Kim, SendBird’s co-founder and CEO, walked on stage and made a simple observation that should unsettle every executive responsible for customers.
We are communicating more than ever.
We are connecting less than ever.
That tension sat at the center of Spark. The day was not about efficiency for its own sake, or one more layer of automation. It was about whether modern brands are willing to move from managing transactions to creating, building, and sustaining relationships at scale.
As someone who has spent three decades leading customer operations, and the last several years building a system to quantify emotional equity and brand affinity, that question is the one that matters.
From tickets to memory
John framed the problem directly.
Most of our systems were built to manage transactions. CRMs, ticketing platforms, routing engines, bots. They count impressions and conversions, but they do not understand people. They track what happened, not why it mattered.
That is why, even as our music and video apps anticipate taste with remarkable precision, most support interactions still feel like 2010. The technology exists. The architecture for genuine understanding has not.
SendBird’s answer is to treat every interaction as part of a living memory, not an isolated event.
The core of the announcement was Delight.AI, a completely new brand from SendBird, an AI concierge that sits on top of three ideas:
- Memory – Agent Memory Platform (AMP): “It remembers every customer and your business goals—so every reply feels intentional, personalized, and never starts from zero.”
- Personalization – For You Conversations (FYC): “Conversations evolve with each interaction, delivering responses that feel tailored to the individual, not the masses.
- Presence – Omnipresence: “Whether customers move from chat to voice to SMS, delight.ai keeps the conversation continuous—no restarts, no repeating, no drop-offs.”
In the demo, that looked like a simple shoe purchase.
A customer starts on a website chat to explore marathon shoes, gets interrupted by real life, and drops off. In most environments, that is where context and opportunity both disappear. In SendBird’s world, the interaction continues over SMS, then moves into a live phone call, pulls in an image of the customer’s current shoes, remembers fit preferences, closes the sale, schedules a follow up call after delivery, and even offers a weather reminder ahead of race day.
One continuous conversation. One unified memory. Many channels. No handoffs.
Technically, this is powered by the Agent Memory Platform (AMP), “FYC” conversations, and omnipresence, all governed by TrustOS as the safety and control layer. Strategically, it signals something else:
Customer experience is moving from workflow design to relationship architecture. This is the critical shift taking place as Customer Experience evolves toward Experience Manufacturing™. The shift from reacting to customers to architecting outcomes — manufacturing experiences that align deeply with your brand values. At Noom, we positioned the CX department as the “heart of the brand – inspiring, motivating, guiding, and empowering” members to be their best selves.
Delight as a system, not a moment
John used a phrase I agree with completely:
Delight is not an accident.
Delight can feel magical to the customer. It cannot be magical to the operator. It has to be designed, governed, and measured — manufactured — like any other critical system in the business.
If you strip away the product names, the design pattern SendBird is leaning toward looks like this:
- Sense intent and emotion in real time
- Remember the history of the relationship, not just the last contact
- Predict what the customer is likely to need next
- Decide what is right for the customer and for the business
- Act consistently, regardless of channel, time, or volume
This is Experience ManufacturingTM – the deliberate design and production of experiences that build emotional equity over time, rather than relying on isolated heroic moments.
What Spark showed is that we now have the infrastructure to run that system at scale. We can listen, remember, and respond in ways that feel tailored and alive, not scripted and brittle.
The Strategic Reality: When Models Become Commodities
As talk of an AI bubble circulates, in the space of Agentic AI, the reality many of the leading companies like Decagon, Sierra, and SendBird are faced with is that the LLMs are becoming a commodity. The nuance of building a profitable, scalable business in that world means managing model usage, protecting consumer trust, and seeking a true differentiation. If a brand can’t support its position long term, Corporate America teaches us that the low-cost provider will win.
With Delight.AI, SendBird is separating itself from the pack, showing a true vision of “service as a utility” and a framework businesses can use to feel safely in control of the AI-to-human interactions between the brand and customer.
It was refreshing to see — and a SPARK of what’s to come in the future.
A Leadership Community Asking the Same Human Question
One of the more humbling experiences for me at SendBird Spark was seeing my book, The MORE Effect Thriving In The Age of AI, as one of gifts for attendees. I was able to personally hand copies to Bhavin Shah, the CEO and co-founder of Moveworks, and Jake Heller, the CEO of Casetext. Two leaders operating at the frontier of applied AI — each asking, in their own way, the same question Spark was built around:
How do we move faster without losing what makes us human?
Their reactions reinforced something I felt throughout the event: the companies defining the next era of CX aren’t just building technology. They’re building philosophies of care, trust, and emotional connection — the very things that AI alone can’t manufacture without intentional leadership.
What leaders should take from Spark
Events like Spark are easy to misinterpret as product showcases. They are that. They are also early signals of where management practice is going. If you lead a customer organization, I would suggest three questions to bring back to your own team:
- Where are we still managing tickets instead of relationships?
Map your current systems against the simple pattern you heard from SendBird: memory, personalization, presence. Where do interactions die instead of continue? Where do we forget what customers have already told us? - What is our definition of delight, and is it operationalized?
If delight is a system, not a surprise, then it must have inputs, routines, checks, and measures. Can your managers describe that system, or are they still relying on frontline heroics? - How will we measure emotional equity over time?
Operational metrics tell you how efficiently you are running. They do not tell you whether customers will follow you into the next product, the next channel, or the next crisis. You will need a way to see and manage the emotional side of the balance sheet.
Peter Drucker often reminded leaders that “the most important thing in communication is hearing what is not said.” At Spark, what was said was impressive. What was implied is even more important.
The era of “good enough” customer experience is over.
We are moving quickly toward a world where every serious brand will need:
- A branded AI concierge that understands, remembers, and acts
- A discipline of experience manufacturing that treats delight as a designed outcome
- A measurement system that captures how those experiences make people feel, and how that feeling changes over time
The technology is here. The infrastructure is emerging. The question now is whether leadership will evolve at the same pace.
As for me, I left SendBird Spark with a pair of images in mind.
On one side, a voice on hold, repeating the same tired script.
On the other, an AI concierge that remembers your marathon, your shoes, your weather, and your name – and a leadership community starting to ask what that power should serve.
The distance between those two worlds is where the next decade of work will be defined.
And if we get it right, we will not only build better systems.
We will build better relationships.
We will create deeper connections.
We will Become More.
Discover The MORE Effect
A field guide for leading and living with courage, hope, and optimism in the AI-first, always-on world.
About Jeremy
Jeremy Victor is a humble, hard working and the author of The MORE Effect. It is the culmination of his life’s work. With one decade as a kid, and three decades in Corporate America, capable of writing, designing, building, and caring, he’s driven to help others become more. No matter the role, no matter the moment, service to others, achieving desired outcomes, and being a good person matter most to him.
He’s the voice behind the Business at the Speed of AI podcast and newsletter, where he explores the intersection of leadership, modernization, and the AI-first always-on economy. His work is guided by a simple belief: that operational excellence and human connection aren’t trade-offs—they’re the future.
His mission: help people—and the systems they work in—Become More.
Join Jeremy online: Website – www.jeremyvictor.com | Newsletter: Business @ the Speed of AI Facebook | Instagram | TikTok OneMORE Minute |
