All things AI with Samya DasSarma

TOGETHER WITH

Enter the apex customer: A new breed of ultra-savvy customer is on the rise. Iterable’s CTO helps you keep up in the new age of high-tech marketing.

Samya DasSarma is the CTO of Iterable, and if you happen to follow our Instagram, you’ll see he’s also officially been “hired” as our first-ever resident CTO, volunteering his time and POV to answer the tech questions marketers are too afraid to ask in their day-to-day.

Before Iterable, he spent over 20 years building technology in Silicon Valley, including stints at Twitter (through its IPO), Medallia, and Box. We sat down with him to talk about where marketers get their relationship with technology wrong, what the next few years of AI actually look like in practice, and the rise of what he calls the “apex customer,” a new breed of shopper savvy enough to add something to a cart just to bait a discount code.

Let’s start with the basics. What’s your background, and what got you here?

I’ve been in the tech industry for 26 years, all of it in Silicon Valley. I’ve seen how the tech gets built here, how hype cycles form, and how great tech survives them. That’s what I like to build. I was at Twitter through the IPO, at Medallia, at Box, before joining Iterable. At the core, I’m a technologist. Companies are almost like children: They have different personalities, and they grow. Having a hand in that growth is what gets me excited.

You’re our first-ever resident CTO. Given that responsibility, what would you say marketers are getting wrong now about how they think about technology?

The fundamental thing is treating it as a strategy instead of an amplifier. Technology is a strategy for the IT department. For marketers, it’s an amplifier. Teams adopt the latest AI tool, expecting it to fix a broken customer journey or messy data. If your foundation is fragmented, technology will only scale that dysfunction faster. One of my mentors used to say automation is the fastest way to disaster if you have the fundamentals wrong. The best marketers understand that tech accelerates what’s already there. You have to get the foundational logic right.

You’ve talked about a line between syntax and logic. What does that mean for marketers as AI takes over more of the technical work?

A modern marketer doesn’t need to write code, but they absolutely need to understand how their data flows. Where does the customer signal originate? What does it trigger? What are the constraints in the system? That’s logic. Syntax is the code itself, and that’s what platforms like Iterable are built to siphon away from the marketer’s responsibility. As agents (like our own Nova) get better, I think the expectation for logic goes up, and the expectation for syntax goes down. Marketers will start working with agents the way a product manager works with an engineer: You convey the strategy, and the machine handles even more execution than in the past.

What’s the question you wish marketers would stop asking you? And what’s the one you wish they’d start?

The question I hear most is “Can this tool do X?” But almost any enterprise tool can check a feature box if you throw enough engineering at it. The better question is “What happens when you scale that feature to 10 million users and five channels?” It’s natural to focus on the problem right in front of you. But when you choose a platform or tool, you’re not solving for today—you’re committing to years of partnership. That means looking beyond questions like “Can this tool do X today?” and instead taking a systems thinking approach: asking how this tool will evolve with your organization and help you get to where your company is headed, not just where it is today.

You’ve mentioned the rise of the “apex customer.” What is that, and what does it mean for how brands need to operate?

The apex customer is the most savvy we’ve seen. The kind of shopper who uses AI tools to shop and looks for ways to play the system. We see it in the numbers. 70% of consumers game their online carts to score discounts. 55% delay their purchases to wait it out for a better offer. 64% start a free trial with no intention of sticking around long-term. That’s not fringe behavior anymore. That’s a majority of customers treating marketing like a system to work.

And honestly, I get it. They’ve figured out how the game is played, so they’re playing it back at us. And that means marketers have to make a real shift to shorten the distance between customer behavior and how they respond. Because if the customer’s adapting in real time, marketing has to do the same, and do it faster. That doesn’t mean automating every decision and hoping it keeps up. It means paying attention to what’s actually happening right now—this cart, this hesitation, this trial about to expire—and closing the gap between that behavior and our response as fast as we possibly can.

Look into your crystal ball for us. What should marketers actually expect from AI in the back half of this year and into next?

Short term, expect a real shift toward agentic workflows, AI moving from a passive copilot to something that can execute multistep tasks on its own. Longer term, the bigger shift is the data layer and the activation layer finally becoming one system instead of two. Once that happens, marketers stop waiting on engineers to build pipelines and get to spend their time on what they’re actually there to do, which is to drive growth.

This conversation has been edited for length and clarity.