How can marketers use AI to drive better customer engagement tactics?

Priya Gill, CMO at Iterable, discusses more about AI-powered personalization, real-time decisioning, customer trust, and the shift toward continuously improving customer engagement programs.

________

Hi Priya, tell us about some recent innovations brands can use to their advantage to create deeper customer engagement.

The biggest shift we’re seeing is that AI is moving beyond content creation into decision-making and execution. Most marketing teams don’t struggle with coming up with campaign ideas, they struggle with orchestrating thousands of decisions across channels, audiences, timing, and personalization fast enough to keep up with customer expectations.

That might sound subtle, but it fundamentally changes the customer experience. It’s the difference between sending a cart reminder after an order has already shipped and recognizing, in real time, that the message is no longer relevant. Customers don’t think in channels; they think in relationships. Every interaction either builds or erodes trust, and AI is finally giving marketers the ability to make those decisions at scale rather than relying on static business rules.

We’re already seeing the impact. Brands using AI-powered decisioning have achieved higher activation rates, open rates, and click-through rates. To me, that’s the real promise of AI – not replacing marketers but removing the operational burden that keeps them from doing their best work.

The other major innovation is making sophisticated marketing accessible to more people inside the organization. Historically, creating highly targeted audiences often required SQL, engineering support, or an analyst. Today, marketers can build and refine those audiences themselves using natural language and AI-assisted workflows, reducing work that once took weeks down to minutes. That means teams can spend less time wrestling with technology and more time creating meaningful customer experiences.

Ultimately, the brands that create deeper customer engagement won’t be the ones sending more messages. They’ll be the ones making smarter decisions in every interaction – using AI to deliver the right experience, through the right channel, at the right moment, while giving marketers more time to focus on strategy instead of execution.

What are the three main benefits of adopting an AI-native customer engagement strategy?

• Changes that used to require an engineering ticket and a release cycle now happen directly within the marketing team’s workflow.
• Better decisions. Good AI doesn’t sort people into buckets; it reasons about what someone’s actually doing right now. That’s the real engine behind 1:1 personalization at scale — not a slide in a deck, an actual mechanism.
• Greater customer trust (which leads to better business results). Every mistimed win-back, every duplicate offer, every stale discount chips away at retention. Flip that around and it becomes an ROI story pretty fast — you see it in reduced churn and reduced friction, not in how impressive the model sounds on paper.

Get all three working together and you end up with two winners instead of one: the marketer who finally builds the thing they actually pitched, and the customer who gets an experience that feels like someone thought about them — because, for once, someone did.

What can marketers do to make automated personalization programs easier to govern and manage?

Personalization only works if marketers trust the system that’s making decisions on their behalf. As AI takes on more execution and autonomous agents become part of the marketing workflow, governance can’t be an afterthought – it has to be built into the architecture from day one. An agent is only as good as the guardrails you give it.

A few things I’d never skip:

• A shared, real-time view of the customer. Order status, open support tickets, lifecycle stage. If those signals are not connected, neither your marketers nor your AI can make good decisions.
• Clear governance over every customer interaction. Teams should know who (or what) is making a decision, why it was made, and be able to audit it afterward. Skip this and you end up with a ‘we miss you’ email landing in the middle of someone’s active support case.
• Treat suppression as a strategic capability, not a checkbox. The best customer experience is often the message you choose not to send.

One thing that’s stuck with me from a customer conversation: it wasn’t just that we could make real-time changes fast, it was that our team actually sat down and worked through their setup instead of handing them a boilerplate answer. Good governance doesn’t slow anyone down, but rather it’s what lets a team move fast without torching the trust they’ve built.

Marketing Technology News: MarTech Interview with Mark Listes, CEO @ Pendulum Intelligence

Why do you think marketing teams still struggle to act quickly on campaign insights and learnings?

It’s almost never a talent problem. It’s an infrastructure problem, plain and simple. Most stacks were built for a much slower world — rigid automations, opaque logic, a lot of technical drag — and customer behavior just doesn’t wait around anymore. Your data’s scattered across a CRM, a commerce platform, a loyalty system, and half a dozen channel tools that were never built to talk to each other. By the time an insight surfaces, an audience gets rebuilt, and a campaign clears approval, the moment that made the insight valuable is long gone.

So real-time orchestration quietly turns into batch processing, one small workaround at a time. And here’s what should actually worry every marketing leader: those workarounds never stay temporary. Enough of them pile up over enough quarters, and a team stops asking what the customer experience actually needs and starts asking what the stack can handle instead. That’s platform latency turning into strategic latency, and it’s exactly the gap the next generation of tools has to close.

What do brands tend to overlook when building long-term customer engagement strategies?

That most of them are still planning campaigns when they should be orchestrating journeys. In a campaign mindset, you decide what you want to send, then go figure out who should receive it. In a customer-centric mindset, you flip that around and start with the person: given everything we know about them right now, what’s actually the most relevant thing we could say or show? Small difference on paper, huge difference in practice.

Making that flip takes three things. You need signals from across the business pulled into a real picture of the customer, not five partial ones. You need decisioning that sits above any single channel, so email, SMS, app, and web are all working off the same shared understanding of that person instead of running their own logic. And you need to actually act on it in real time, because relevance has a shelf life — the right message a day late is just a wrong message. The brands that get this right stop thinking in individual campaigns and start thinking in one continuous conversation with the customer.

Most engagement roadmaps are still built channel by channel — email performance, app performance, web performance — instead of around the one relationship the customer actually thinks they’re having with you. That’s exactly how you get three ‘exclusive offers’ landing in someone’s inbox in a single day. Each channel is winning on its own scorecard while the customer just experiences it as noise.

I actually keep a running list of these — call it my personalization hall of shame. ‘Hi {First_Name}’ in a subject line. A cart reminder for something that already showed up at your door. A birthday discount that’s three weeks late. Nobody plans this stuff. Every marketer I know has shipped some version of it. But it’s a useful gut check, because every single entry traces back to the same root cause: nobody stopped to check whether the message still made sense for the person on the receiving end.

The second thing brands miss is how much the small compromises actually cost. Each one feels like a totally reasonable call in the moment, made to hit a deadline. But make that call enough times and it becomes a culture that designs around the platform’s limitations instead of the customer’s actual context. That shift is way more expensive than any single broken send. The whole strategy only works if both sides win — the marketer who ships the idea, and the customer who never has to think about the plumbing behind it.

Excessive messaging could be a barrier to stronger engagement. How can brands determine the right communication frequency so as to refrain from overwhelming customers?

Frequency should follow behavior, not a calendar. A volume-based send schedule is just a proxy for engagement, and not a great one. What actually works is event-triggered messaging that reflects how someone is interacting with what you’ve already sent, where they are in the lifecycle, and whether they’re in an active support or purchase journey.

I believe that when engagement dips, the instinct is almost always to send more. Usually the stronger move is the opposite — send less, but be more precise about it. The goal is identifying the point where an additional message begins to hurt engagement instead of improving it, and adjusting cadence accordingly instead of relying on blanket rules. Pair that with capturing someone’s attention at the moment they’re actually likely to act instead of guessing at a send time, and frequency stops being a policy and starts being something that’s actually managed based on real behavior.

If someone’s opened three emails about hiking boots and still hasn’t purchased them, the answer isn’t a fourth email with a bigger discount. It’s one well-timed question about what they’re actually looking for. Fewer messages, doing a lot more work. That’s the whole point. And this doesn’t have to be done by the marketer alone; AI can help identify the signals, the timing, and the next best action.

Tell us some early, often-overlooked signs that customer engagement is actually improving.

Before looking at day-to-day engagement metrics, I’d start with a more fundamental question: are you measuring the right thing? The most valuable signals aren’t opens or clicks, they’re the customer behaviors that predict long-term business outcomes. If your goal is lifetime value, what actually leads to it? A second purchase? Faster product adoption? Greater feature usage? Once you’ve identified those leading indicators, you can design journeys to influence them and measure whether they’re actually changing. And just as importantly, measure those outcomes at the journey level, not the campaign level. Customers don’t decide because of one email or one push notification, they respond to a series of interactions over time. The goal isn’t proving someone opened a message; it’s proving your engagement changed what they would have done otherwise.

From there, I look for a handful of leading indicators:

• Customers engage without needing bigger incentives. If repeat purchases, retention, or product adoption improve without increasingly aggressive discounts or promotions, that’s a strong sign your engagement strategy (not your offers) is doing the work.
• Customer friction starts to disappear. Fewer unsubscribes, spam complaints, and mistimed lifecycle messages tell you the experience is becoming more relevant, not simply less frequent.
• Time-to-value gets shorter. When new customers reach meaningful milestones faster, it’s usually because onboarding is delivering the right guidance at the right moment instead of following a generic sequence.
• Experiences become more consistent across channels. Marketing, customer support, and product interactions begin reinforcing one another instead of contradicting each other. Customers may never notice that explicitly, but they absolutely notice when every interaction feels like it’s coming from the same brand.

Increasingly, AI helps surface these signals much earlier than marketers could on their own. Instead of waiting for quarterly retention numbers, teams can identify shifts in engagement patterns, adjust journeys in real time, and continuously optimize toward the customer behaviors that ultimately drive growth.

Before we wrap up, can you share an example of an effective workflow or practice that helps brands build continuously improving customer engagement programs?

The best programs run on a genuinely tight test-and-learn loop. Ship a change, measure it against a real engagement or retention outcome, not just opens and clicks, then feed that straight into the next iteration on a cadence measured in days, not quarters. It’s literally our name sake — to iterate — and it’s exactly where AI is proving most useful right now: testing variants, fast, at a volume no human team could keep up with on their own. This is where operational discipline matters just as much as the idea itself. The teams I admire most treat this loop like any other operating rhythm — a standing review, a clear owner, and everyone agreeing up front on what ‘working’ actually means before the test even ships.

What makes the loop work isn’t the testing part, honestly. It’s killing the friction between the learning and the next move. If someone has to file a ticket and wait on engineering to act on what a test just told them, the loop breaks and the learning goes stale fast. The teams that keep improving are the ones where the person who spots the insight is the same person who can act on it, that same day.

That’s the model we push every customer toward, and it’s the same challenge I’d put to the whole industry: most teams are going to sit around waiting for someone else to hand them the AI-era playbook. The ones who actually win this decade are going to build the system that lets them write it themselves.

Marketing Technology News: Stop Measuring AI Adoption. Start Measuring AI Decisions

Iterable is an AI-powered customer engagement platform that helps marketers use customer data to create personalized experiences across channels and manage engagement at scale. 

About Priya Gill
Priya Gill is the Chief Marketing Officer at Iterable, where she oversees global marketing strategy, brand development, and demand generation.