Platform loyalty is collapsing. Audiences hop between streaming services, sign up for tools they abandon in weeks, and choose experiences over brands. The companies holding attention are doing it not through better acquisition funnels but through continuous, predictive engagement that meets customers between purchases. Most brands have not caught up, and the gap is widening.
Alexandra Pompei is Head of Go-to-Market at Alyena Legacy and a Global Partner GTM Executive. She spent nearly 20 years at Microsoft, most recently as Global Partner GTM Strategy Director, where she drove $2B+ in partner-influenced pipeline and led AI readiness programs across the U.S., EMEA, and Asia.
"AI is not a tool anymore. It's an engagement fabric. It's what connects the audience, holds their attention, and predicts what they'll do next," says Pompei.
Pompei frames the shift around a concept she calls the "engagement net," a data-driven experience layer that keeps customers connected continuously, not just at the point of transaction. The platforms doing this well, Netflix, Spotify, TikTok, are all building it on the same foundation: centralized customer data fed into predictive models.
"When you talk about loyalty, for me it is what we are earning from the customer in what I call the in-between," Pompei says. She points to moments like an AI assistant surfacing a recommendation inside a spreadsheet, or a workflow that saves 10 minutes without being asked. "That's how you create loyalty nowadays. And that's the big shift in marketing from now on."
"If you are doing marketing to get a funnel, forget it," Pompei says. "You get the acquisition. But then how can you keep this customer? The acquisition means nothing." She argues pipeline measurement should shift toward how well a brand keeps its audience engaged after the first action, not how many people entered the top of the funnel.
The engagement net runs on customer data, but Pompei sees most brands chasing the wrong signals. "The concern of the brands should be how can I capture data from my customer, instead of 'should I gain more clicks on Instagram' or 'I want more reviews,'" she says.
"When we talk about experience, we have to talk about data. How we are collecting the data, how we are translating the data, and how we are creating predictive models," Pompei says. The tools exist. What most organizations lack is the architecture to centralize data from multiple touchpoints and turn it into a continuous engagement model. "We have a bunch of tools. What we have less of today is an architect who understands how to combine all of this and create the net."
Pompei pushes back on the idea that only large platforms can build engagement nets. "It doesn't matter if it is a big company or a small company," she says. "When you are building the go-to-market, you should look at different touchpoints and predict what should be done." She is building this model at her own company, using AI across social platforms to track what customers open, read, and engage with, then delivering the right message at the right moment.
Pompei's strongest argument is organizational. Building continuous engagement requires more than a marketing team. It requires the entire company.
"You cannot build the engagement net only with marketing," Pompei says. "You have to have the entire organization working together with the same message, delivering the experience." She points to companies merging CX and marketing into a single organization, and in some cases embedding CX inside product development, because the people closest to the customer have the most valuable insights for building what comes next.
Pompei highlights Airbnb's expansion into services as a tangible example. After booking a rental, users are now offered massages, wine deliveries, and local experiences. "That's exactly what I'm expecting when I go on vacation," Pompei says. "That's the point from the brand: how can I keep this desire alive during the entire life of the customer?" Airbnb knows what to offer because it has the data and the predictive models to anticipate needs after the initial transaction.
Pompei sees this model becoming standard within three years. The data infrastructure is being built now. The AI layer is functional. "We are just at the start of AI," she says. "But with the minimum data we have today, we can already start building this engagement net. And it's working. It's a matter of how quickly companies understand this and how much they want to invest."
