The Future of Onsite Engagement: Key Takeaways from Our Latest Webinar

Earlier this week, we hosted a webinar to walk through the latest product updates and innovations we've rolled out.

Led by Ben, our founder and CEO, and joined by Lisa from our customer success team, the session covered the new and AI-powered features we've been building in 2025.

We wanted to show you how these new tools address a real challenge ecommerce businesses face: 80+ percent of website traffic leaves without converting.

Our focus is helping you close that gap and capture lost attention through smarter, more intelligent strategies.

The four new game-changing features for Wisepops users

During the webinar, we showcased four powerful tools that are redefining onsite marketing. They are built on our proprietary data intelligence system and AI capabilities, and they can work together seamlessly to transform your customer engagement.

Web push: reaching customers where they are

Web push notifications open an entirely new channel for connecting with your audience. Unlike traditional email and SMS, web push lets you reach visitors when they're browsing online but not currently on your site—directly through their browser. 

What makes this particularly powerful is how complementary it is to your existing marketing channels. We typically see opt-in rates between 0.2 to 1.5 percent of visitors, and web push campaigns can amplify your email marketing performance by up to 10 percent.

One of our key innovations with web push is flexibility in the opt-in experience. Rather than interrupting new visitors with a generic notification permission request, you can integrate the opt-in prompt into your popup strategy. This means you can invite subscribers at the right moment—after they've browsed a few pages, viewed a specific product, or shown genuine interest—and combine it with your broader engagement strategy for a smoother user experience.

AI cart recovery: smarter abandonment detection

The reality of ecommerce is sobering: only 30 to 40 percent of carts are converted, leaving 60 to 70 percent abandoned. Even more challenging, only a small portion of those abandoned carts can be recovered through email alone. That's where exit prediction comes in.

Traditional exit-intent popups have real limitations. They trigger too late (when someone's already decided to leave), generate false positives (triggering when visitors aren't actually leaving), and miss real abandoners entirely. 

Our AI cart recovery feature changes this game by analyzing multiple behavioral signals—page views, time on site, cart contents, and other factors—to predict which visitors are likely to abandon before they actually do.

The results speak for themselves. Our exit prediction algorithm is twice as accurate as traditional exit-intent, and delivers almost double the engagement. 

In fact, we recover 70 percent more carts compared to exit-intent approaches. Plus, the experience is smoother for customers, with reminders of their cart items, time-limited offers, and the option to save for later.

Experiments platform: measure what actually works

One challenge every marketer faces is isolating the true impact of their campaigns. With attribution models everywhere claiming credit for the same conversion, it's hard to know what's actually working. That's why we built a full-featured experimentation platform directly into Wisepops.

You can now run sophisticated A/B tests on any campaign element—popups, bars, notifications, embeds—against control groups to measure real incrementality.

Our platform tracks not just click-through rates, but your entire conversion funnel: bounce rates, page views, session data, time on site, revenue, and custom goals. We even use generative AI to interpret your results and deliver an expert-level report for each test.

This is particularly valuable for testing those high-stakes decisions like welcome offer coupons. Should you offer 10 percent off or $10 off? Instead of guessing, now you can test and know exactly which approach drives more revenue and profitability. And unlike enterprise-grade testing solutions, our platform is no-code, self-service, and lets you launch tests in about ten minutes.

AI product recommendations: personalization at scale

Product recommendations can be complex and expensive to set up, but they shouldn't be. Our collaborative filtering algorithm makes recommendations incredibly simple: one-click setup that powers real-time personalization for every visitor.

Rather than just showing products that work well together, our system learns what each individual visitor likes and makes recommendations in real time throughout their session. You can easily customize which products to include or exclude, and recommendations refresh weekly based on your catalog and traffic patterns.

The impact is significant—we see an average of 5 percent attributed revenue from product recommendation campaigns.

You can deploy them across embeds, notifications, and popups, with different strategies for different stages of the customer journey. For new visitors where you don't have much data, best-seller recommendations work great. As the session progresses, you can shift to more personalized suggestions like frequently-bought-together items displayed in your feed—where they drive the highest revenue per impression.

What's next

We're just getting started. Our roadmap includes multi-branch campaigns for creating dynamic quizzes and conversion funnels, enhanced cookie consent management, expansion of product recommendations to non-Shopify platforms, and AI-powered segmentation to help you understand your audience even better.

Looking ahead to next year, we're focused on automating even more of your onsite marketing using AI. We're identifying new opportunities automatically, creating more meaningful messages, and helping you scale personalization in ways that weren't possible before.

Thank you for being with us on this journey!