7 Personalization Things Klaviyo Can't Do (and What You Can Do Instead)

Personalization algorithms face several key challenges in ecommerce:

• Sparse data, especially for new or infrequent shoppers, creates cold start issues and forces reliance on generic defaults—like showing bestsellers.
• Disjointed data sources—user profiles, event logs, store catalogs, and web analytics—are often siloed, making it hard to generate a unified view of the shopper.
• Buying journeys are increasingly fragmented, with shoppers interacting across multiple sessions, devices, and channels.
• Every shopper is at a different stage in the buying lifecycle, from first-time visitors to loyal customers, requiring different messaging strategies.
• Shopper behavior is time-sensitive and influenced by seasonality, trends, and context, which personalization systems must continuously adapt to in order to stay relevant.

Conventional generative AI in email marketing typically focuses on creating or personalizing written content—like subject lines or body copy—based on user profile data. This isn’t new; it’s simply the application of language models using word prompts to generate text. While useful, it’s a direct and limited use of existing generative AI capabilities.

Generative AI for ecommerce takes personalization to an entirely new level. Rather than simply generating text, it draws on a wide range of data inputs—shopper clickstream behavior, user profiles, product metadata, and performance uplift metrics—to dynamically predict which products are most relevant to each individual shopper. In this context, the input is customer and product data, and the output is a personalized stream of shoppable products, rendered just as they would appear in the online store.

• Rule-based personalization is simple and scalable for broad segments but lacks depth and becomes unmanageable at scale.
• Collaborative filtering offers strong one-to-one relevance by using crowd behavior but struggles with new users or products.
• Content-based filtering personalizes based on product attributes and individual preferences, working well even with limited data—but risks becoming repetitive and narrow in scope.
• Hybrid approaches combine methods to boost accuracy and cover each other’s weaknesses. They offer excellent relevance and personalization depth but require more technical complexity and computing resources.
• Generative AI represents the most advanced form of personalization. It not only recommends relevant products but dynamically generates tailored content, tone, and language for each recipient. This enables one-to-one messaging at scale with unmatched depth, though it comes with higher infrastructure and data handling requirements.

In short, the more advanced the method, the deeper the personalization—but also the greater the need for data and processing power. Each approach has trade-offs, and the right choice depends on your goals, scale, and available data.

Klaviyo’s data layer introduces several structural limitations that impact usability and restrict advanced personalization:

• Personalization data is fragmented across user profiles and event datasets, making it difficult to form a unified view of the shopper.
• Behavioral signals are embedded within Klaviyo’s event structure and are not directly accessible at the time of email delivery, limiting real-time personalization.
• While Klaviyo supports event-triggered automated flows, it does not allow multiple events to be combined within a single flow—restricting more advanced, behavior-driven personalization strategies.
• The Klaviyo–Shopify integration forms a closed loop that blocks external access to Shopify’s product catalog and performance data—two essential inputs for generating dynamic and highly relevant product recommendations.

Email Pulse is designed to work around Klaviyo’s input data limitations:

• It is creating its own unified data layer that combines behavioral signals, customer profiles, product metadata, and performance analytics—all outside Klaviyo’s closed ecosystem and accessible in real-time.
• Rather than relying on a single event trigger like Klaviyo flows, Email Pulse uses multi-signal modeling to evaluate a shopper’s lifecycle stage, intent, and preferences in real time.
• It bypasses the rigid structure of Klaviyo product feeds by generating personalized product streams independently, then embedding them directly into emails using universal content blocks.
• Email Pulse taps into Shopify data and Google Analytics to access store performance and product engagement metrics that Klaviyo can’t see.

Only a handful of companies attempt to enhance Klaviyo emails with personalized product recommendations—and most follow Klaviyo’s pattern. The prevailing technique is to use Klaviyo’s event tracking to “pre-bake” product recommendations as an HTML content block embedded within the event itself.

However, these solutions inherit the same limitations as Klaviyo itself:

• They rely on outdated personalization technology.
• They use a single input parameter.
• They generate recommendations that are static—based on behavior captured at the time of the event.
• These systems are also closed, offering little to no flexibility for customization.
• Critically, they lack the ability to track user interactions with specific product recommendations, making it impossible to measure performance or optimize based on actual outcomes.

Klaviyo provides only surface-level email performance metrics—deliverability, open rates, click rates, and attributed revenue. What’s missing is visibility into what happens after a shopper clicks through and visits the online store. This disconnect blinds email marketers to customer behavior and preferences beyond the inbox, making it nearly impossible to establish a feedback loop for ongoing optimization.

Email Pulse bridges the data divide by connecting what Klaviyo leaves disconnected—email behavior and on-site shopping activity. It integrates data from Shopify, Google Analytics, and Klaviyo to create a complete picture of each shopper’s journey, from inbox to purchase. This unified view enables real-time insights into how email-driven traffic behaves in the store, what products they engage with, and what drives actual revenue.

Klaviyo offers personalization as a value-added feature to its core email and customer data platform. This approach gives merchants a helpful starting point, enabling basic personalization capabilities out of the box. However, when the goal is to drive maximum performance and deliver truly personalized experiences at scale, brands turn to solutions built by personalization specialists—like Email Pulse.

Email Pulse is a natural extension of our Recommend and Personalized application, which is used by hundreds of leading Shopify brands to deliver personalized online shopping experiences. What sets Email Pulse apart is its real-time personalization capability—leveraging live shopper clicks as buying signals to dynamically adapt and present the most relevant products in the moment.

This innovation was inspired by our collaboration with Amazon’s machine learning team, where we explored ways to make Amazon’s advanced personalization technology accessible to the broader Shopify ecosystem. That experience gave us a deep appreciation for the power of clickstream modeling, a technique Amazon pioneered over years of optimizing its own marketplace.

Email Pulse was born out of our ambition to bring this level of intelligence beyond the storefront and into the inbox. To minimize complexity and reduce workload for email marketers, we packaged all of this advanced personalization logic into a single, easy-to-use component: InMail Shop.

In Klaviyo, it is technically possible to use the “Products a customer may also like” product feed in batch campaigns, but whether it’s advisable or effective depends on the data richness of your audience.

What’s Possible:

• You can insert the “Products a customer may also like” block into batch campaigns using dynamic content blocks.
• Klaviyo will attempt to generate recommendations based on available behavioral data.

Key Limitations:

• Sparse Data: For contacts with no meaningful interaction history, the feed defaults to a generic fallback product set.
• No Event Context: In batch campaigns, there’s no real-time event trigger to refine the recommendation context.
• Generic Output: You’re likely to get diluted or irrelevant results for a significant portion of your list.

When It’s Advisable:

• You’re targeting a segment with rich behavioral data (e.g., highly engaged shoppers, recent purchasers).
• You include fallback logic (like showing bestsellers) for contacts with insufficient data.

Email Pulse’s InMail Shop is a powerful email personalization component that brings AI-generated, highly personalized product assortments directly into shoppers’ inboxes:

• Leverages all available shopper data for comprehensive personalization.
• Accurately identifies the shopper’s lifecycle stage, ensuring relevance to their current buying intent.
• Detects the strongest behavioral signals indicating product interest.
• Intelligently determines the most relevant personalized product streams based on detailed data analysis.
• Automatically generates and delivers product streams in real-time.

For marketers, using InMail Shop is effortless—simply drag and drop the universal component into an email template, and the entire personalization process is seamlessly automated.

Klaviyo’s email file sizes are unnecessarily increased due to inefficiently structured product feed components. These feeds include excessive inline styles, redundant HTML, unnecessary whitespace, and repetitive elements—all of which contribute to substantial code bloat. The resulting bulky HTML increases the risk of email clipping, especially in Gmail.

Email Pulse is designed with performance and deliverability in mind. Unlike conventional product feed blocks that often generate bulky HTML filled with inline styles and repetitive code, Email Pulse uses a lean, modular structure to keep email file sizes minimal. As a result, marketers can confidently include engaging, personalized shopping experiences without sacrificing inbox placement or user experience.

The examples below compare a typical Klaviyo product feed with an InMail Shop content block, demonstrating how Email Pulse can deliver significantly more content without a significant increase in email payload.

Klaviyo Product Feed — Payload: 20kb

InMail Shop — Payload: 30kb

Klaviyo provides basic email performance metrics—such as opens, clicks, and attributed revenue—but stops short of offering the deeper insights marketers need:

• There’s no visibility into how shoppers behave after clicking through to the online store.
• No way to measure the effectiveness of individual product recommendations.
• No reporting tied to lifecycle stages or shopper intent.

Without these insights, marketers are left with surface-level data and little ability to connect email activity to real outcomes.

Email Pulse was built to solve a critical blind spot in ecommerce marketing—the disconnect between email engagement and what happens next in the online store. While most email platforms stop tracking at the click, Email Pulse goes further by combining email metrics with onsite behavior, product interactions, and lifecycle data. This unified view gives marketers real insights into how shoppers respond to email-driven product recommendations, what drives conversions, and where opportunities are being missed.

Uplift modeling is a technique used to measure the true incremental impact of a campaign by comparing the behavior of those exposed to it with a similar group that wasn’t. Instead of simply tracking clicks or attributed revenue, it isolates what changed because of the campaign.

At Email Pulse, we apply uplift modeling to both entire email campaigns and the personalized product recommendations within them—measuring how each influences shopper behavior beyond the inbox.

Learn More: Email Uplift Modeling as a Game Changer