Less managing, more selling: how AI is changing product bundling
Bundles have always been a reliable tactic in ecommerce. They help guide shoppers toward complementary products, whether it’s completing an outfit, building a routine, or adding relevant extras; they naturally increase average order value.
Most teams know bundles work. Few have the time to do it well. Creating the right combinations, keeping them live as inventory shifts, and spotting missed opportunities quickly become difficult to manage at scale.
Agentic AI changes that. It reshapes how bundles are created, maintained, and optimized over time, making it easier to unlock their full potential.
Why bundling is still difficult to scale
Bundles are simple in concept, but more complex in practice.
Creation is slow. Finding the right product combinations still relies on catalog browsing, intuition, or partial data. There’s rarely a clear signal explaining why a specific pairing makes sense, or how confident you should be in it.
Bundles are fragile. When a product goes out of stock, bundles either break or disappear. Fixing this requires someone to notice the problem, investigate it, and manually resolve it. That process doesn’t always happen before the bundle has already cost you revenue.
Opportunities are invisible. Your best-performing products aren’t always bundled. Spotting which ones should be requires time and analysis that most teams don’t have bandwidth for. Opportunities exist but often go untapped.
Together, these challenges create an operational overhead that limits the impact bundling could have.
What AI Bundles change
AI Bundles extend our Dynamic Bundles solution by introducing AI across two distinct areas: the bundle creation flow and an ongoing opportunities engine that proactively surfaces suggestions. Powered by Huginn, Nosto’s AI agent, they’re part of our agentic capabilities, continuously monitoring performance and suggesting actions so your team can focus on impact instead of investigation or manual execution.
Building bundles faster with data-backed suggestions
Creating bundles becomes significantly more straightforward.
As soon as you add a product, AI suggests complementary items based on real co-purchase behavior. Each suggestion is supported by a clear signal showing how often products have been bought together, making it easier to understand why a product is recommended.
This removes the need to manually explore the catalog or rely on assumptions. You can build bundles faster, with more confidence in the combinations you choose.
Keeping bundles live when inventory changes
Bundles often lose effectiveness when products go out of stock.
AI Bundles address this directly by introducing fallback suggestions during the creation flow. For each product in a bundle, you can define alternative items that will automatically replace it if needed.
This ensures bundles remain complete and continue to perform, without requiring manual updates. Instead of reacting to issues after they occur, bundles are designed to handle them upfront.
Surfacing opportunities that would otherwise be missed
Most bundling tools focus on creation. AI Bundles go further.
Within the AI Opportunities section, the system highlights bundles that need attention, such as those impacted by out-of-stock products, and suggests replacement strategies based on either performance or similarity.
At the same time, it identifies high-performing products that are not yet bundled and recommends relevant combinations to unlock additional value.
What previously required time-consuming analysis becomes a continuous flow of actionable suggestions, directly available to you.
Bundling before and after AI
| Traditional bundling tools | With AI Bundles | |
| Building a new bundle from scratch | – Browsing the catalog to find the right combos – Testing option – Relying on intuition – No clear signal on whether a pairing makes sense | – AI automatically suggests complementary items – Products chosen based on real co-purchase behavior – Eliminates guesswork |
| A product in an existing bundle goes out of stock | – Bundle shows as incomplete on the storefront, or disappears entirely. – Revenue lost before anyone realises – Requires a person to spot the error – Needs to be manually fixed | – AI agent flags OOS immediately – 2 replacement options given: a cross-sell option to preserve or improve performance, and a similarity option to preserve the bundle’s original intent – Revenue is protected |
| A top-performing product has no active bundle | – Spotting it requires manually cross-referencing performance data with the bundle catalog – Opportunity goes unnoticed | – Huginn continuously monitors top-performing products – Bundle opportunities are automatically surfaced – Opportunity becomes revenue |
Get started with AI Bundles
AI Bundles are available in the latest version of Dynamic Bundles, accessible from the bundle creation flow and the AI Opportunities section. Learn more in our help article.
Existing customers can reach out to their CSM or book a demo here.
New to Nosto? Book a demo here to learn more
