Nosto’s unique approach brings significant benefits over traditional methods of back-end integration

Ease and speed of implementation

Nosto can be added to any online store, regardless of size or platform without it becoming a complicated, costly IT project.

Real-time relevance

Unlike other solutions which often have to wait for a product refresh to get the most up-to-date data on your store, Nosto’s algorithms are updated in real-time with every page load.


There is no risk of slowing down site speed with Nosto – actually often it’s the opposite: the product pages load quicker because they are now loading just a single product, whereas previously it had to lookup multiple products.

Less technical risk

Nosto integrates through the UI layer, therefore it touches less of the core code base.


All the above combine to reduce the cost of implementation and ongoing management significantly.

Patent 1

Relates to the way we gather the product data from the retailer’s store (eg product price, description, availability, etc.).

Most personalization solutions do this via a product feed. The problem with this method is that it requires a back-end integration which can be complicated, costly and time-consuming to set up. In addition, the updates are typically scheduled at intervals (eg every 8 hours) meaning that the data is not always up to date, leaving a risk that the retailer could be missing out on more relevant recommendations or even recommending out of stock products.

Nosto’s innovation is that we circumvent the need for a product feed by using the visitors of the website to provide the product information. By adding just a few snippets of HTML and javascript code to the store, we set up a real-time data exchange between the retailer’s store and our cloud. Every time a user then loads a page on the store, we use that as a trigger to capture the latest product information, together with the user’s behavioral data on how they have interacted with the store. In this way, Nosto has a real-time replica of the retailer’s product catalogue at all times as well as up-to-date profiles on each user (who they are, where they come from, what they viewed, added to cart and bought, etc.).

Patent 2

Extends Patent 1 to add: Sharing the product catalogue information with other service providers (e.g. ad networks), and syncing the user identifier across to other service providers. This enables personalization across other service providers by using the personalization and catalogue data gathered in the original patent.

Although we work on unique recommendation algorithm approaches every day, we chose not to start patenting those. Instead of focusing our patents on the recommendation algorithms, we took a different path, one similar to the Wright brothers when learning to fly. We focused on the integration and delivery required for recommendation algorithms to work, creating a system that is low-cost and fast for retailers to implement.

It’s said that the reason Wright brothers were the first to fly is because they developed a way to iterate fast, enabling quick trial and error. Developing recommendations algorithms is similarly a very iterative process. By focusing on delivery and integration in the first place, we now have thousands of merchants using Nosto, enabling us to continuously develop our recommendations faster and faster with shorter iteration cycles. It is this innovation that has made it possible for Nosto to become the fastest growing personalisation engine in the world.