Optimize load speed with machine learning

Load speed-optimization with Intelligent Prefetching

The load speed of your webpage has a great impact on both the user experience and conversions. A single second slower load speed generally results in a 20% lower conversion rate, meaning that a lot can be achieved by improving the load speed on your webpage.

Based on machine learning, we have developed a tool that can improve your load speed without changing the design, content or infrastructure of your page. Intelligent Prefetching uses machine learning to predict what page a user will visit as their next. Based on the prediction, the browser can load the content of that page before the user have clicked on the link. The user will experience a page that opens in an instant, without any waiting time, which is an exceptional user experience.

Steen Rasmussen

Digital Strategist & Senior Partner
+45 70 20 29 19

Continuous improvement of load speed

The predictions are made on user level and are based on anonymous data from the single users current web-session. The calculations are built on analytics-data and by constantly analyzing the precision of the predictions, the system will continuously learn and improve its precision. Through intelligent use of data and machine learning, our Intelligent Prefetching tool will constantly learn from itself and become more intelligent and even better at optimizing the load speed and give your users a better experience on your site.

This is how Intelligent Prefetching work

Intelligent Prefetching work by using machine learning to track the single user’s session and use analytics data to calculate where the user will click next. instead of running the algorithm in the browser, and slow down the site, Cloud-technology is used to execute the calculations. 

Once the calculations have been made in the Cloud, the browser starts to load the content of the predicted page, making it load instantly once the user clicks on the link. Load speed and the precision of the prediction are monitored, for the system to constantly analyze its efficiency and use the information to be able to make even better predictions. The system learns from itself and becomes smarter and smarter, the more users it gets to analyze.

The Intelligent Prefetching engine predicts and prefetches the next page before the user clicks

Case: Significant increase in visitors for VisitDenmark

After the implementation of our Intelligent Prefetching tool, VisitDenmark has experienced a 40% improvement in their load speed and a significant increase in the number of visitors on their site.

  • VisitDenmark has traffic from all over the world, which is a challenge for the users perceived load speed.

  • We implemented a load speed engine, that based on user data estimated the page the single user was most likely to click into, and with Intelligent Prefetching we improved the average load speed by 40% across their whole page. At specific pages, the improvement was close to 75% improved load speed.

  • We gave the visitors on VisitDenmark an improved user experience, which has lead to a significant increase in the number of visitors of their site. The project was done on a global level and across VisitDenmark 11 sites.