Examples of how AI can be used on a website
Examples of how rek.ai with the help of anonymized user behavior and artificial intelligence improves websites' user experience, usefulness and contributes to reduced costs. All examples are based on a standard version of rek.ai, but with different types of presentation for the visitor.
1. Swedish Public Employment Service
Swedish Public Employment Service wanted to solve the problem that jobseekers who use Platsbanken received too few job proposals. They used rek.ai to build an AI model that learned which occupational categories are related, in order to provide suggestions for additions to the search, which increases the possibility of finding a job. The AI model uses all user interaction to train a model with knowledge of behavior in the “search for new work” field.
2. Halmstad municipality
Halmstad Municipality enhances the user experience and helps its visitors to find the right information on their website more quickly by recommending pages and e-service to the visitor. The recommendations take into account, among other things, which topics the visitor has previously read about on the website.
The utilization of keyword suggestions generated with the help of rek.ai also increases the personification of the user experience and leads the visitor to the correct search result faster.
3. Länsförsäkringars Fastighetsförmedling
Länsförsäkringars Fastighetsförmedling uses rek.ai to recommend properties. Through a “trained” AI model, you automatically lift the real estate objects that will appeal to the visitor.
The factors that affect the recommendation include position, price level and type of real estate object. The model also understands if an object becomes “viral” and can then lift the object, which leads to ** raising the brand **.
4. Experience Växjö
Växjö helps visitors to discover events and places within the municipality by using rek.ai on their tourist website. The model learns what visitors demand by training a new AI model every night that takes into account factors such as position and previously shown interests.
5. Swedish Public Employment Service: Intranet
Swedish Public Employment Service’s intranet is a complex website with a large amount of content. To elevate information to the employees, AI is used, which is trained in user behavior and therefore can prioritize which administrator support, news and content pages are most likely that each employee will benefit from in a given situation. In order for the recommendations to be as high quality as possible, the employee’s professional role and previous interests are taken into account when the recommendation is made.
6. Luleå municipality
Luleå municipality uses rek.ai to highlight on both the start page and subpages and services that the visitor is most likely to request depending on geographical position, device, visitor pattern etc. The possibility of making these sorted links available already in a first step increases the usability of the website on more than one level.
The result is that fewer people contact customer service or cancel their case.
7. Recommend more reading in Skellefteå municipality
By recommending further reading after a news item, Skellefteå helps its visitors to discover new material and inspire them to read more. The recommended news is selected by rek.ai based on all the parameters the AI model takes into account.
The result is that all the material that the editors produce reaches the visitors of the website
8. Integrate external website
Upplands-Bro integrates an external e-service portal on its website. This results in a dedicated platform for building e-services, but leads to the fact that it can be more difficult for the visitor to find the services. By adding the external e-service portal in rek.ai, the services can be displayed both among the keyword suggestions and on their search results page.
An external website is included in all rek.ai projects at no extra cost.
On the content management system SiteVision’s help web, rek.ai is used to predict which questions the visitor is interested in. The function is based on data on user behavior and how visitors navigate the website.
10. Lärarnas Riksförbund
Lärarnas Riksförbund wants to inspire and help its members and visitors to find inspiring and educational material available on the site. By using rek.ai, they personify the pages and highlight articles, podcasts and movie clips that previous visitors with the same “profile” have shown interest in.
11. Swedish Public Employment Service: Digitalportalen
An internal portal for Swedish Public Employment Service’s employees uses rek.ai to recommend articles and courses for employees.
The courses are in an external system, but can still be part of the AI model that learns which content suits different roles within the organization.
The recommendations increase the use of available content and help employees discover new information.
12. Falkenberg Destination
Destination Falkenberg wants visitors to discover all the interesting things that are happening in the municipality. Therefore rek.ai is used to recommend events and pages for the visitor. The recommendations are created based on a model that is trained on how previous visitors have shown interest in the various events in relation to geographical position, number of visits and language in the browser. By using AI on the site increased the display of events by 10%.
13. Falkenberg Intranet
The intranet for employees in Falkenberg municipality is built on the platform SiteVision with Social Collaboration. The intranet uses rek.ai to show links and groups that may interest each employee based on which department he or she works in, and based on previously shown interests. The result is that employees can more easily find information that helps them in their professional role.
14. Falkenberg municipality
By recommending pages and e-services on its website, Falkenberg Municipality gives its visitors a personalized experience.
The website recommends subpages on the various sub-websites and e-services on the page that collects all e-services. The AI model for the recommendations takes into account, among other things, the location of the visitor, and whether he or she uses a mobile or desktop computer (and a further more than 80 other factors).
The result from an A / B test on the start page showed that the implementation of rek.ai led to a reduction in bounce rate by 13%.
15. Götene municipality
Götene municipality has used rek.ai on its home page to create the navigation that the visitor encounters. The result is that the visitor more quickly finds the links that are of interest.
On the subpages of the website, related pages have also been created with the help of AI. In this way, visitors who have not found the right information can get suggestions on pages whose topic is relevant to the page visited.
Götene also uses ** keyword suggestions ** developed by rek.ai.
Götene also uses rek. ai to recommend news after the visitor has read a news item. In this way, they increase the number of news items discovered by visitors.
16. The Swedish Migration Board
The Swedish Migration Board has a large amount of documentation on its intranet, which the employees use in their daily work. By using rek.ai, the intranet can show the documentation that the employee is most likely to seek, depending on, for example, which department the employee works in, and which pages have previously been visited.
The result is that the employee more easily finds the right information and avoids asking colleagues and support.
17. Järfälla municipality
Järfälla municipality uses rek.ai to sort the navigation of subpages, which gives the visitor a more easily navigated website. The sorting takes place automatically and generates a menu structure that dynamically adapts to all factors (more than 80) that the AI model has included in the training. Editors therefore do not have to spend time prioritizing among links, but can focus on producing content.
18. Stockholms Studentbostäder
Stockholms Studentbostäder is a website with visitors from all over the world. By using rek.ai on the personalized start page, the pages that are most likely to be visited by the visitor are recommended depending on their position, browser language, etc.
The search box also uses keyword suggestions generated using an AI model trained on data from the website.
19. Upplands-Bro municipality
Upplands-Bro municipality uses rek.ai to show recommendations for the subpages that the visitor is most likely to request.
Together with using AI trained on visitor patterns in their keyword suggestions, they increase the usability for the municipality’s visitors. By using AI-generated keyword suggestions, the number of completed searches increased by 152%.
20. Swedish Public Employment Service: Activities
For jobseekers who visit the Swedish Public Employment Service’s website have information on activities they can take part in, both physical and online. When the jobseeker reads about the activities, the Swedish Public Employment Service wants to recommend additional activities and therefore uses rek.ai to develop suitable activities based on subject and geographical position.
By using AI, the display of activities has increased by over 14%.