LupaSearch 1.23 is Live: Rules Engine and Personalization testing panel
A new version of LupaSearch is here and there are a couple of new features that we are very excited to present.
Previously the search query dashboard only allowed boosting specific brands, values or applying other changes globally. It means that the same configuration was active for all users, no matter what they were searching for.
The updated rules engine now allows having different query configurations depending on individual user’s behaviour.
Rule condition is used to define when a rule is applied: it can be an exact or a partial search query text match, specific day, date range or a weekday.
Here are some examples of what can be achieved when using the rules engine:
Boosting specific brands in search results only on weekends;
Only apply brand boosting, if user query starts with specific keywords, i.e., water or mineral;
Use numeric boost for products by discount size only on the day of the Black Friday;
Show only the most popular products if user enters an empty query;
Automatically sort results from the cheapest to the most expensive, if user query includes word cheapest or inexpensive;
Include additional facets that allow filtering by screen size or resolution, if the user query includes phrases like tv, monitor or television.
With rule actions it is possible to override any property in the original configuration, if the condition is matched: including boosting, adding additional facets or filters, enabling did you mean, similar queries or personalization functionality.
Each search query can have multiple rules that are applied in a specified order:
Important: Rules engine is still in BETA functionality, so make sure to thoroughly test your configurations before using it in production.
If you want to learn more about rules engine, you can checkout our documentation: https://console.lupasearch.com/docs/concepts/rules
Personalization testing panel
LupaSearch already had personalised search results. However, this feature might have been difficult to test, since each user could have gotten different results for the same query, depending on their engagement history.
For this reason, we are introducing a “Boosted ranking comparison” panel, with integrated personalization testing functionality, which allows you to visually compare the differences in product ranking, when one or more boosting features (value, numeric, phrase boost, rules) are enabled.
To test personalization, the dashboard generates a random user id, allowing you to simulate user engagement actions (like clicking on a product or adding it to the cart) and see how the personalised ranking changes in real time.
Boosted ranking comparison can be accessed from the search query configuration screen.
Important: For personalization to work, an Event API, that reports user interactions with the search results, must be implemented in your LupaSearch integration.
Filter for field selection elements
All of the dropdown select fields in LupaSearch dashboard (environment, mapping field, query configuration selectors, and so on) are now searchable, which should save you a lot of time when configuring various settings, especially if your environment contains a high amount of search indices and mapping fields.
LupaSearch is a continuously growing product. If you encounter any issues or have any suggestions on improving it, feel free to contact our team at email@example.com. This will help us further improve our service.