Utilizing the potential of query clusters has led to a significant surge in search-redirect traffic, while simultaneously streamlining the management of these redirects. This success has convinced the majority of our clients to adopt SearchHub as their primary system for creating and managing search-redirects.

However, with this increased usage, certain challenges have surfaced in effectively managing and analyzing the search-redirects. Additionally, users have expressed their desire for specific features, as outlined in the following requests:

In response, we have decided to undertake a complete redesign of the Redirects view.

This redesign aims to address the emerging challenges and accommodate the requested functionalities progressively.

searchHub uses data to make every onsite search technology on the planet better. Why should optimizing the structure of our user interface be any different?

Our announcement today, is both the product of in-person customer interviews, and anonymous usage data. The result is a user-experience design focused on the task at hand.

So, what's changing?

The right-hand corner is now populated with:

Data-driven by design, our left-hand navigation now groups action items for an intuitive workflow:

The primary objective of SearchHub is to facilitate incremental enhancements to search systems for search teams on a large scale over time. With the introduction of the Query-Testing Module, we aim to elevate the process of determining relevance from mere guesswork to conducting detailed search experiments on a large scale.

This module enables the evaluation of modifications to your search system, such as integrating synonyms, adjusting field boosts, incorporating vector search, or making product assortment modifications, using real data rather than solely relying on intuition.

In its current form, the Query-Testing capability automatically identifies clusters of queries that should be compared against each other for testing purposes. These clusters are selected based on the need to exhibit sufficient differences in user engagement and to have a significant number of searches to ensure statistically meaningful experiments. These selected queries are then isolated from any mapping adjustments that SearchHub might otherwise apply.

Tests are automatically concluded once a statistically significant and sufficiently powered outcome is detected, or if the test duration exceeds 28 days.

Upon completion of a standard test, one of the following outcomes will be observed:

After running this capability with our BETA customers for a few months, we've received overwhelmingly positive feedback. We're thrilled with the results and excited to announce that it's now accessible to all our customers using our searchCollector.

Search managers occasionally need to review and enhance search results with minimal result sizes.

It's essential to acknowledge that even the most advanced Search Engines can face challenges in consistently retrieving all relevant results for any given query.

In terms of improvement, even highly specific queries yielding relevant results may benefit from additional related results. For instance, when searching for a specific product, the search engine might successfully locate the exact product and a couple of associated items.

To simplify this process for you, we have introduced an additional Quick-Filter called "." This feature is designed to make your lives a little easier when working with searches that yield limited outcomes.

In response to user feedback expressing a desire for more comprehensive insights into the performance of long-tail search queries and their key performance indicators (KPIs), we have incorporated a valuable addition to our SearchInsights Top Queries view. A minor yet impactful feature has been introduced, allowing users to directly filter queries displayed in the tables based on the minimum number of words or tokens.

In the accompanying screenshot, you can observe that by adjusting the "#words" value to 2, the queries in the tables were instantly refined to include only those with a minimum of 2 words or tokens.

We trust that this enhancement will continue to enhance your productivity as you optimize your search engine.

As an increasing number of users depend on searchHub for their search redirects, we've gathered feedback suggesting the convenience of exporting these redirects in a structured format.

Moving forward, you can effortlessly export all your redirections as a ".csv" file by clicking the newly introduced "Export" button in the QueryRedirects view.

We received feedback suggesting the need to export Underperforming Key Performance Indicators (KPIs) over various time intervals for evaluation purposes.

– we've just rolled out a new feature! You can now export the raw values for the top graph as a CSV file with a simple click on the download button located in the top right corner. We believe this addition will enhance your workflow significantly as you manage and optimize your search processes.

We are thrilled to announce the release of smartQuery's new major version, V2.1, which comes with a range of exciting enhancements and full support for our planned mid-term features.

In smartQuery V2.1, we have completely overhauled our underlying data structures, enabling us to seamlessly integrate additional models / features while significantly reducing memory usage at the same time. This means that smartQuery can now deliver twice as much information with up to a 70% reduction in memory footprint, all while maintaining its exceptional performance.

We've introduced our real-time inference library, preDICT, which empowers smartQuery to optimize user search queries, even if it has never encountered them before. This is achieved through one of the following approaches:

We have also already incorporated all the necessary APIs and data structures to support our upcoming planned features: QueryRecommendations and NeuralInfusedProducts. By doing so, we can seamlessly activate these features on our end without requiring any further updates to smartQuery on your side.

We sincerely hope you'll love these improvements.

Yesterday, we implemented a new core update with the goal of enhancing the precision and efficiency of smartQuery. While we are thrilled to already be receiving billions of queries each month, we recognize that this high volume also includes a considerable amount of low-quality data.

Consequently, our primary focus for improvement has been on developing methods to identify and eliminate such subpar data before it reaches your smartQuery clients. You may see noticeable declines in the "learned search phrases (active)" chart. However, please rest assured that we are actively monitoring these changes to assess their impact on your Key Performance Indicators (KPIs) and will make adjustments as needed to ensure a positive outcome.

We've recently introduced an improved system for managing roles and permissions, offering more precise control over user access and editing privileges. This update allows us to finely tune access rights, encompassing the ability to view, edit, and save permissions for all your colleagues.

To accomplish this, we'll be introducing predefined user roles. For instance, some roles will limit the ability to view SearchInsights, including modifying time slots for Top Queries, while preventing changes to Query Clusters, Transformations, or Management Configurations.

We'd like to hear your preferences regarding these roles, and we'll work to find the most suitable options for all our customers. Please let us know your preferences.