Feature Requests

For daily work on existing clusters, it would be helpful to have filter options on top of the lists, e.g. show only all popularity equal/smaller than 5. Also: the infinity scrolling makes it hard to find the right spot where one was working on the other day. sometimes it even has loading issues, meaning one can`t scroll down to popularity zero...

When you open a cluster, you could see, that there is a link for a redirect. But to see if the redirect is active, you have to scroll (depending on how many kpis you see) far to the right. It would help to see the redirect active / not active info on the left side.

Make the power of searchHub usable for optimizing the product data before it's even indexed into the search engine. Normalize essential data fields like title, product type, etc. so all values are streamlined Empower facet cleansing and normalization for text and numeric facets -- Add number/unit conversions, so people can search in inch, cm, mm, whatever they like (MB/GB/TB is another example) Merge multiple source fields into one target field (e.g. if multiple sources for the same facet exist) Let the user decide, which facet is valuable in which section (category) of the assortment Make entity recognition available for descriptions to generate more valuable facets Verify all indexed data against searchHub query clusters for perfect matching

When looking at the new dashboard in searchhub, the "revenue" KPI could be interpreted as the actual revenue created on the site. This is not the case and could lead to misinterpretations. Therefore, please clarify where the "revenue" metric stems from (and, if applicable, the limitations).

It would be great if you can click directly at the AI Training as at the redirects to modify the done changes. At the moment the user has to click first on Edit -> do the change -> save.

We use Queries by Category a lot to identify popular assortments and improve them together with our category managers. The overview helps us see what’s currently in high demand — but we’re missing the option to track changes over time. It would be helpful to compare the selected timeframe with a previous one (similar to what’s already possible in Performance Charts). That way, we could better understand if certain actions — like optimizing the garden furniture category — actually led to improvements. Here’s what we’d like to see: In the tile overview: the number of searches from the comparison timeframe shown in brackets. Simple trend indicators (e.g. arrows up/down) to show whether performance has increased or decreased. An easy way to spot optimization potential — for example, categories with high search volume but low engagement. In the results table: all KPIs shown for both timeframes side by side. This would help us evaluate the impact of our work and spot new opportunities more easily.

The current solution shows the list of queries based on there whole performance. Sessions and Clicks that happened within the customers session. It would be helpful to see the clicks that actually happened in the Category that has been selected for exploration. For e.g. in category vine and spirits currently the top query is airconditioning. From the selected category this is not true. Because there where just a few clicks on the product of this category. This is misleading. The wrong conclusion could be that it would be helpful to place a teaser for vine and spirits in the categorypage. In oposite to that it should sort the queries by real clicks on products.

To enable users: 1) to see the entire list of words (including total number of words available) 2) allow for filtering of the words by the different categories that it is mapped to in the popular categories categories 3) include some KPIs of the words so users know which ones to focus on first if there is limited resources working on it

We request the implementation of a filter option in Searchhub that allows users to generate a list of the most visited clusters while excluding those with an active redirect. This functionality would significantly reduce manual effort, as the current process involves reviewing high-traffic clusters individually to determine whether a redirect is already in place. Automating this step would optimize workflow efficiency and help focus optimization efforts on unresolved clusters only.

We need the ability to flag clusters as "checked but not relevant" (in underperforming queries etc.). There are terms that refer to products we will never carry in our portfolio. We would like to mark these terms in order to exclude them from further review (e.g., via a filter). As of today, the terms remain permanently in the evaluation. And their number keeps increasing.