Understand & Delight Your Search & AI Users
To delight your users, first you need to understand them. We help you analyze your user experience and behaviour to help identify exactly what they are looking for. Qualitative user research, data analysis and visualisation will show you what users search for and whether they find it successfully, allowing you to quickly identify where you can improve search result quality and reveal opportunities for search product development.. Gathering and analyzing click data and human judgements will allow you to calculate metrics for search quality, a useful first step in creating an offline Search Lab for tuning search & AI and a way of measuring how well search is helping your business succeed.
Analyze Search User Experience
Using industry standards and our deep experience of creating and improving search user interfaces, we review your user experience on several levels to identify pain points, optimise user flows and allow you to guide users on their search journey.
We’ve worked on hundreds of websites and search products and we can tell you which interface patterns will work best for your users. We can produce detailed reports and work with your UX teams to improve how users interact with your site or service and help your team verify changes to the search UX in user testing.
Categorize User Queries
The queries your users type can indicate their information needs and what tasks they are attempting – to inform themselves, research a product or look for a new job. We can help you build powerful analysis tools and dashboards so you can constantly monitor what users are searching for – and if they are successful.
For an example see how we helped leading EU parts supplier Rubix build the dashboards shown here, which identifies queries that produce no results, queries that produce only a few results and also the most popular searches.
Measure & Tune Relevance with Quepid
We built Quepid as a search relevance workbench, allowing you to gather human judgements of search result quality and calculate overall metrics. Quepid is used by relevance engineers across the world to assess, tune and improve search. It’s a standard part of our process of improving search quality for our clients.
Quepid works with open source search engines such as Apache Solr, Elasticsearch & OpenSearch, vector engines like Vectara and has APIs allowing it to be connected to potentially any search engine or service. Quepid is open source under the Apache 2 license and available as a free hosted service, so you can use it both inside and outside the firewall.
Human judgements are just one way to evaluate search result quality – we can also help you collect implicit judgements based on clicks and develop automatically generated judgements using AI.
Learn how to focus on your users
We know how to delight them with search & AI!