Elasticsearch, Events

ElasticON 17 – One to Remember!

We were really excited to announce the availability of our Elasticsearch Service offering, and the timing couldn’t have been better – right before our trip to ElasticON 17 in San Francisco!

It was great checking out the Elasticsearch sessions- hanging out with the community and our pals from Hitachi. SearchStax paired with Hitachi Credeon Secure Full-Text Search offers the most secure way to perform search against confidential data.

Contact us to learn more!

Elasticsearch, New Release

Introducing Fully Managed Hosted Elasticsearch Service

We’re really excited to announce the availability of our fully-managed Elasticsearch Service. Our goal was to enable developers to focus on the interesting parts of their jobs: application development and relevance tuning. And leave the boring stuff: maintenance, support, care and feeding of their Elasticsearch infrastructure to us – the leading experts in open source search. 

We’ve seen some of our customers struggle with their hosted Elasticsearch applications, and realized there was a strong need for a fully-managed solution.

We’ve taken the lessons learned from our Solr-as-a-Service solution and applied them to our new service. We want to help companies make amazing Elasticsearch applications faster, easier, and with greater confidence.

Our Elasticsearch Service solution offers automated provisioning and scaling of Elasticsearch, backups, 24/7 monitoring and alerting, comprehensive query-level analytics, and class leading administrative tools.

SearchStax, combined with our Managed Services, powers the leading fully-managed cloud-agnostic Elasticsearch-as-a-Service.

For more information or to schedule a demo, please contact us.

New Feature, Search Analytics

How Fast is Your Search? Latency Statistics from SearchStax Analytics

How quickly does your search application respond to a typical user query? Since every query is different, this question isn’t easy to answer. Latency statistics let us describe the responsiveness seen by a typical user, based on actual queries from the field.

At Measured Search, we are excited to announce the addition of  Average, 95th percentile, and 99th percentile latency to our suite of SearchStax analytics reports. Below are the three new statistics shown in SearchStax Analytic’s App Overview display:

Average, 95th, and 99th Percentile Latency in SearchStax Analytics from Measured Search.

This performance metric is expressed in milliseconds.

  • Average Latency: Average of all queries responded within this time.
  • 95th Percentile Latency: 95% of all queries responded within this time.
  • 99th Percentile Latency: 99% of all queries responded within this time.

In the example above,  the average query was returned within a second, which is encouraging. 95% returned in a second and a half (or so). Then there were a few queries that required multiple seconds of computation. Nearly all queries (99%) returned within 12.6 seconds. Overall, this is a very responsive system. But we might want to take a closer look at the slow queries. SearchStax lets us view the slow queries in the Top Searches display.

For additional information see the App Overview discussion in the SearchStax User’s Guide.  Click the link to learn more about SearchStax, the leading Solr-as-a-Service platform.

New Feature, Search Analytics

SearchStax Unveils Similar Search Query Analytics

We’re excited to announce our latest enhancement to SearchStax Analytics, Similar Search Query Analytics.  Similar Search Query Analytics groups search queries by the item that is ultimately clicked.  This allows your search team to understand the keywords users are typing to find things.  You can then leverage these analytics into actionable insights by enhancing your synonym lists with these similar search queries.

Searchers think uniquely. Ten people searching for the same product may type in ten different queries. They won’t use all (or any) of the keywords in the product title. They may misspell or mistype unfamiliar labels. Finally, a significant number of users experiment with keywords looking for the magic phrase that turns their vague desires into concrete product recommendations. As your organization’s search expert, your job is to help your users find what they are looking for faster and easier.

Similar Search Query Table

SearchStax’s Similar Search Query Analytics gathers all of the queries that led to a particular product and ranks them by popularity. This shows us what the user typed in before finding that specific item. We turn these search analytics into actionable insights by either enhancing the synonyms associated with this item or we can even modify the product descriptions to make them resemble the language most often used by our customers in their queries. Either of these would results in faster, more successful search sessions that lead the user to the items she’s looking for, even if she didn’t quite ask for it the way you might have originally envisioned.

For more information, see Similar Searches in the SearchStax User Guide.  Contact us to get a demo of our Analytics Suite or to learn more about SearchStax, the leading Solr-as-a-Service platform.

New Feature, Search Analytics

Searches per Session – SearchStax Analytics

Introducing Searches per Session – the latest search metric from SearchStax Analytics. How many queries did your user submit before she either succeeded (and clicked through) or she gave up and ended the session? This statistic reveals a lot about your user’s search efficiency. Does she need multiple searches to find the thing she’s looking for OR is she able to find what she’s looking for on her first try.  For transaction and intent heavy properties, a Searches per Session closer to one is typically better than a higher number.  For browse or navigational search heavy properties, a higher searches per session number is appropriate. In either case, keeping a close eye on Searches per Session and monitoring for large swings in either direction provides actionable insight for your search team.

Searches Per Session from SearchStax Analytics

SearchStax Search Analytics reports Searches per Session in two formats. Configure a time window and then view the Analytics App Overview screen. In the upper right is a single figure reporting the mean queries/session over the requested time period. The points on the graph indicate searches/session for individuals days (or hours, if appropriate) during the requested interval.

For additional information see the App Overview discussion in the SearchStax User’s Guide.  Click the link to learn more about SearchStax, the leading Solr-as-a-Service platform.

New Feature, Search Analytics

Average Click Position – Analytics Metric from SearchStax

We’re excited to announce our latest enhancement to SearchStax Analytics, the Average Click Position metric.  Average Click Position can be a very strong indicator of the effectiveness of your Solr search relevance. It is a measure of where your users are clicking on a list of search results. For example, if the user clicks on the second result, two will be recorded.  If the user clicks on the sixth result, six will be recorded. A low Average Click Position (one to three) indicates that your users are generally finding what they are looking for within the first few results. Alternatively, a high Average Click Position (ten plus) indicates that users are generally having to go to your second page (assuming ten results per page) to find what they are looking for.  That as you know is not a good thing.  This metric gives you insight into your search relevance and over time how your relevance is trending. It is a powerful metric that can help guide search relevance improvements.

To learn more or for a demo of SearchStax Analytics, contact us.