Top Frequently Asked Questions
Website and eCommerce Search tools that exist today tend to solve for relevancy by optimizing indexing and/or results ranking. Current tools also focus on the content and keywords, trying their best to match a result with what they think the user intended. However, they do not factor in user behavior. SearchBoost.ai tackles this problem by focusing on the users and their experience - after all it is them who will consume search results. We observe your anonymous and known users, their behaviors and patterns. Then we train that data using a number of machine learning models and techniques to ultimately serve your users with a predictive, relevant and personalized search results that cater to the highest clicks and conversions.
SearchBoost.ai imports your user behavior from your data source using our out of the box connectors with leading data tools like Adobe Analytics or Google Analytics. Then pulls in search results from your search tool with connectors for leading search engines like Apache Solr or Elasticsearch as well as connectors to some of the top experience platforms like Adobe, SAP and Acquia Drupal. Using the user data and search results, SearchBoost.ai trains a machine learning model for better results. After the new smart results are produced, SearchBoost.ai pushes those results back to your search engine. Once your search engine has the new results, it would serve your users on your site with them. Also you will have a number of reports and dashboards showing you insights on the performance of SearchBoost.ai, including in your data analytics and in our own dashboard.
Yes, SearchBoost.ai doesnʼt replace your current search tool - it just makes it better.
|Field||Field Description||Must have|
|Search keyword||This represents the keyword that was used for the search||Yes|
|DocumentID||Unique identifier for each document||Yes|
|SessionID||This is the id from each user session||Yes|
|Next Page||The next page after users visit||Yes|
|UserID||Unique user Id||No|
|Page Number||Page number in the sequence||No|
|Page URL||Page URL||No|
|TimeStamp||If Next page is not present timestamp is must have||Depends on Next Page|
|Search result Click||This is a must to have.||Yes|
|Add to Cart||The user clicked to Add to cart after being driven from Search||No|
|Form fill||The user performed a form fill after being driven from Search||No|
|Buy button||User converted after being driven from Search||No|
|Subscribe||User Subscribed after being driven from Search||No|
SearchBoost.ai works on any searchable content and products that have been parsed by your existing search tool.
Check out some of our trusted clients list in the homepage.
- SearchBoost.ai uses AutoML for its relevancy models which uses 15 ML algorithms with 14 preprocessing methods yielding a total of 110 hyperparameters. This approach ensures the most optimal set of results based on the highest confidence levels.
- Markov Chain attribution model is also utilized for optimal scoring for conversion attribution to provide the proper set of search results that yield that highest potential of conversion!
- See our AI/ML Page to learn more.
SearchBoost.ai doesnʼt affect how your application consumes or displays search results, we only ensure that your search engine has the most optimal and relevant results ranking. If your current application leverages faceted search, so will SearchBoost.ai.
SearchBoost.ai provides relevant search results for all users and all also provides an additional level of personalization for known users.
SearchBoost.ai uses the following security measures:
- Security Groups with proper Inbound and Outbound rules
- Load Balancer for exposing public DNS to the private network
- Web Application Firewall (WAF) for prevention from DDoS attacks
- and other general web vulnerabilities
- Data Resilience by creating backup copies of data
- Separate infrastructure for each client
- SSL encryption in-flight between server and client
- Data encryption and allowing access to only authorized requests
- Client-side and server-side validations
- Filter HTML and request parameters to secure from XSS attacks
- Setup CORS to prevent unauthorized usage of the site and APIs
- Use CSRF tokens to avoid spamming from Bots
Since SearchBoost.ai doesnʼt interact directly with your application, just your analytics and search engine, it adds minimal potential exposure to your infrastructure.
SearchBoost.ai provides with dashboard templates for Adobe or Google Analytics.
SearchBoost.ai also offers search performance dashboard within the tool dashboard.
SearchBoost.ai is very simple to signup for and configure. The majority of the time is spent ensuring your analytics tool is capturing the right data regarding your user behavior that will produce the most relevant results for your future users.
Because SearchBoost.ai uses machine learning and relies heavily on the incoming data from your data tool, there is no exact timeline that can be estimated. However, most implementations of SearchBoost.ai see a difference in search clicks and search exits (relevancy indicators) within the first 3 months.
Since SearchBoost.ai doesnʼt interact directly with your application, just your analytics and search engine, it doesnʼt affect your application availability at all.
Since SearchBoost.ai doesnʼt interact directly with your application, just your analytics and search engine, it doesnʼt affect your application availability at all. The following are SearchBoost.ai dashboard SLAs only.
|Support Window||Product Support|
|Critical / P1 on Prod Environment||< 2 hours||8 hours||24 Hours|
|High / P2 on Prod Environment||< 4 hours||12 hours||48 Hours|
|Medium / P3 on Prod Environment||< 8 hours||16 hours||1 Week|
|Low / P4 on Prod Environment||< 16 hours||24 hours||2 Weeks|
Please contact us so we can help you cancel your account.
Your search engine has an on/off switch to stop incoming search results from SearchBoost.ai.