How norm Measures Success: The Key Metrics We Track, and How they Can Help your Business

January 15, 2025

Adrentech’s norm - our AI knowledge base platform - ability to provide AI generated answers to user questions based on knowledge articles provided by your business is one of its key features. You can read more about this feature, and norm here. Measuring customer perceived success of these AI generated answers, along with other data collected through customer use of the platform, is key to your businesss' ability to harness norm at its full potential. To ensure this, we track this data alongside a wide range of additional metrics that give us – and your business by extension – valuable insights into norm’s performance. These metrics are particularly important as they can allow your business to identify areas for improvement within the platform's knowledge base, judge customer satisfaction, and ensure customers are getting the best answers quickly. In today's article, we will outline two categories of these metrics - top articles and generative search data - key to measuring success.

Top articles searched are the first of the categories of data collected by norm. The collection of these most frequently customer accessed articles is particularly helpful as this information can provide insights into what topic areas users need the most help in. To track this, we collect metrics within this such as session count, which tracks the number of unique visits an article has received over a given period. The higher this session count, the higher the likelihood that the content within an article is particularly relevant to users. Customers are also capable of leaving helpfulness ratings on individual articles. This data, paired with session count, can grant your business the ability to drill down to knowledge gaps, as well as know when and where to eliminate articles that are no longer useful. Hit count is another of the metrics we collect. Hit count measures the total number of times an article has been accessed, including repeat visits. A high hit count for an article typically indicates that a user could be having difficulty locating what they need on their first try, or that the content within the article is particularly helpful enough to warrant repeat views. In conclusion, tracking the two metrics of session count and hit count can help identify popular topics, as well as help to identify gaps in content that may need to be addressed.

Generative search data is another of the categories of information collected. This data is particularly important as it grants your business knowledge of the customer’s perceived success of their searches, among other things. One of the metrics collected to track this is the success metric. This data point indicates whether a customer search has been successful in fetching at least one relevant article for an answer. The success metric is particularly useful in that it can over time indicate article gaps in the knowledge base, in addition to letting your business know where an abundance of information is available. Another of the useful metrics collected is the numeric score. This metric allows customers to rank the success of generative responses, from one star to five. This metric allows your business to collect quantifiable feedback from customers that can signal areas that need attention, as well as ensure future knowledge base expansions are based on user preferences. These metrics, when integrated, can allow for more targeted improvements in content quality as well as help to ensure customers' real-time needs are always being addressed.

Curious to learn more about the metrics collected by Adrentech's norm platform, and how their collection can help your business to better meet customer's needs? Schedule a demo here.