What factors/variables contribute to score/rank?

  • Created

All of these measures relate to groups of features that get fed into the AI model to determine which accounts should be qualified and with what ranks. Depending on the feature, there are different mathematical and data science tools that we apply to them, but in most cases we’re doing some comparison for each candidate account to see how similar it is to the customers in your profile. These range from very simple things (how many followers does this account have on FB, how many followers do each of your customers have?) to very complicated things (for all your profile intent keywords, build multi-dimensional vectors representing each accounts intent throughout the internet on those keywords, and then do some linear algebra to compare these N-dimensional vectors).

Once we have calculated the features for each candidate account (and your profile customers) we run a machine learning model to classify and score each candidate account so that for each candidate account, we can expose which features were the most important in scoring and ranking them.   

Was this article helpful?

0 out of 6 found this helpful