My data consists of individual messages, tagged with the userID of the user who sent them. I want to count the number of users who say "burger" and "fries", but not necessarily in the same message. In the example UserID Message 1 "I'd like to order a burger" 2 "The weather is nice" 1 "I'd also like some fries" 2 "I'd only like a burger"   User 1 should be counted by user 2 shouldn't. I believe a way to do this would be inner joining by the userID on two separate searches       index=idx_chatbot_gb_p component=chatbot-ms | spath "userID" | spath input=payload output=Message path=messages.message{}.plaintext | search (Message=* burger *) | join type=inner userID [ seach (Message=* fries *) ]         I get zero results when I try this, even though I get results on the individual searches and many users order burgers and fries. Does anyone know a better way to do this or can spot what I've done wrong? Thanks ... View more