A conversation is one question to one answer engine, with the answer that the answer engine gave. Genezio calculates each number from conversations. Thus, at this level you find why a score has its value.

​
What a result holds

The question. The scenario, as Genezio really asked it. The answer. The full response, as the answer engine wrote it. The detections. Genezio records these items:
  • If your brand appeared
  • If the answer engine recommended your brand
  • The competitors that the answer named
  • The sources that the answer cited

​
How to read a result correctly

Read the answer before the detections. The classification tells you what occurred. The text tells you why, and the reason is usually clear when you read it. Ask these three questions about each answer:
  • Where did this answer come from? The citations are the material that made the answer. If the answer engine recommends a competitor because of one review, work on that review.
  • How does the answer describe my brand? If the answer names you as the expensive option, it is a mention and a problem at the same time.
  • Is this answer sufficient for a buyer? If yes and you are not in it, you have a real gap. If the answer is poor for all brands, the category is open.

​
Why two runs of the same question are different

Answer engines are not deterministic. The same scenario gives different words, and sometimes different brands, from one run to the next run. Thus, one conversation is evidence, but it is not proof. Read some conversations before you make a conclusion. Use trends as the signal, not single values.

​
When a result looks incorrect

First, examine the scenario. Many unexpected results come from a scenario that does not ask what you think it asks. The scenario can be too broad or too narrow. Or it can use words that no buyer uses.