How Nova works
We measure real buyers, in a live AI conversation, and record both sides
Nova follows real people before, during and after an AI-assisted buyer research conversation, from the brands or products they considered beforehand, to the ones the walked away with.
We measure both B2B and B2C purchases.
A study has three stages:
Before. Buyers tell us about their situation, their buying context, their needs and the brands they already have in mind. This can be tailored to the client’s needs, and allows us to segment the data. The AI is blind to this information.
During. They engage with the AI assistant, in their own words, to explore their needs. The AI has live we search, and reacts to the buyer’s needs to match them to suitable brands and products.
After. Many buyers end with a final pick, some don’t, everyone gives us an updated view of their shortlist. We capture how the conversation impacted their decision at that point, and what they intend to next in their buying journey. This can be tailored to client needs.
Because the AI is blind to which brands a buyer brought with them, we can separate what the buyer brought from what the AI brought.
There’s two sides to an AI conversation, we record both.
What the AI says
- Every claim about each brand
- The topic, tone and firmness of each claim
- Whether each claim cited a source
- Every search it ran, and every result
- The links it placed in its replies
What the buyer does
- Who raised each brand first
- Whether they followed it up
- Whether they turned it down, and why
- Which brands they let pass
- Their final pick, and their stated next step
Data handling
Personal details are removed before anything is stored. Participants can ask for their data to be deleted at any time.
Let's talk.
Tell us what you need to know. We'll go find out.