Ranking questions are broken. Ask someone to order 15 items from most to least important and you will get data shaped by fatigue, not by preference. Respondents satisfice. They anchor on whatever is at the top. The result: noisy data that looks precise but is not.

Researchers have known this for years. That is why MaxDiff was invented: instead of ranking everything at once, show participants a small subset (3 to 5 items), ask them to pick their most and least preferred, and repeat across multiple rounds. The math behind it, rooted in choice modelling, produces a preference scale with real discrimination between items. Not just "people liked A" but "A is 3.2x more valued than B."

MaxDiff has lived in quantitative survey tools since the 1980s. Grid-based. Static. Disconnected from the conversations where people actually explain their reasoning.

We brought it into Conveo.

How it works

Add a MaxDiff question to any topic guide. Define your list of items, choose how many appear per set, and set how many rounds each participant completes. During the interview, Conveo's moderator weaves the MaxDiff sets into the conversation naturally, presenting each set as a question and recording the participant's choices.

Behind the scenes, Conveo handles rotation, set composition, and response capture. You see a clean preference ranking in your analysis, ready to use alongside your qualitative themes and quotes.

Why this changes your research

• You get quantitative-grade preference data inside a qualitative study. No separate survey needed, no context switching for participants.

• MaxDiff forces real trade-offs. Participants cannot say "everything is important." The methodology surfaces what truly matters.

• The AI moderator manages all the complexity: item rotation, set balancing, response tracking. You focus on your research question.

• Results integrate directly into your Conveo analysis. Compare preference rankings across segments, overlay them with qualitative themes, or export for further modelling.

Coming next week

Intelligent probing on MaxDiff responses. After a participant makes their choice, the moderator follows up: "You picked X as most important and Y as least. Tell me more about that." This combines the statistical power of MaxDiff with the depth that only qualitative can deliver. A first in the industry.

To get started, open any study, add a MaxDiff question in your topic guide, and configure your items. Your next study can have both the depth of qual and the precision of quant, in a single conversation.