Structuring Multi-Concept and Multi-Segment Studies

Many studies test more than one concept, or need to talk to several different userships with different questions. This guide covers how to structure that in Tellet: when to keep everything in one project, when to split into separate projects, and how quotas and comparisons behave in each case.

The core rule: one discussion guide per project

A Tellet project (sometimes called an "environment") runs on a single discussion guide. If every respondent should get the same questions, keep them in one project and use screening or filters to separate them into segments afterward.

The moment a segment, usership, or concept needs different questions, it needs a separate project.

Segments that share a discussion guide can live in the same project. Segments that need different questions need their own project.

Example: a four-concept study

Say you're testing four concepts and each pair needs its own set of questions. Split the work across two projects instead of one:

• Project 1 - Concepts 1 and 2

• Project 2 - Concepts 3 and 4

This keeps each project's sample balanced between the two concepts it covers. If the discussion guide is identical across both projects (same questions, just different stimuli), you can combine the results afterward, either manually or by exporting both projects and merging the data outside Tellet.

You aren't limited to two projects. Splitting into more, smaller projects gives you more flexibility in how you pair and compare concepts, at the cost of more setup and coordination work, since each project needs its own screening, quota, and publishing pass.

Example: four userships with different guides

When four userships each need a distinct discussion guide, create one workspace and add four projects inside it, one per usership. The workspace keeps everything under one roof for the team, while each project's discussion guide, screening, and quotas stay independent.

See Workspaces and Projects for how workspaces and projects relate.

Comparing results

The Compare tab generates side-by-side segment reports, but it only works within a single project, comparing two audience filters built from that project's own data.

• Segments inside one project (e.g. two screener-defined groups, or an age split): use Compare directly.

• Separate projects or environments (different concepts, different userships): Compare cannot combine them. Export the data from each project and merge or compare the results manually outside Tellet.

Plan for this before you launch: if you expect to compare two concepts side by side later, it's simpler to keep them in one project as segments than to reconcile two exports afterward.

Screening and quotas across projects

Build your screener directly in Tellet rather than relying only on panel-level targeting; it qualifies and segments respondents as they enter the interview.

A project supports quotas on exactly one screening question. Enable the "Enable quotas" toggle on that single-select question, then assign a percentage to each qualifying answer so the distribution sums to 100%. You set the quota question and its target distribution during project setup, before you publish.

If your study needs quotas on more than one dimension at once (for example, both age group and usership), and those dimensions require different discussion guides, that's another signal to split the study into separate projects rather than trying to fit it into one project's single quota question.

For the full mechanics of quotas and screening rules, see Screening Questions and Max Participants.

Quick decision guide