Glossary
The terms used in theconsumer.ai from A to Z.
The terms you meet in the app, each with a short explanation and a link to the page that covers it in detail. How the main terms relate to each other is shown in Core concepts.
A
- Action field: The area a recommendation targets: Product, Price, Communication, Target group, Positioning or Distribution. Read a report
- Ad-hoc question: A single open question you ask a whole panel straight from a thread. It costs credits like a survey run with one question, gives you themes instead of a report, and only you can see it. If an integration asks it, it belongs to the team. Ad-hoc questions & recontact
- Additional Panel: A team panel generated without a free panel slot and paid once at 0.30 credits per person. It takes up no slot; archiving refunds nothing. Pricing and credits
- Analysis: The step after the survey. Answers are read and condensed, quotes are verified, charts are built and the report is written. Progress
- Audience: The label the app uses for a panel in several places, for example in the input field and on cards. Audience
B
- Background: PDFs a study's analysis reads too, for example a briefing. Never evidence. The launch keeps each file as it stands then, and the report names them in a footer. In the Library such files carry the Read chip. Background
C
- Calculation: How a KPI's answers are summarised into one value: top 2 box, top box, bottom 2 box, mean, or the share of one option. Called Aggregation in the KPI settings. KPIs
- Choice question: A question type with 2 to 12 set options. Respondents pick one answer (Single) or several (Multi). Survey design
- Closed scale: The scale question type. Respondents rate on a scale with labelled ends. Survey design
- Connected app: Claude or ChatGPT, connected by a member to exactly one team by signing in. It acts on that member's behalf: designs name the member as author, only that member confirms and uploads, ad-hoc questions stay private to them. It needs no team API key and can be disconnected at any time. Integrations
- Confirmation: A card in the thread where you allow or cancel a paid or consequential step, for example Start the study?. If the design, panel or price changes first, it expires or asks again. Tools & approvals
- Credit: The unit survey runs, ad-hoc questions and additional panels are billed in: one Dynamic Persona answering 15 questions. The price of a survey run is participants × questions ÷ 15, ten times that with Premium Twins. Pricing and credits
- Credit limit: A ceiling owners can set for individual members. It applies per period. Credits, usage & limits
- Crosstab: A CSV download of each question's results, broken down by attributes such as age or region. Export
D
- Design: Short for survey design. It is the name of the Designs tab and of the Design view in the Dock. Survey design
- Dock: The work area to the right of the chat. This is where you work on a study: goal, background and survey runs, plus the views Design, Panel and Report. Run a study
- Dynamic Persona (formerly: Synthetic Consumer, Synthetic Panel): An AI persona generated from population statistics. It does not represent any single real person. Results from a panel of Dynamic Personas show tendencies. Audience
E
- Evidence: The verified respondent quotes a statement in the report is based on. Judge reliability
G
- Goal: The decision a study should support, in one sentence. Every survey run keeps it at launch. The report's headline, summary and recommendations focus on it. Plan a study
I
- Integration: Access to your team from other tools via REST or MCP. An integration can't launch or cancel anything on its own: a member always confirms in the browser. Integrations
- Iris: The AI assistant you talk to in threads. Working with Iris
K
- KPI: A starred scale or single-choice question. In the report it appears as a card with one value. KPIs
- KPI template: A reusable KPI with fixed wording, scale and calculation. It comes as Standard from theconsumer.ai or as a Team template from your team. KPIs
L
- Library: The tabs Panels, Threads, Studies, Designs and Materials below the input field on the home page. The interface
- Link: Bring a study or report from another thread into the Dock so you can read it next to the chat. Run a study
M
- Machine translated: A marker on text that was translated into your language automatically. Languages
- Material: A file you uploaded: an image, a PDF or an Office document. It is either personal or belongs to the team. Iris reads and searches PDFs. Files & materials
- Member: The role for doing research in the team, without managing billing, members or team settings. Members & roles
- More common than in the population: Traits that are expected to occur more often in a new panel than in the population with the same demographic requirements. You see them before you create a panel. Choose or create a panel
N
- No answer (technical): An answer that did not come through for technical reasons, despite retries. It is left out of the question's base and costs no credits. Judge reliability
- Not interpretable: A marker on a question or KPI where more than 20 % of the answers failed for technical reasons. Judge reliability
O
- Open question: A question type with a free-text answer. The interviewer follows up as the probing instructions say. Survey design
- Overview: The first view of a study in the Dock and on the study page. It shows the study status, goal, background and survey runs. Run a study
- Owner: The role with full control over the team, including billing, members, team context and creating panels. One owner is the Primary Owner. Members & roles
P
- Package: 50, 250 or 1,000 credits, bought at the package price of the current plan and valid for 12 months from purchase. Pricing and credits
- Panel: The group of AI respondents who answer your questions. Also labelled Audience in places. Audience
- Panel card: A panel's card showing its type, participant count, age and gender. Flip it to see the distribution and individual profiles. The panel card
- Panel Slot: Room for one team panel you generated for free. How many you have depends on your plan. Archived and additional panels don't take up a slot. Manage panels
- Panel version: A frozen state of a panel with a fixed member list. The report names the version that was surveyed. Audience
- Participants: The number of members in the chosen panel version. Together with the number of questions it sets the price of a survey run. Launch & costs
- Plan: Your team's level: On Demand, Start, Grow or Enterprise. It sets the monthly credits, the package price, panel slots, panel size, Premium Twins and integrations. Pricing and credits
- Premium Twin (formerly: Digital Twin, Twin Panel): A replica of a real person who was interviewed in depth beforehand. A survey run with them costs ten times as much; panels from the Twin Corpus are available from Grow, your own in Enterprise only. Audience
- Preparation: A design with a chosen panel, ready for a survey run to be launched. Launch & costs
- Probing instructions: Tell the interviewer when and how to follow up on an open question, for example after a vague answer. Survey design
R
- Raw data: A ZIP download of all answers of a survey run, one row per respondent. Export
- Recommendation: A prioritised recommendation in the report. Expanded, it shows What to do and Why with a piece of evidence. Read a report
- Recontact: A conversation with a single respondent after an ad-hoc question. Free, up to 30 messages, visible only to you and never part of an analysis. Ad-hoc questions & recontact
- Report: The result of a survey run, with the report basis, summary, KPIs, recommendations and the results of every question. It is created automatically and the status then reads Report ready. Read a report
- Report basis: The top of a report. It states the panel type, panel name, version, number of respondents and when the survey ended. Read a report
- Research question: The central question a survey design sets out to answer. The report answers it in the summary. Survey design
S
- Standard: A label for panels and KPI templates provided by theconsumer.ai. The counterpart is Team. Audience
- Stimulus: An image or PDF a design shows to respondents, for all questions or for single ones. At most 10 per design. Stimuli & materials
- Study: The container for one research project. It holds one or more survey runs. Run a study
- Study page: The full-page view of a study, with the same overview as in the Dock, only to read. Run a study
- Summary: The opening of a report: a headline and a few points, followed by the KPIs and recommendations. Read a report
- Survey design: The research question, questions, stimuli and KPIs of a planned survey run. The team can see and reuse it. Survey design
- Survey language: The language respondents are asked in. Currently always German. Languages
- Survey run: One concrete execution of a study with a fixed design, a fixed panel, its own answers and its own report. The Runs section labels them Run 1, Run 2 and so on. The phase in which the panel answers is the survey (Survey running). Run a study
T
- Target Definition: The Create a new Panel? card, in which Iris describes a new audience before you have it created. Choose or create a panel
- Team: Your shared workspace. Anything that serves the research is visible to the whole team. Manage your team
- Team API key: A key another tool uses to access your team. It is shown only once and can be revoked. Integrations
- Team context: A short text about your team, such as industry, products and wording rules, that Iris takes into account with every request. Team context
- Template: A ready-made survey design you copy and adapt. There are seven standard templates plus your own team templates. Templates
- Theme: A group of similar answers to an open question, with its share and evidence. Read a report
- Thread: A saved conversation with Iris. Only you can see your threads. Threads
U
- Usage: The page with your team's available credits, your usage and the credit limits. Credits, usage & limits