Getting Started with Decode

What is Decode?

Decode is Entropik's unified research platform built for researchers, insights teams, marketers and product teams who need to understand how people think, feel and behave. It brings together user research, consumer insights, creative testing, AI-moderated interviews and a central research repository in one place.

Most research tools capture what people say. Decode captures what people actually experience. It does this using Emotion AI, a combination of facial coding, eye tracking, voice analysis and behavioural signals, that runs alongside every study to measure attention, emotional response and behaviour in real time. The result is a more complete picture of how someone responded to a product, creative or experience, not just their stated opinion of it.

The platform is organised around five pillars. Each pillar covers a specific type of research or capability. You can run a study using one pillar or combine several across a larger research programme. 

The Five Pillars 

User Research captures how people think, feel and behave when using a product, feature or experience. It uses Emotion AI alongside survey and task responses to measure attention, emotional reactions and behavioural patterns during sessions. 

When to use it 

Use User Research when you are testing a product or feature before launch, identifying where users struggle or drop off, comparing design or prototype options, or trying to understand why users behave in a particular way. 

How Emotion AI works here 

Facial coding identifies emotional responses in real time. Eye tracking records where a participant looks and for how long. Mouse tracking captures hesitation and click patterns. This data is captured alongside survey or task responses, giving a more complete picture of how someone experienced something.   

Research methods 

• General Research: Covers what users think, expect or prefer. Use this for surveys and structured feedback sessions.
• Advanced Research: Covers how users compare options or evaluate usability. Use this for prototype testing and design preference studies.
• Task Research: Captures whether users can complete real tasks on a live site or app. Use this when testing a live website, app or prototype in a real environment.
• Live Interviews: Capture the story or motivation behind a user's behaviour. Use this for one-on-one sessions with a human moderator where depth of understanding matters.
• Group Discussions: Surface patterns and perspectives across a group. Use this for focus groups and collaborative sessions where shared attitudes are the focus. 

For a full guide to setting up a User Research study, refer to the User Research article 

https://help.getdecode.io/en/article/7a1e2f4e-5937-428f-988e-f787e3ababde

Consumer Insights covers research into how people respond to brands, products and content. It includes quantitative methods such as surveys, conjoint studies and NPS, as well as media testing and in-context creative testing. 

When to use it 

Use Consumer Insights when you are understanding consumer attitudes toward a brand or category, testing how people respond emotionally to an ad, video or image, identifying which product features matter most to consumers, or testing content on a simulated social media feed before publishing. 

How Emotion AI works here 

In Media Testing, facial coding captures emotional reactions as a participant watches a video or views an image, frame by frame. Eye tracking shows what they notice first and what they miss. In In-Context Testing, participants interact with content as it appears inside a simulated Instagram, TikTok, YouTube or LinkedIn feed, and Decode records both their stated feedback and their behavioural responses.   

Research methods 

• General Research: Covers consumer attitudes, sentiments and preferences. Use this for brand surveys and category research. 
• Advanced Research: Covers which product features matter most and how consumers weigh them against each other. Use this for conjoint analysis, NPS and satisfaction surveys.
• Media Testing: Captures how consumers respond emotionally to video or image content, frame by frame. Use this for ad testing and creative evaluation before a campaign goes live.
• In-Context Testing: Measures how content performs inside a simulated social media feed. Use this for Instagram, TikTok, YouTube and LinkedIn content before publishing.
• Live Interviews: Surface the deeper motivations behind a consumer's choices. Use this for one-on-one sessions with a human moderator.
• Group Discussions: Reveal shared attitudes or perceptions across an audience. Use this for category exploration and brand perception studies.

For a full guide to setting up a Consumer Insights study, refer to the Consumer Insights article 

https://help.getdecode.io/en/article/advanced-blocks-overview-copy

AI Creative Insights analyses visual creatives using predictive AI. You upload an image such as an ad, banner, packaging design or landing page and Decode produces scores and visual maps showing where attention lands, whether key elements are being noticed and how clear the message is. No participants are needed. 

When to use it 

Use AI Creative Insights when you are evaluating multiple creative options quickly before committing to live research, checking whether a logo, CTA or headline is getting noticed, comparing two creative versions to identify which performs better, or understanding what a viewer is likely to notice first. 

What you get 

• Attention Score: Measures how much attention the creative captures overall.
• Focus Score: Shows whether attention is going to the right elements such as the CTA, logo or headline.
• Clarity Score: Shows whether the core message is easy to understand at a glance. 
• Heatmap: Shows which areas of the creative draw the most and least attention. 
• Fogmap: Shows the areas of the creative that viewers are least likely to notice.

For a full guide to using AI Creative Insights, refer to the AI Creative Insights articles

https://help.getdecode.io/en/article/dea4866b-d78e-4634-8a0d-2961e4900607

AI Moderation replaces the human moderator in a qualitative interview session. You create a discussion guide, a structured list of questions and topics, and upload it to Decode. Participants join a session and respond to questions posed by the AI. The AI follows up based on each response, keeping the conversation structured. Once all sessions are complete, Decode analyses responses and identifies themes and patterns. 

What a participant experiences 

Participants join a session and interact with the AI moderator in a conversational format. They read a question, respond by typing or speaking, and the AI follows up based on what they said. Sessions are structured but conversational, not a form to fill in. A typical session runs between 15 and 30 minutes.  

When to use it 

Use AI Moderation when you are running qualitative research across a large number of participants, reaching participants across multiple locations without scheduling constraints, ensuring consistency across all sessions, or getting qualitative findings faster than live moderation allows. 

Live Interviews vs AI Moderation 

Both methods produce qualitative insight. The right choice depends on the topic, the number of participants and the timeline. 

For a full guide to setting up an AI Interview, refer to the AI Interview article 

https://help.getdecode.io/en/article/ba707c34-9a32-41ae-b7d9-db9e441ad01a

Insights Hub is where research materials are stored in Decode. When a study is complete, you can upload the report, raw data, presentation or any related documents and link them to the project they belong to. Anyone on the team can search for past studies, access documents and see what research has already been done on a topic. 

What you can do 

• Upload and store files: Store reports, presentations, PDFs, video recordings and raw data linked to the project they belong to.
• Search past research: Search across all studies and files in your workspace to find relevant research without asking around or going through emails.
• Share with your team: Share research directly from Insights Hub with other teams or stakeholders. 
• Review before starting: Check what research has already been done on a topic before commissioning a new study.

For a full guide to using Insights Hub, refer to the Insights Hub articles 

https://help.getdecode.io/en/article/0398a30b-5063-4291-856a-30493c152f35

How the Pillars Work Together 

The pillars are designed to work together. Not every project will use all five. Your starting point depends on what you are trying to find out. 

A team testing a new product might start with AI Creative Insights to evaluate early concepts, run a User Research study to identify usability issues, follow up with Live Interviews to understand the reasons behind specific behaviours, and store all findings in Insights Hub for future reference.  

A marketing team evaluating a campaign might only use AI Creative Insights and Media Testing. A research team running a large qualitative programme might use AI Moderation and Insights Hub. Start with the pillar that fits the research question.