DESIGN

Feedback Labs Members' Community Learning Site

Step 1: Design

Designing an AI- and tech-enabled feedback loop is about planning the “what, when, where, and how” of your entire process, with your constituents at the center. It starts by identifying who needs to give and use feedback, what technologies and AI tools are actually accessible to them, and how to reach them in ways that feel simple, engaging, and constructive. This includes using visual and interactive tools to make information clear and appealing, and carefully crafting open- and closed-ended questions that use your audience’s own language.

AI tools and large language models (LLMs) can help you to draft, translate, and simplify surveys, but their outputs must be checked for bias, nuance, and inclusivity by staff.

Design is iterative: you sketch a plan, test it with colleagues and constituents, refine it, and keep returning to it. A strong, user-centered design step sets up the rest of the feedback loop to run more smoothly and makes it more likely that feedback will be both high-quality and acted upon.

The design phase is where feedback questions, methods, and workflows are developed. Getting this phase right, with or without technology, determines the quality of everything that follows. Technology should serve the design, not drive it.

Typical Challenges Organizations Face

Here are common challenges that we have found organizations experience in the Design stage of their feedback loop:

Designing for the tool rather than the community

Organizations sometimes build feedback questions around what a survey platform makes easy, rather than what communities actually need to express.
Technology choices made during buy-in can inadvertently constrain the design, limiting question types, languages, or accessibility options.

Question bias and leading language

Survey and interview questions often reflect the priorities and assumptions of the organization rather than the lived experience of communities.
Translating questions across languages and cultural contexts is frequently under-resourced, leading to instruments that do not translate meaningfully.

Accessibility and inclusion gaps

Digital feedback instruments default to text-based, screen-based, individual formats, which exclude people with low literacy, disabilities, shared devices, or unreliable connectivity.
• Organizations often design a single instrument and assume it will work for everyone.
4. Survey fatigue and respondent burden
• Communities, especially those already navigating hardship, are frequently asked for feedback by multiple organizations. Long, poorly designed surveys compound this burden.
• Without piloting instruments, organizations miss opportunities to catch confusing questions before they reach communities.

Tech Tools & Use Cases

Here are two ways that we have found organizations using technology tools in the Design stage of their feedback loop:

Visual tools

Simplify your design and make it more visually appealing using tools such as:
- Canva
- Loom

Interactive Tools

Improve engagement capacity and make your design interactive with tools such as:

AI Tools & Use Cases

Some organizations have found using Large Language Models (LLMs) to help them design their listening and feedback strategy.

Large Language Models (LLMs)
- ChatGPT, Latimer, Gemini, CoPilot, Claude.ai

Here is a resource to consider if using an LLM to design your strategy is the right move for your organization.

Understanding AI and Tech Guide

Ethical Considerations

Staff, Organization, and AI Model Bias

Ensure the AI model is trained on diverse linguistic patterns; staff review outputs for nuance.

Learn more about ethical considerations with our AI & Feedback Resource.

Page updated

Report abuse