FeedAIde: Guiding App Users to Submit Rich Feedback Reports by Asking Context-Aware Follow-Up Questions

Ali Ebrahimi Pourasad; Meyssam Saghiri; Walid Maalej

cs.SEMar 4, 2026

FeedAIde: Guiding App Users to Submit Rich Feedback Reports by Asking Context-Aware Follow-Up Questions

Authors: Ali Ebrahimi Pourasad, Meyssam Saghiri, Walid Maalej

Organizations: University of Hamburg, Hamburg, Germany

Abstract

User feedback is essential for the success of mobile apps, yet what users report and what developers need often diverge. Research shows that users often submit vague feedback and omit essential contextual details. This leads to incomplete reports and time-consuming clarification discussions. To overcome this challenge, we propose FeedAIde, a context-aware, interactive feedback approach that supports users during the reporting process by leveraging the reasoning capabilities of Multimodal Large Language Models. FeedAIde captures contextual information, such as the screenshot where the issue emerges, and uses it for adaptive follow-up questions to collaboratively refine with the user a rich feedback report that contains information relevant to developers. We implemented an iOS framework of FeedAIde and evaluated it on a gym's app with its users. Compared to the app's simple feedback form, participants rated FeedAIde as easier and more helpful for reporting feedback. An assessment by two industry experts of the resulting 54 reports showed that FeedAIde improved the quality of both bug reports and feature requests, particularly in terms of completeness. The findings of our study demonstrate the potential of context-aware, GenAI-powered feedback reporting to enhance the experience for users and increase the information value for developers.

Figures & tables

Figure 1. Overview of $FeedAIde$ . When a user triggers the feedback process, $FeedAIde$ collects contextual data, such as a screenshot (1). $FeedAIde$ then predicts possible feedback, such as a bug report, from which the user can choose or enter their own (2). Next, $FeedAIde$ generates $i$ context-aware follow-up questions (3), asks the user for answers to gather additional details (4), combines all information into a rich feedback report (5), and sends it to the developers (6).

Figure 2. $FeedAIde$ iOS framework guiding users through a context-aware feedback flow. In this example, a user reports that their daily streak in a language learning app has unexpectedly reset to one (Screen 1). Once triggered, e.g. via shake-to-report, $FeedAIde$ detects the context and proposes possible feedback (Screen 2). After the user selects “My daily streak suddenly reset”, the system asks adaptive follow-up questions to collect valuable information for the developers (Screens 3–4). It then creates a rich feedback report and confirms its submission (Screen 5).

Figure 3. UML class diagram of the feedback report generated with $FeedAIde$ .

Figure 4. Existing feedback mechanism in the PPEmployee app, where users can submit feedback via the FAQ section by tapping the light bulb icon.

Participant	Age	Gender	Occupation
P1	21	M	Fitness trainer
P2	51	M	Club General Manager
P3	21	M	Fitness Trainer
P4	25	F	Fitness Trainer & Service
P5	25	M	Personal & Fitness Trainer
P6	27	M	Team Leader

Table 1. Participants in the user testing.

After completing all scenarios (per system)
Q1	How easy did you find the process of submitting your feedback?
	Scale: Very difficult (1) – Somewhat difficult (2) – Somewhat easy (3) – Very easy (4)
Q2	How helpful do you find the system for submitting your feedback?
	Scale: Not helpful (1) – Somewhat helpful (2) – Helpful (3) – Very helpful (4)
At the end of the session (Responses given verbally)
Q3	Explain which feedback system you prefer.

Table 2. Survey questions to evaluate participants’ perceptions of ease, helpfulness, and overall experience when comparing the traditional feedback system and

FeedAIde

Participant	Traditional		$FeedAIde$		Preference
Participant	Ease	Helpfulness	E.	H.	Preference
P1	3	1	3	4	New
P2	3	2	3	3	New
P3	4	1	4	3	New
P4	3	1	4	4	New
P5	3	1	4	4	New

Table 3. User ratings of the traditional feedback system (text field) versus

FeedAIde

Quality Dimension	Traditional	$FeedAIde$
Observed Behavior	0.93	1.50
Expected Behavior	2.00	2.00
Steps to Reproduce	0.14	2.00

Table 4. Average bug report quality ratings (n=28; Traditional System n=14;

FeedAIde

n=14). Ratings were given on a 3-point scale (0,1,2).

Quality Dimension	Traditional (%)	$FeedAIde$ (%)
Summary	100.0	100.0
Description	0.0	100.0
Specify problem	0.0	53.8
Correct summary	69.2	92.3
Product version	0.0	100.0
Correct language	92.3	100.0

Table 5. Average feature request quality (n=26; Traditional n=13;

FeedAIde

n=13).

Explore similar work

LikeThis! Empowering App Users to Submit UI Improvement Suggestions Instead of Complaints
Mar 4, 2026Jialiang Wei, Ali Ebrahimi Pourasad, Walid MaalejSoftware Engineering Human-AI Collaboration
University of Hamburg, Hamburg, Germany
From Reviews to Requirements: Can LLMs Generate Human-Like User Stories?
Mar 30, 2026Shadman Sakib, Oishy Fatema Akhand, Tasnia Tasneem +1Requirements Engineering Text Generation
Department of Computer Science and Engineering, Islamic University of Technology (IUT), Gazipur, Bangladesh
Designing the Future of User Feedback for Generative AI
Oct 2, 2026Alisa Frik, Julia Bernd, Amitis Karami +1Generative AI Evaluation Human-in-the-Loop Evaluation
International Computer Science Institute, Berkeley, CA, USA · University of California, Berkeley, Berkeley, CA, USA · eBay Inc., San Jose, CA, USA

Mar 4, 2026J/K move · Enter open · S save

LikeThis! Empowering App Users to Submit UI Improvement Suggestions Instead of Complaints

Jialiang Wei, Ali Ebrahimi Pourasad, Walid Maalej

University of Hamburg, Hamburg, Germany

Software Engineering Human-AI Collaboration

Open paper AskPDF arXiv

User feedback is crucial for the evolution of mobile apps. However, research suggests that users tend to submit uninformative, vague, or destructive feedback. Unlike recent AI4SE approaches that focus on generating code and other development artifacts, our work aims at empowering users to submit better and more constructive UI feedback with concrete suggestions on how to improve the app. We propose LikeThis!, a GenAI-based approach that takes a user comment with the corresponding screenshot to immediately generate multiple improvement alternatives, from which the user can easily choose their preferred option. To evaluate LikeThis!, we first conducted a model benchmarking study based on a public dataset of carefully critiqued UI designs. The results show that GPT-Image-1 significantly outperformed three other state-of-the-art image generation models in improving the designs to address UI issues while keeping the fidelity and without introducing new issues. An intermediate step in LikeThis! is to generate a solution specification before sketching the design as a key to achieving effective improvement. Second, we conducted a user study with 10 production apps, where 15 users used LikeThis! to submit their feedback on encountered issues. Later, the developers of the apps assessed the understandability and actionability of the feedback with and without generated improvements. The results show that our approach helps generate better feedback from both user and developer perspectives, paving the way for AI-assisted user-developer collaboration.

Abstract

Explore similar work

LikeThis! Empowering App Users to Submit UI Improvement Suggestions Instead of Complaints

From Reviews to Requirements: Can LLMs Generate Human-Like User Stories?

Designing the Future of User Feedback for Generative AI