Aha! Launches Virtual User Testing: Find Issues in AI-Built Apps Before Launch
In July 2026, Aha! added Virtual User Testing to Aha! Builder, its AI app coding tool. With help from Elle, the AI assistant, Builder turns an idea into a proof of concept or application. Virtual user testing gives you an objective starting point: a way to catch what you might miss when reviewing your own work.
Nothing replaces feedback from real end users. But it is helpful to resolve as many issues as possible before sharing your concepts with real people.
Test before you deploy
Virtual user testing works like this:
- Define the job to be done: focus the virtual user on a specific workflow or scenario.
- Choose an optional persona: test from the perspective of a particular customer or user type.
- Review structured feedback: see what is working well and where there is room to improve.
- Turn feedback into action: create ideas from the findings, or ask Elle to help implement changes right away.
As you make changes, run another virtual user test to see whether the fix resolved what the last one flagged. It is a fast way to check progress between iterations. The virtual user moves through your application with screenshots of each step, along with the issues and recommendations it found.
How to get started
- Click the Run virtual user testing icon from Preview or Features/Ideas.
- Describe the test scenario.
- Optionally, select one of your application's personas.
- Click Submit to start the test.
- Review the full report and prioritize any updates.
- Click Create idea to save a recommendation as an idea record.
- Ask Elle to implement the changes you want to make.
Once you are confident in the app experience, bring in actual users — the feedback widget can route what they tell you into the same ideas list the virtual tests already started. Between the two, you get a fuller picture of what to fix before your next release.
Virtual user testing is now available for Aha! Builder customers on the Team and Scale plans.
How virtual and real users work together
Virtual user testing is valuable because it is cheap, fast, and repeatable: based on the task and persona you define, it walks the critical paths in minutes, which makes it ideal for a quick regression after every iteration. Real user testing is valuable because it is real: people bring usage patterns, emotional reactions, and context you never anticipated. The right sequence is to clear the obvious issues with virtual users first, then use real users to validate assumptions and dig out deeper needs.
Aha! connects the two through the ideas list — recommendations generated by virtual tests and comments submitted by real users through the feedback widget land in the same queue, so the team can prioritize them together. That way "validation" stops being a one-off gate before launch and becomes a daily habit across prototyping, development, and release.
A concrete example: walking through a checkout flow
What virtual user testing actually looks like is easiest to grasp with an example. Say you have built a landing page with a cart and want to confirm the checkout flow has no obvious friction. You could describe the task like this:
As a visitor who wants to buy the Business plan, I start from the homepage, choose a plan, complete the billing information, and confirm I receive an order-success message; please record the outcome of every step and anything that confuses me along the way.
After picking or customizing a persona (for example, "a small-business owner visiting for the first time who is not very familiar with web operations"), the virtual user walks the path with screenshots. It might find that the plan comparison table gets cut off on mobile, the billing form's country field has no default value, or the confirmation page lacks a clear order summary. Fix each issue, run another round, and keep going until the critical path is clean. Each round takes minutes, which makes it ideal for a quick regression before every commit.
Virtual vs real user testing at a glance
| Dimension | Virtual user testing | Real user testing |
|---|---|---|
| Cost | Very low, included in the plan | High, requires recruitment and incentives |
| Speed | Report in minutes | Days to weeks |
| Coverage | The task paths you define | How users actually behave |
| Emotional feedback | Limited, mostly rational | Rich, with emotion and context |
| Best stage | Quick regression after each iteration | Validation before launch and after release |
The two are not either-or but a relay: clear the obvious issues with virtual users first, then let real users run deep validation on a clean app.
A few usage tips
- Make the task specific: "try it out" is too vague; write the job to be done as executable steps so the virtual user knows which path to take.
- Do not overdo personas: the default persona is usually enough; only customize when you care about a specific audience (mobile users, new users).
- Do not treat the report as conclusions: the virtual user surfaces "candidate problems"; merge duplicates and then prioritize so you are not buried in dozens of findings.
- Keep real feedback in the same queue: route virtual tests and feedback-widget comments into one ideas list for a single decision, avoiding two competing sets of priorities.
Reference: Aha! Builder product page https://www.aha.io/aha-builder
16IDC perspective
For teams using AI to generate prototypes and applications quickly, virtual user testing automates the validation step too — a key move in Requirements Analysis from "idea" to "verifiable product": expose friction cheaply with virtual users before investing in real users and resources. When building websites, use landing page design to define core conversion paths early, and tech stack evaluation to decide the stack for prototypes versus production.
Source: https://www.aha.io/blog/test-ai-built-apps-with-virtual-users-in-aha-builder