|
This article covers the following: |
Overview
Wingz Agents is a powerful feature within the Wingify platform that enables you to create, customize, and deploy a suite of specialized AI agents.
Each agent is configured using multiple components, such as Agent Name, Agent ID, Goal, Prompt. These components provide full customization and control over your AI experience, enabling you to build agents tailored to specific analytical needs and workflows.
Once you create an agent, you can deep-dive into your campaign data by interacting with it using instructions in natural language. This helps you generate insights faster and makes optimization easy for everyone, regardless of technical skills.
You can also use multiple agents within the Wingz Chat and Workflows to solve a single complex business problem.
For example, a Conversion Rate Optimization (CRO) specialist can use a sequence of agents that are configured to perform specific tasks in the following order to develop a data-backed redesign plan for a checkout page.
-
Segment Analysis: The CRO specialist invokes an agent, for example, Segment Analyst to identify underperforming segments, such as Mobile users in North America with a high bounce rate.
- Behavioral Analysis: The CRO specialist uses these findings in the same chat by invoking another agent, for example, a UX Researcher to analyze session recordings and heatmaps for the Mobile segment to identify friction points. The agent identifies that the Calculate Shipping button is getting overlapped by the mobile keyboard.
-
Strategic Recommendations: Finally, the CRO specialist invokes another agent for example, Design Optimizer, in the same chat to generate ideas for addressing the keyboard overlap issue. This results in high-impact plans such as adding a sticky Proceed button or an accordion-style form in the existing checkout page.
By combining qualitative (recordings, heatmaps) and quantitative (campaign data) agents, you get a 360-degree view of user behavior that a single prompt might miss.
Let us understand the benefits of using Agents.
Key Benefits
- Automate repetitive analysis: Reduces time spent on manual data analysis tasks, such as applying complex filters to campaign reports.
- Generate data-backed ideas: Overcomes creative bottlenecks by using AI to conduct research and suggest fresh test ideas.
- Reuse complex prompts: With Agents, you can save and reuse complex prompts across the Wingz ecosystem, including Chat and Workflows.
Let us understand the key components required to build an agent.
Components of a Wingz Agent
When you create a Wingz Agent, you define its identity and capabilities through a set of core components. These components determine how the agent interprets your requests, what data it accesses, and how it analyzes that data.
The primary components of a Wingz Agent include:
- Agent Name: A descriptive label for the agent that conveys its specific purpose within your workspace, for example, CheckoutPage Analysis Agent.
- Agent ID: A unique identifier used by Wingify to recognize and invoke the agent during chat interactions.
- Goal: A high-level description of what you want the agent to accomplish, for example, Analyze the conversion rates on the New Checkout Page. This serves as the foundation for the agent's logic.
-
Prompt: The detailed workflow and instructions that tell the agent how to process inputs, and structure its output. A detailed prompt includes the following components:
- Role: Defines the persona the AI must adopt, for example, a Conversion Rate Optimization specialist.
- Workflow: A sequential set of steps the agent must follow, such as asking for clarification if a URL is missing or using specific tools to analyze funnels.
- Output: A structured format in which the agent must present the analysis along with recommendations.
Configuring these components allows you to create an agent customized for your specific use cases.
Sample Agent Use Cases
The following table lists sample agents (with their use cases) that you can create to simplify your campaign analysis and manage your campaigns effectively as per your business use case.
| Agents | Use Cases |
|---|---|
| A/B Test Performance Analysis | Tracks A/B test performance, identifies winning variations, evaluates conversion uplift and statistical significance, and provides actionable recommendations. |
| Test Ideation | Generates and prioritizes data-backed test hypotheses for a webpage or campaign using page analysis, CRO best practices, and historical test results. |
| Portfolio Health Monitoring | Provides an account-wide view of the test campaigns by analyzing campaign stages, winners, stalled tests, and campaigns awaiting launch. |
| Heatmap & Behavior Analysis | Analyzes clicks, scrolls, and attention patterns to uncover user behavior trends, friction points, and optimization opportunities. |
| Session Recording Analysis | Analyzes session recordings in bulk to identify recurring behavioral friction, including rage clicks, dead clicks, hesitation, and scroll fatigue. |
| Design-to-Test Automation | Converts a plain-language UI change into a test variation, aligns it with the website's existing design system, verifies the change, and saves it as a draft campaign. |
| Weekly Optimization Reporting | Creates a weekly summary of optimization campaigns, covering campaign status, performance metrics, visitor data, statistical significance, and recommendations. |
| At-Risk Campaign Detection | Identifies campaigns that are underpowered, plateaued, or inconclusive and flags them for attention before they stall. |
| Draft Campaign Configuration Review | Reviews campaign settings and variations before launch to identify potential issues with targeting, goals, traffic allocation, or implementation. |
| Campaign Performance Comparison | Compares campaign performance across variations, audience segments, or time periods to identify meaningful differences and performance drivers. |
| Executive Portfolio Summary | Creates a high-level summary of the optimization program, highlighting key outcomes, winners, underperforming campaigns, and overall impact. |
| Campaign Debrief | Creates a structured post-test summary covering what was tested, the outcome, key learnings, and recommended next steps. |
To start configuring and using specialized agents, see Create, Configure and Use Wingz Agents.
Need more help?
For further assistance and more information, contact Wingify Support.