# AI Analytics co-pilot

### **Overview**

AI Analytics Assistant is an integrated analytics co-pilot built directly into PageFly Analytics, helping you understand and optimize your store's performance without needing to be an analytics or e-commerce expert.

What makes it special: **AI can analyze your data, detect anomalies, and provide specific recommendations** based on your real data and personalized info, even before you ask.

### **How to start using**

1. Access **Analytics** page
2. Click the **AI button** in the top right (available on all Analytics screens)
3. **Chat with AI:** Ask anything about your data
4. **Customize AI:** Click adjustment icon → declare goals/business model/experience level

<figure><img src="https://2800568630-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FeLf471SpmnojuhVGQb2E%2Fuploads%2FM29GocKocKKGP2txHDQ5%2Fimage.png?alt=media&#x26;token=62403224-de49-480c-9df2-04dd2842eb66" alt=""><figcaption></figcaption></figure>

### **What features are available?**

#### **1. Learning & Support**

AI helps you understand analytics from the basics to the advanced:

* **Explains all metrics:** Bounce rate, Conversion rate, Sessions, Gross sales, etc. – AI explains in plain language, avoiding technical jargon.
* **Feature guidance:** How to read and navigate dashboards, how to monitor data by devices/countries, how to compare time periods, etc.
* **Technical troubleshooting:** GA4 won't connect? Data discrepancies? AI walks you through step-by-step troubleshooting.
* **Analytics skill building:** Instead of just giving answers, AI guides you on how to analyze data yourself so you can make informed decisions.

***

#### **2. Contextual data analysis**

AI answers specific questions based on your available store's data on your GA4:

* **Time-based comparisons:** This week vs last week, this month vs last month, or custom date ranges, etc.
* **Segment analysis:** Mobile vs desktop, traffic sources, new vs returning visitors, etc.
* **Page-level comparisons:** Which pages convert best? Which needs improvement?

**Key strength:** When you ask about any metric, AI automatically:

1. Retrieves current data
2. Compares to the previous period
3. Scans for anomalies
4. Provides insights and action recommendations – all in a single response.

***

#### **3. Automatic anomaly detection**

AI proactively identifies issues or opportunities every time you query any metric. If an anomaly is detected, AI highlights it prominently in the response.

**Some types of anomalies:**

* **Sudden changes:** Metrics change >30% in <7 days, traffic spike/drop >50%
* **Pattern breaks:** Weekly trends disrupted, expected weekend peaks don't occur
* **Segment-specific issues:** Mobile performs abnormally worse than desktop, specific traffic source conversion drops suddenly

***

#### **4. Specific action recommendations**

AI doesn't stop at identifying problems – it provides concrete next steps based on your data and customized knowledge base:

**Some types of recommendations:**

* **Quick wins:** Mobile load time 4.8s → Consider compressing images
* **Technical improvements:** High bounce rate on slow pages → Optimize assets
* **Business intelligence:** Traffic source X converts 3x better → Increase ad spend there

***

#### **5. AI customization**

You can customize how AI responds to match your store's specific context:

<figure><img src="https://2800568630-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FeLf471SpmnojuhVGQb2E%2Fuploads%2F6PhhjSMlp5ZKUqef7Dnj%2Fimage.png?alt=media&#x26;token=f1a8cbf3-9946-4162-bbbd-46b44a72d13b" alt=""><figcaption></figcaption></figure>

**Information you can configure:**

* **Main goals:** Increase conversion rate, reduce bounce rate, increase traffic, increase AOV, etc.
* **Business model:** E-commerce, Dropshipping, Subscription, POD, D2C, etc.
* **Experience level:** Beginner, Average, Expert
* **Your role:** Business owner, Marketer, Developer, Agency

**Impacts:**

* **Explanation depth:** Beginners get every term explained, Experts receive denser insights
* **Insight priorities:** If the goal is "Increase conversion rate", AI highlights CVR-related insights first
* **Presentation style:** Business owners receive insights framed by revenue impact, Marketers by campaign performance

***

#### **6. Confidence levels – Transparency about reliability**

AI always clarifies the confidence level of insights:

* **High confidence:** Statistical significance present, clear pattern
* **Medium confidence:** Limited data, suggestive trend
* **Low confidence:** Hypothesis only, needs testing

→ **You know when to trust actions, when to proceed cautiously.**

***

### **What AI can't do**

* Can't access data outside your connected GA4 account
* Can't predict future performance with certainty
* Can't access personally identifiable information about your customers
* Can't override PageFly's or Shopify's policies
* Can't directly modify your store: AI only analyzes and recommends - you have to implement changes yourself
* Can't guarantee results: AI only gives suggestions without any commitments
* Can't replace GA4 or Shopify analytics: AI analyzes data from these sources, doesn't collect data itself
* Can't provide links/CTAs: AI won't redirect you outside PageFly or sell anything
* Can't work well with insufficient data: If the sample size is too small, AI may answer with low confidence

***

**More questions?** Chat with AI directly in Analytics – it will explain almost everything! 🚀