Google Analytics is designed largely for marketers, so product managers have to adapt it carefully to extract the insights they need.
Google Analytics is one of the oldest and most widely used web analytics tools. It enables you to measure how users interact with your website or app, providing insights that can guide marketing, content optimization, and product development. But the actual job is to translate raw data into actionable signals for your product decisions.
Google Analytics originated from a company called Urchin, which Google acquired in 2005. Since then, it has evolved into a powerful platform that tracks visitors by embedding a small piece of JavaScript code on your pages. When a user arrives, this code sends data about their activity to Google’s servers, which process and display it in dashboards.
Understanding what Google Analytics tracks and how it organizes data is critical to avoid confusion and misinterpretation.
The pattern is consistent: Metrics are numbers, dimensions provide context
The foundation of Google Analytics reporting rests on two concepts: metrics and dimensions.
Metrics are numerical measurements of user interactions on your site. They are standalone values that summarize activity — for example, the number of visits, pageviews, or bounce rate. Metrics always appear as numbers and form the columns in reports.
Dimensions are non-numerical attributes that describe data. Examples include the user’s country, the landing page URL, or the traffic source. Dimensions are not meaningful in isolation but provide essential context when paired with metrics. They enable you to segment metrics, such as visits by country or bounce rate by device type.
Talvinder explains:
"A metric is a numerical measure of the user interaction. Metrics will always be expressed in the form of a number. Metrics are standalone entities. When you look at a metric in a standalone fashion, it provides you with information about site-wide performance. These metrics form the columns of a report structure in Google Analytics."
"Dimensions, on the other hand, are non-numerical data fields. Unlike metrics, dimensions are not standalone entities — they are not generally meaningful when viewed individually. But when coupled with metrics, they provide meaningful context and can be used to segment a metric."
For example, knowing your site had 10,000 visits (metric) is useful. Knowing that 4,000 of those visits came from Bangalore (dimension) helps you tailor your product or marketing to that audience.
Common metrics and dimensions product managers should know
| Metric | What it measures | Example use case in product management |
|---|---|---|
| Visits (Sessions) | Number of user sessions on the site | Track changes in overall traffic after a launch |
| Pageviews | Total pages viewed | Identify popular content or features |
| Bounce Rate | % of sessions with a single pageview | Assess landing page effectiveness |
| Average Session Duration | Average time spent per session | Gauge engagement depth |
| Conversion Rate | % of users completing a goal | Measure success of signup or purchase flows |
| Dimension | What it describes | Example use case |
|---|---|---|
| Traffic Source | Origin of user traffic (Google, direct, referral) | Identify which channels drive valuable users |
| Device Category | Desktop, mobile, tablet | Optimize UX for dominant devices |
| Landing Page | First page visited | Analyze entry points and optimize onboarding |
| Location | User’s geographic location | Localize content or prioritize markets |
Filters: the essential tool to clean and segment your data
Raw Google Analytics data can be noisy or misleading if not filtered properly. Filters allow you to exclude irrelevant traffic or focus on specific segments.
Filters apply only to new data after they are set — they do not affect historical data. Talvinder emphasizes:
"Filters are used to clean and segment your data. They provide segmentation to gain a better understanding of a particular subset of activities happening on your website. Filters help customize reports so the most useful data can be highlighted. They also help clean unwanted data so irrelevant information is filtered away."
Common filters product teams use
| Filter Name | Purpose | Indian context example |
|---|---|---|
| Exclude traffic from internal IPs | Remove company employee visits to avoid skewing data | Filter out visits from your office IP range |
| Include only traffic to a subdirectory | Focus on a specific product feature or microsite | Track activity only on your payments portal |
| Exclude spam or bot traffic | Remove suspicious or automated visits | Clean data from fake traffic sources |
Filters must be applied carefully. Talvinder warns:
"Always keep the default profile without filters as a backup. Adding multiple include filters can cause data to disappear unexpectedly because filters are executed in order, and output from one filter becomes input for the next."
Types of filters available
| Filter Type | Description |
|---|---|
| Exclude Pattern | Excludes data matching a pattern (e.g., IP addresses) |
| Include Pattern | Includes only data matching a pattern |
| Search & Replace | Finds and replaces text within fields |
| Uppercase / Lowercase | Forces a field’s text to all upper or lower case |
| Advanced | Combines multiple fields into one for complex filtering |
Goals: measuring what matters for your business
A goal in Google Analytics is any user action that aligns with your business objectives. This could be a purchase, a signup, or reaching a thank-you page.
Talvinder defines goals clearly:
"A goal can be any activity on your website that is important to the success of your business. For simplicity, a web page that displays a confirmation for submitting an order could act as a goal. Each time a visitor meets a particular criterion, a goal is recorded. During a single session, a goal can only be counted once."
Why goals matter
Setting up goals lets you measure conversion rates and understand how well your site or product is driving desired outcomes. Without goals, you only see raw traffic but not business impact.
Common goal types
| Goal Type | Description | Example in Indian product context |
|---|---|---|
| URL Destination | Triggered when user visits a specific page | Order confirmation page after checkout |
| Time on Site | Triggered when user spends more or less than a threshold | Measuring engagement on a content portal |
| Pages/Visit | Triggered when user views more or fewer pages than threshold | Assessing depth of browsing on an e-commerce site |
How Google Analytics works under the hood
Google Analytics tracks users by setting first-party cookies via JavaScript embedded on your pages. When a visitor arrives, the tracking code runs and sends data to Google’s servers.
Some of the cookies used include:
| Cookie Name | Purpose | Expiry |
|---|---|---|
| _utmc | Temporary session identifier | Deleted when browser closes |
| _utmb | Session duration tracking | Expires after 30 minutes of inactivity |
| _utma | Unique visitor identifier | Expires after 2 years |
| _utmz | Campaign/source tracking | Expires after 6 months |
| _utmv | Visitor segmentation | Expires after 2 years |
This cookie-based tracking is how Google Analytics distinguishes new versus returning users, sessions, and traffic sources.
The origins of web analytics and how it became actionable
After the internet’s birth, IT teams maintained server logs capturing IP addresses, browser types, operating systems, and referrers. Scripts extracted useful insights from these logs, birthing web analytics.
Unlike simple web reporting, web analytics is actionable — it enables you to make informed decisions to improve your online strategy.
Talvinder highlights:
"Web analytics helps in making informed decisions about changing your online strategy. It is not just about reports; it is about action."
MeetingScene: A product team discusses Google Analytics data challenges
Product team weekly review at an Indian e-commerce startup in Bangalore
Neha (Product Manager): “Our Google Analytics dashboard shows a spike in bounce rate last week. But I’m not sure if it’s real or due to internal traffic.”
Rahul (Data Analyst): “We forgot to filter out our office IPs after the last release. That likely inflated the bounce rate.”
Priya (Growth Lead): “Can we set up filters to exclude internal traffic and segment by device type? Mobile bounce has been a concern.”
Neha (Product Manager): “Also, let’s define goals for checkout completion and newsletter signups so we can track conversion.”
This is a typical conversation illustrating the importance of clean data and meaningful metrics.
Clean data is the foundation of trustworthy product insights.
SlackChat: Product and engineering discuss implementing Google Analytics goals
FieldExercise: Set up a Google Analytics goal and filter for your product
Title="Implement GA basics for your product" time="15 min"
- Identify one key user action that aligns with your business success (e.g., signup, purchase, content completion).
- In Google Analytics, create a goal tracking that action. Use URL destination or event tracking as appropriate.
- Determine if you have internal or test traffic that skews your data. Set up a filter to exclude that IP range.
- Explore reports by pairing metrics (e.g., sessions, bounce rate) with dimensions (e.g., device, source).
- Write a short note on what you learned about your users and what product question this data could help answer.
JudgmentExercise
scenario="You are the PM at a Series A Indian SaaS startup. Your marketing team reports a sudden drop in signups via Google Analytics. The data shows a 30% increase in bounce rate on the landing page, but your product team has not changed the page recently. You suspect data issues." question="What steps do you take to diagnose the problem and ensure the data reflects reality?" expertReasoning="First, check if any filters were recently added or removed, especially those excluding internal traffic. Verify if the tracking code on the landing page is intact and firing correctly. Investigate whether recent marketing campaigns introduced new UTM parameters that might be misconfigured. Cross-validate GA data with other analytics tools or server logs if available. Communicate with engineering to confirm no deployment broke tracking. Finally, avoid making product decisions until data integrity is confirmed." commonMistake="Jumping to conclusions that the product change caused the drop and directing engineering to fix the landing page without validating data accuracy. Misinterpreting raw metrics without context leads to wasted effort and poor prioritization." />
You are the PM at a Series A Indian SaaS startup. Your marketing team reports a sudden drop in signups via Google Analytics. The data shows a 30% increase in bounce rate on the landing page, but your product team has not changed the page recently. You suspect data issues.
Your task: What steps do you take to diagnose the problem and ensure the data reflects reality?
your reasoning:
FromTheField context="from a Pragmatic Leaders Analytics workshop"
I have seen many PMs struggle with Google Analytics because it is designed primarily for marketing teams. Product managers must torture it to work for them — creating custom reports, setting up meaningful goals, and applying filters to get clean data. The trap is relying on raw metrics without context. Metrics alone are meaningless; dimensions and segmentation provide the story behind the numbers. The real skill is in translating GA data into product insights that drive decisions.
Where to go next
- If you want to learn how to run user research that complements analytics data: User Research Methods
- If you want to master product metrics and KPIs: Metrics and KPIs
- If you want to understand how to run experiments and A/B tests: Experimentation and A/B Testing
- If you want to explore advanced product analytics tools like Mixpanel and Amplitude: Advanced Product Analytics