Website Analytics Metrics: Which Numbers Actually Matter

Website analytics metrics fall into two categories: vanity metrics that look impressive in reports but don't predict business performance, and signal metrics that directly correlate with business outcomes. Most analytics dashboards show both without distinguishing between them.

The goal of analytics is not to monitor everything — it is to focus on the website analytics metrics that tell you whether your marketing is working and where it needs improvement.

⠀

Traffic Metrics: Volume and Quality

⠀

Sessions vs. users

Sessions count visits; users count individual visitors (approximately — cookie-based tracking has limitations). For most analysis, sessions is the more useful metric because it reflects how often people are engaging with your site. Users is relevant for reach analysis (how many distinct people are you reaching?).

Traffic by channel

Total traffic volume is rarely actionable. Traffic broken down by channel (Organic Search, Paid Search, Email, Direct, Referral, Social) tells you how your marketing mix is performing. Track channel mix changes over time — a shift from organic to direct often indicates branded search growth, while organic growth confirms SEO investment is returning.

New vs. returning users

New users represent acquisition; returning users represent retention. For informational content sites, high returning user rates indicate loyal audience. For e-commerce, returning users who haven't purchased indicate a nurturing opportunity. Context determines whether the ratio is positive or negative.

Organic search traffic

For businesses investing in SEO, organic search traffic is a primary KPI. Track it separately from total traffic, segment by branded vs. non-branded queries (via Google Search Console), and monitor the trend over 6+ months — SEO ROI is visible at 3–6 month timescales, not week-to-week.

⠀

Engagement Metrics: Are Visitors Finding Value?

⠀

⠀

⠀

Engagement rate (GA4)

GA4 replaced bounce rate with engagement rate — the percentage of sessions where a user was engaged (scrolled, clicked, spent 10+ seconds on the page, or triggered a conversion event). Engagement rate above 60–65% is generally positive for informational content. Below 50% on key landing pages indicates a mismatch between the traffic and the page content.

Average engagement time per session

How long users spend on the site actively engaged (not just with the tab open). This metric reflects content quality and relevance better than session duration did in Universal Analytics because it excludes idle time. Benchmarks vary by industry, but 2–3 minutes for service-based business sites is a reasonable baseline.

Pages per session

The average number of pages a user views per visit. Higher is not always better — it depends on whether you want users to browse deeply (e-commerce, content sites) or convert quickly (single landing page campaigns). Compare pages per session by channel and by landing page to identify where users are engaging versus dropping off.

Scroll depth

What percentage of users scroll to the bottom of key pages? Scroll depth below 50% on long-form content or service pages indicates users are not reading — which means either the content isn't compelling or it isn't matching their intent. Scroll depth is one of the most underused website analytics metrics for diagnosing page quality problems.

⠀

Conversion Metrics: The Numbers That Link to Revenue

⠀

Conversion rate

The percentage of sessions that result in a conversion event. This is the most important metric on any site where conversions are defined. Industry averages range from 1–3% for general e-commerce, 2–5% for service business lead generation, and higher for highly targeted paid traffic campaigns.

Track conversion rate by channel separately — a 5% overall conversion rate masking a 0.5% organic rate and an 8% email rate means very different things for investment decisions.

Cost per conversion (for paid channels)

Cost per acquisition (CPA) connects marketing spend to business outcomes. Define CPA as total channel spend divided by conversions from that channel. Monitor CPA by campaign, ad group, and keyword to identify where spend is efficient and where it needs reallocation.

Revenue per session (for e-commerce)

Total revenue divided by total sessions. This normalizes revenue data for traffic volume changes, making it possible to compare periods accurately even when traffic volume has changed.

⠀

Diagnostic Metrics: Where Are Things Breaking?

⠀

⠀

⠀

Exit rate by page

The percentage of sessions that end on a specific page. High exit rates on transactional pages (cart, checkout, contact form) indicate friction. High exit rates on informational pages (blog posts) are normal. Context determines whether an exit rate is a problem.

404 error frequency

Track 404 errors via Google Search Console (Indexing → Not Found) and GA4 event tracking. A high volume of 404 errors on previously indexed URLs indicates broken links or missing redirects — both hurt user experience and SEO.

Form submission failure rate

If your analytics tracks form submissions, compare form view events to form submission events. A large gap (many views, few submissions) indicates form friction — too many fields, confusing labels, or technical errors. This is one of the most impactful conversion improvements most businesses make.

⠀

Website Analytics Metrics to Stop Tracking

⠀

Total pageviews (in isolation): A page that receives 10,000 views but no conversions is contributing nothing to business outcomes. Pair pageviews with engagement metrics and conversion attribution to make them meaningful.

Social media followers and likes from within website analytics: These belong in social platform reporting, not website analytics. They measure audience size on platforms you don't own.

Alexa rank and domain authority (as standalone KPIs): These are estimated metrics from third-party tools. They correlate with SEO performance but are not direct business outcomes.

Blakfy configures website analytics metrics reporting for clients that focuses on the data that informs decisions — eliminating dashboard noise and building the measurement frameworks that connect web performance to business results.

⠀

⠀

Frequently Asked Questions

⠀

What is a good conversion rate for a website?

Benchmarks vary significantly by industry, traffic source, and conversion type. As general context: e-commerce sites average 1–3% overall (higher for email and branded search traffic). Service business lead generation typically sees 2–5% on well-optimized landing pages. More useful than industry benchmarks is tracking your own conversion rate trend over time and segmenting by channel to identify where the biggest improvement opportunities lie.

Why does my bounce rate look different in GA4 versus Universal Analytics?

GA4 replaced bounce rate with engagement rate (and an inverse engagement rate that functions like bounce rate). The calculations differ significantly — GA4's engagement rate counts any session where a user scrolled, clicked, or spent 10+ seconds as engaged, while Universal Analytics' bounce rate only excluded single-page visits without any interaction. This means GA4 typically shows higher engagement rates (lower bounce equivalents) than Universal Analytics did for the same traffic.

How often should I review my website analytics?

For anomaly detection: configure automated alerts in GA4 and review when triggered. For trend analysis: weekly review of key metrics (conversion volume, traffic by channel). For strategic decisions: monthly deep analysis connecting metrics to business outcomes. Daily analytics review without a specific diagnostic purpose creates more anxiety than insight for most businesses.

What is the difference between sessions and pageviews?

A session is a user's visit to your site — it begins when they arrive and ends after 30 minutes of inactivity (by default in GA4). Pageviews count each individual page loaded within sessions. One session can include multiple pageviews. For measuring traffic, sessions is the more useful metric; for measuring content popularity, pageviews per page is relevant.