Free tools Windows power users keep installed
One-click scans. No signup required.
Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Website metrics help you understand how people find your site, what they do there, and whether they take an action that matters to you. Traffic alone cannot tell you whether a blog is building an audience, a small business is getting qualified leads, or an online store is making sales.
Start with four layers: reach (how many visits and views), acquisition (where visitors came from), engagement (what they did), and outcomes (leads, sales, sign-ups, or another goal). A small set of trustworthy measurements is more useful than a dashboard full of unexplained numbers.
What are website metrics?
A metric is a number, such as sessions, page views, or revenue. A dimension describes or groups that number, such as source, device, country, or landing page. An event records an interaction, such as a form submission, purchase, download, or video start. A key event—called a conversion in some tools—is an event you have designated as important to your objectives. A key performance indicator (KPI) is a measurement tied directly to a strategic goal.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →For example, a site might record 1,000 sessions (metric), including 400 from organic search (dimension breakdown), and 30 form submissions (event). If 12 submissions are qualified leads, those may be the business outcome to prioritize. The qualified-lead rate, defined against a stated denominator, could be a KPI. A scroll or button click may help explain behavior, but it is not automatically a business result.
#1 Best Overall
Website analytics is most useful when it connects these questions: How much activity occurred? Where did it come from? What did people do? Did the activity produce value?
The essential website metrics
Reach: users, sessions, and views
- Users: An analytics platform’s estimate of users it recognizes. In GA4, total, active, new, and returning users are distinct measures; Google defines active users around engagement, while new users are associated with first visits or first opens. See Google’s user metric definitions. Use users to gauge audience reach and change over time, not as a count of unique human beings. A person may use several browsers or devices, while consent choices, blockers, and modeling can affect what is recorded.
- Sessions: A measure of visits, useful for comparing channels and landing pages. In GA4, a session generally starts when a user views a page or screen and no session is active; the default timeout is 30 minutes of inactivity. Definitions vary by tool, and sessions are not unique people. Repeat visits, bots, refreshes, or low-quality traffic can raise the count. GA4 traffic acquisition documentation describes its session and report metrics.
- Views or page views: Counts of pages or screens viewed. Views can help identify popular content and editorial demand, but do not prove that visitors read, understood, or valued a page.
Ask what a change means before celebrating it. More users may indicate broader reach; more sessions with no increase in qualified leads or sales may instead reflect repeat visits, irrelevant traffic, or a tracking issue.
Acquisition: how people arrive
Review traffic by source and medium—for example, organic search, direct, referral, social, email, paid search, or display—and by campaign and landing page. GA4’s Traffic acquisition report includes sessions, engaged sessions, engagement rate, key events, and session key-event rate. The useful question is not just which channel brought the most visits, but which brought visitors who reached the relevant outcome.
Use consistent campaign tags on links you control. Common UTM parameters are utm_source, utm_medium, utm_campaign, utm_content, and utm_term. For example:
https://example.com/offer?utm_source=newsletter&utm_medium=email&utm_campaign=summer_sale&utm_content=hero_buttonChoose lowercase values and a shared naming convention before a campaign starts, then document it. Avoid tagging internal links unless you have a specific reason: doing so can overwrite or fragment the original source attribution. Never place sensitive personal information in a URL. Plausible’s metrics documentation describes UTM support and campaign reporting.
Search Console and Analytics answer different questions. Google Search Console reports Google Search visibility, including queries, impressions, clicks, and click-through rate. An analytics tool reports what visitors do on the site after arriving, such as sessions, engagement, events, and leads. Google explains how to connect search performance with landing-page behavior in its guide to using Search Console and Analytics together. Their totals will not necessarily match because they collect and process different data.
Engagement: what visitors do
In GA4, an engaged session is a session that lasts longer than 10 seconds, has a key event, or includes at least two page or screen views. Engagement rate is engaged sessions divided by total sessions. GA4 bounce rate is the inverse: the percentage of sessions that were not engaged. These are GA4-specific definitions, not universal standards; see Google’s engagement and bounce rate definitions.
Recommended Free Tools
Use engagement rate to compare similar landing pages, campaigns, devices, or audience groups—not as a standalone quality score. A reader may get a complete answer from one article and leave; a page with multiple events may still be confusing or irrelevant. A single-page site or landing page may need an appropriate custom event to register meaningful interaction. Plausible’s documentation discusses custom events and this single-page measurement caveat.
Average engagement time can help compare similar content, but time metrics need context. A long visit can mean useful attention, confusion, or an abandoned tab; background-tab behavior and implementation affect what is counted. Do not impose a universal expectation such as a fixed number of minutes per page.
Scroll depth, downloads, video plays, outbound clicks, and form starts are behavioral diagnostics. Low scroll depth could point to a weak opening or a long page. Downloads without inquiries could mean the next step is unclear. Outbound clicks may be exactly the intended result for a directory or affiliate site. Interpret each event against the page’s purpose.
Outcomes: key events, leads, and sales
Plan a small, stable event vocabulary for meaningful actions: for example, generate_lead, sign_up, purchase, subscribe, download, or begin_checkout. Mark only business-relevant actions as key events or conversions. Counting every page view or scroll as a conversion makes outcome reports hard to interpret.
State the denominator whenever you report conversion rate. Two common formulas are:
Session conversion rate = sessions with a key event ÷ total sessions
User conversion rate = users with a key event ÷ total usersThose rates answer different questions. Pick the one that suits the decision, use it consistently, and label it clearly. GA4’s Traffic acquisition report includes key-event reporting once important events are configured.
For a lead-generation site, raw form submissions are only a start. Track qualified leads, calls or appointment bookings, lead conversion rate, cost per lead, cost per qualified lead, lead-to-customer rate, and—where possible—revenue per lead. Spam, duplicates, unqualified inquiries, and missing CRM attribution can inflate form counts.
For an online store, monitor purchases and revenue alongside add-to-cart, checkout-start, and purchase-completion rates, average order value, and refunds. Revenue reporting depends on sending correct ecommerce values, currency, and transaction details and avoiding duplicate purchase events. GA4’s traffic acquisition documentation describes its revenue measures; implementation quality still determines whether those figures are dependable.
A 10-metric beginner starter set
- Users — Is recognized audience reach changing?
- Sessions — How many visits occurred?
- Views — Which content or pages are being viewed?
- Source/medium — Which channels bring visits?
- Landing pages — Where do visits begin?
- Engagement rate — How many sessions meet the tool’s engagement criteria?
- Average engagement time — How does active attention compare among similar pages or segments?
- Key events — How often did important actions occur?
- Conversion rate — What share of sessions or users completed the chosen action?
- Revenue or qualified leads — Did the activity create business value?
You may not need every item. A personal blog could prioritize returning readers and newsletter subscriptions; a service business may care most about qualified leads and booked calls; a store needs reliable purchase and revenue data.
Start with a measurement plan
Decide what success means before choosing reports or tools. A simple plan might look like this:
| Objective | User action | Event | Primary KPI | Useful diagnostics |
|---|---|---|---|---|
| Sell products | Complete checkout | purchase |
Revenue or purchases | Product views, add-to-cart, checkout starts |
| Generate leads | Send a contact form | generate_lead |
Qualified leads | Form starts, landing page, source/medium |
| Grow a newsletter | Subscribe and confirm | sign_up |
Confirmed subscribers | CTA clicks, form completion rate |
| Publish useful content | Read and continue or return | Relevant custom event, if needed | Returning readers or assisted outcomes | Views, engagement, scroll depth |
| Promote a local business | Call or request directions | Call or direction event | Calls or bookings | Device, geography, landing page |
The tool should follow this plan. Otherwise, it is easy to collect a large volume of events without knowing which ones should guide a decision.
How to choose an analytics tool
There is no universal best platform. Consider the reports and integrations you need, how much setup and maintenance your team can handle, your privacy requirements, and whether you want a hosted service or to run software yourself. Privacy obligations depend on geography, implementation, consent choices, vendor terms, and legal advice; no analytics vendor is automatically compliant in every situation.
| Tool | Best fit | Strength | Main trade-off |
|---|---|---|---|
| Google Analytics 4 | Sites using Google products or needing broad event-based reporting | Flexible measurement and integrations | Setup and interpretation can be complex, and noisy event plans can make reports misleading |
| Matomo | Organizations valuing data ownership, self-hosting, or a broad reporting suite | Extensive reports and an open-source self-hosted option | Self-hosting means responsibility for infrastructure, updates, backups, and security; cloud and self-hosted costs and features differ |
| Plausible | Bloggers, creators, and small businesses wanting simpler traffic and goal reporting | Clean dashboard, campaign reporting, goals, and custom events | Less suited to complex product analytics or elaborate advertising workflows; hosted service is paid |
| Fathom | Teams preferring hosted simplicity and willing to pay for low-maintenance reporting | Event and ecommerce tracking, exports, API access, and email reports are advertised | Paid service and potentially less suitable for highly customized analysis |
- Start with GA4 if Google Ads, integrations, or detailed event reporting matter and you can maintain the setup.
- Consider Matomo if ownership, self-hosting, or broader capabilities such as funnels and ecommerce reporting matter and you have the operational capacity. Matomo describes its open-source software as free to use, but that does not mean hosting and maintenance have no cost. Its reports guide covers its broad feature set.
- Consider Plausible if straightforward traffic, campaign, goal, and custom-event reports are enough. Check its current plans and usage limits before committing; its documentation describes metrics, goals, and UTM reporting.
- Consider Fathom if you want hosted simplicity and are comfortable with a paid service. Verify current plans and features on its pricing page.
Pricing, limits, and features can change and may vary by plan, billing period, or region. Verify them directly before purchase. “Free” can mean free software, not a free hosted plan or zero cost to operate. A heatmap or session-recording tool, tag manager, ad platform, CRM, ecommerce platform, and server logs each serve different jobs; none is a substitute for a well-defined analytics measurement plan.
Set up basic tracking
For GA4, Google’s beginner workflow covers creating a property, adding a web data stream, installing Analytics, using reports, and setting up conversions or key events. A practical sequence is:
- Create an Analytics account or property and a web data stream for the site.
- Install the tag using the method appropriate to your site or tag manager. Confirm it is on the correct site and not duplicated by another installation method.
- Check that activity is arriving. Visit the site in a test browser and use the available real-time or debugging reports to verify collection.
- Define the actions that matter, create or identify the corresponding events, and mark only the important ones as key events.
- Test the events on desktop and mobile. Test both successful and unsuccessful form paths, and make sure a purchase or lead is recorded once.
- Use consistent UTM tags for campaigns you control and connect relevant services, such as Search Console, if useful.
- Write down the event names, definitions, campaign conventions, and any filters so future changes are understandable.
If expected data does not appear, check in this order: (1) the tag is present on the intended site; (2) the selected property and stream are correct; (3) consent settings permit the collection being tested; (4) an ad blocker is not interfering; (5) the event name and trigger are correct; (6) a form or checkout is not redirecting before the event fires; (7) cross-domain or payment-provider transitions are configured where needed; (8) the report’s date range and filters are correct; (9) allow for processing time; and (10) use browser developer tools or the vendor’s debugging tools to inspect whether a request was sent.
Reports do not all update instantly, and installing tracking cannot recreate data from before collection began. For purchases, check that transaction details and currency are sent correctly and that refreshing a confirmation page does not count the same order twice. Use a transaction ID for deduplication where the implementation supports it.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsBuild a small dashboard
A beginner dashboard should answer a few recurring questions, not display every available number. Organize it around:
Best Value
- Overview: users, sessions, views, key events, and revenue if relevant.
- Acquisition: sessions and key-event rate by source/medium, top landing pages, and Search Console clicks and impressions.
- Content: top pages, views by page, engagement rate by landing page, average engagement time, and relevant scroll or download events.
- Outcomes: key events by landing page and channel, funnel steps, revenue, or qualified leads.
Useful comparisons include mobile versus desktop, new versus returning users, organic versus paid versus referral traffic, and country or region where relevant. If Search Console data supports it, compare brand and non-brand search. Keep the dashboard small enough that you can explain what you would do if a number moved.
Interpret trends before acting
There is no universal “good” engagement rate, bounce rate, conversion rate, or time on page. Results depend on the site’s purpose, page type, traffic source, device, geography, audience, season, and measurement setup. Compare like with like: the same metric definition, similar page types, appropriate time periods, and comparable segments.
- Traffic rises but conversions fall: Check whether the new visitors come from a different channel, device, or landing page. Review offer relevance, page clarity, form or checkout functionality, and event tracking before concluding that the channel failed.
- Traffic falls but revenue rises: Fewer visits may be more qualified, or order value may have increased. Check purchases, average order value, channel mix, and tracking before treating the drop as a problem.
- Engagement falls after a redesign: Compare the same pages, sources, devices, and periods. Check whether a tracking event or consent setting changed, as well as whether the new layout made key actions harder.
- Search Console clicks fall but Analytics sessions do not: The systems measure different stages and process data differently. Compare their trends and definitions rather than requiring matching totals. Review query, landing-page, and date-range changes.
- A popular landing page produces few outcomes: Inspect the traffic source and visitor intent, then test the page’s offer, calls to action, mobile experience, and form or checkout. Check that the outcome event fires correctly.
- Mobile conversion is lower than desktop: Compare page speed, layout, form usability, payment flow, and traffic sources. A device split is a clue to investigate, not proof that the device itself caused the gap.
- Engagement looks low on a successful one-page experience: Visitors may have found the answer or completed the intended action without navigating further. Check outcome events and whether the tool needs a custom event to represent meaningful interaction.
High traffic with no results can reflect a weak offer, the wrong audience, poor landing-page relevance, a broken form, bots or spam, or attribution lost at a cross-domain transition. Test the experience and tracking before reallocating a budget based on a single report.
Common analytics mistakes and data checks
- Tracking everything: Too many poorly named events create noise. Track actions that answer a real question.
- Calling every event a conversion: Keep behavioral diagnostics separate from outcomes that matter to the business.
- Comparing unlike periods or tools: Definitions, time zones, attribution, consent rates, bot filtering, modeling, and processing delays can differ. Compare direction and trends where exact totals cannot be reconciled.
- Trusting benchmarks without context: A rate is not good or bad without its definition, objective, and relevant comparison group.
- Ignoring privacy and blocked tracking: Consent choices, ad blockers, and implementation can leave data incomplete. Interpret counts as measured activity, not a perfect census.
- Skipping form and purchase tests: An apparent performance drop can be a broken event or checkout, rather than a change in visitor behavior.
- Changing tags without documentation: A CMS plugin and tag manager may install the same tag twice; form events may fire on both submit and thank-you-page load; a refreshed order confirmation may duplicate purchases; and a single-page app may count route changes as page views. Review event counts, keep an implementation record, and deduplicate transactions where supported.
- Assuming every visit is human: Suspicious referrers, unusual geographic patterns, repeated requests, short sessions, and sudden bursts may signal bot or spam traffic. When analytics data looks implausible, review hostnames and server logs as well.
A practical trust checklist: Is the tag firing on the right pages? Are test or internal visits excluded or clearly labeled? Do key events fire once? Are campaign tags consistent? Have forms and purchases been tested end to end? Are consent limitations understood? Do the date range and time zone make sense?
A simple monthly review
- Compare the current period with a relevant prior period; account for seasonality or campaign timing.
- Look at outcomes first: qualified leads, bookings, purchases, or revenue.
- Find the largest meaningful change, then segment by source, device, landing page, or audience.
- Check data quality and the user journey before assigning a cause.
- Write one testable hypothesis, make one focused change, and record what you changed.
- Review the result over a suitable period before deciding whether to keep, revise, or reverse it.
The aim is not a perfect count of every visitor. It is a measurement system consistent enough to show what is changing, where it is changing, and which action is worth taking next.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
