What Should I Track On My Website

The Short Answer: Track What Connects to Decisions

You should track the specific actions on your website that tell you whether it is doing its job. That means form submissions, phone calls, content engagement patterns, and the journeys visitors take before they convert. Everything else is noise until those fundamentals are solid. The mistake most mid-market companies make is tracking too much (or too little) because nobody sat down and asked, “What decisions will this data actually inform?”

At NexusBond, we work through this question at the start of every website project because what you track shapes how you build. If you wait until after launch to figure out your measurement approach, you end up with a mess of pageview data that tells you almost nothing useful. This article walks you through exactly what to track, why each metric matters, and how to set priorities so your analytics actually help you run the business.

Start With Your Website’s Job, Not a Generic Checklist

Before you open Google Analytics or start listing metrics, you need to be clear about what your website is supposed to accomplish. This sounds obvious, but most tracking setups fail because they are built around a generic template rather than the site’s actual purpose. A B2B services company with a six-month sales cycle has completely different tracking needs than an ecommerce store or a media publisher.

Ask yourself these questions:

• What does a “successful visit” look like? Is it a contact form submission, a demo request, a phone call, a downloaded resource, or something else?
• What does the sales team need to know about leads that come through the website?
• Which pages are supposed to persuade, and which are supposed to inform?
• How do you currently decide what is working and what isn’t?

Your answers define your measurement requirements. In our projects, we define these requirements during prototyping, before a single line of tracking code is written. The reason is simple: if you know what success looks like before you build, you can design the site to make that success measurable. If you retrofit tracking after launch, you are always playing catch-up.

The Core Metrics Every Business Website Should Track

Regardless of your industry, there is a baseline set of metrics that every business website needs visibility into. Think of these as the foundation. You can add sophistication later, but without these, you are flying blind.

Macro Conversions

Macro conversions are the primary actions that directly generate business value. For most B2B companies, these are contact form submissions, demo requests, quote requests, or phone calls. For ecommerce, it is completed purchases. These are the metrics your leadership team cares about, and they should be the first things you configure in your analytics platform.

What we typically find on mid-market sites is that macro conversions are either not tracked at all, tracked incorrectly (counting page loads instead of actual submissions), or tracked in a way that double-counts. A form that redirects to a thank-you page, for example, can inflate your conversion numbers every time someone refreshes that page or bookmarks it. Accurate macro conversion tracking requires event-based measurement, where you fire a tracking event when the form is actually submitted successfully, not when a page loads.

Micro Conversions

These are the smaller actions that indicate a visitor is moving toward a macro conversion. They matter because most visitors will not convert on their first visit. For B2B companies with longer consideration cycles, micro conversions are often the only meaningful signal you have for weeks or months before a deal closes.

Common micro conversions worth tracking include:

• PDF or resource downloads
• Video plays (and how much of the video was watched)
• Email link clicks
• Scroll depth on key pages (pricing, case studies, about us)
• Clicks on “Get Directions” or phone number links on mobile
• Newsletter signups
• Use of interactive tools like calculators or configurators

The trick with micro conversions is being selective. If you track every possible click, your data becomes cluttered and hard to interpret. Only track micro conversions that genuinely indicate purchase intent or meaningful engagement. A visitor who downloads your implementation guide is telling you something very different from someone who clicks your logo.

Traffic Sources and Acquisition Channels

Knowing where your visitors come from is essential for understanding which marketing efforts are working. Your analytics should cleanly separate organic search, paid search, social media, email campaigns, referral traffic, and direct visits. This sounds basic, but clean channel attribution requires discipline. If your team shares links without proper UTM parameters, or if your paid campaigns use inconsistent naming conventions, your channel data will be unreliable.

We recommend establishing a UTM naming convention before you launch any campaigns. Document it, share it with everyone who creates links, and enforce it. The cost of sloppy UTM practices is that your “direct” traffic bucket balloons with visits that actually came from identifiable sources, and you lose the ability to evaluate campaign performance.

Engagement Quality

Raw pageviews are not very useful on their own. What matters is whether visitors are engaging meaningfully with your content. In Google Analytics 4, the engagement rate metric (sessions where users spent at least 10 seconds, had a conversion event, or viewed at least 2 pages) is a significant improvement over the old bounce rate. But you can go further.

Track scroll depth on your most important pages. If you have a services page that is 2,000 words long, knowing that 70% of visitors never scroll past the first third tells you the content structure needs work. Pair scroll data with time-on-page data, and you start to see whether people are actually reading or just leaving the tab open.

Events That Most Companies Miss

Beyond the obvious conversions and engagement metrics, there are several high-value events that most companies neglect. These are the data points that, once you have them, make you wonder how you ever made decisions without them.

Internal Site Search

If your website has a search function, track what people are searching for. This is one of the most underused data sources we see. Site search queries tell you exactly what visitors are looking for and cannot find through your navigation. If dozens of people are searching for “pricing” every month, that is a clear signal your pricing page is buried or missing. If people search for a service you offer but the search returns no results, you have a content gap costing you leads.

Configure your analytics to capture the search term, the page the search was initiated from, and whether the search returned results. This small investment in setup pays for itself within weeks.

Outbound Link Clicks

When visitors click links that take them off your site (to a third-party review platform, a partner site, or a social media profile), that click is valuable data. It tells you what external resources your audience values and, in some cases, signals intent. A visitor who clicks through to your LinkedIn company page after reading a case study is exhibiting different behaviour from one who clicks an external resource link and never returns.

Error Events

Track 404 errors, form validation failures, and broken interactive elements. These are not just technical issues; they are revenue leaks. If 15% of visitors who start your contact form encounter a validation error and abandon it, that is a conversion problem disguised as a technical one. Our team recommends tracking form abandonment as a standard event on every project because the insights it reveals are consistently actionable.

CTA Button Clicks

If you have multiple calls to action on a page, knowing which ones get clicked and which get ignored helps you optimise layout and messaging. Do not just track that “a button was clicked.” Track which button, on which page, in which position. The hero CTA, the mid-page CTA, and the footer CTA may all say “Get a Quote,” but they perform very differently depending on where the visitor is in their reading journey.

What Not to Track (and Why Restraint Matters)

There is a strong temptation to track everything. Modern tag management tools make it easy to fire events on every hover, every click, every scroll pixel. Resist this temptation. Over-tracking creates three problems that are worse than under-tracking.

First, it slows your site down. Every tracking script adds load time, and every event fires a network request. On mobile connections, this adds up. Second, it creates so much data that your team cannot find the signals that matter. When your event list has 200 entries, nobody is reviewing them regularly. Third, it increases your privacy and compliance exposure. Every data point you collect is a data point you need to account for under GDPR, CCPA, or whatever regulatory framework applies to your audience.

The right question is never “can we track this?” but “will someone actually use this data to make a decision?” If the answer is no, do not track it. You can always add tracking later when a specific question arises. What you cannot easily do is untangle a cluttered analytics setup that has been accumulating junk events for two years.

Setting Up a Measurement Framework

A measurement framework is a structured document that maps your business objectives to specific metrics, the events required to capture those metrics, and the tools used to collect the data. It is the single most important artefact in any analytics setup, and most companies do not have one.

Here is what a practical measurement framework covers:

• Business objectives: What the organisation is trying to achieve (e.g., increase qualified leads by 20%)
• Website goals: How the website contributes to those objectives (e.g., generate demo requests from target industries)
• Key performance indicators: The specific metrics that indicate progress (e.g., demo request form submissions, segmented by traffic source)
• Events and triggers: The technical implementation details (e.g., fire event “demo_request_submit” when form ID #demo-form returns a success response)
• Reporting cadence: Who reviews which metrics and how often

Building this framework before you configure any tools ensures that your tracking setup is intentional rather than reactive. For a detailed walkthrough of how this fits into a broader measurement system, see our measurement systems guide.

Tools and Implementation: Getting the Technical Side Right

For most mid-market companies, the right toolset is Google Analytics 4 (GA4) paired with Google Tag Manager (GTM). GA4 handles data collection and reporting. GTM handles the deployment of tracking tags without requiring code changes every time you want to track something new. This separation matters because it lets your marketing team adjust tracking without waiting on developers for every change.

Google Tag Manager Best Practices

Your GTM container should be organised and documented. Name your tags, triggers, and variables using a consistent naming convention so that anyone on the team (or any future consultant) can understand what each element does. A tag named “GA4 – Event – Form Submit – Contact Page” is infinitely more useful than one named “Tag 14.”

Use GTM’s built-in variables and triggers before writing custom JavaScript. GTM offers click tracking, scroll tracking, form submission tracking, and visibility triggers out of the box. Custom code introduces fragility and maintenance burden. Only resort to it when the built-in options genuinely cannot do what you need.

Test every tag in GTM’s Preview mode before publishing. This connects your browser to a debugging interface that shows exactly which tags fired, which triggers activated, and what data was sent. Never publish a GTM container change without testing it first. A misconfigured tag can silently corrupt your data for weeks before anyone notices.

Data Layer Implementation

For more sophisticated tracking, your developers should implement a data layer. This is a JavaScript object that sits on your pages and contains structured information about the page, the user, and any events that occur. Instead of scraping information from the page’s HTML (which breaks every time the design changes), your tags read from the data layer, which is stable and controlled.

Common data layer variables include page type, user login status, form IDs, product categories, and content groupings. On B2B sites, we often push information like the visitor’s industry (if known from a previous form fill) or the stage of content they are viewing (awareness, consideration, decision) into the data layer. This lets you segment your analytics by intent rather than just by page URL.

Connecting Website Data to Business Outcomes

Tracking events in isolation is useful but limited. The real value comes when you connect website data to what happens after the visit. For B2B companies, this typically means connecting your analytics to your CRM so you can see which website interactions preceded closed deals, not just which ones preceded form fills.

There are several practical ways to do this. The simplest is hidden form fields that capture the visitor’s traffic source, landing page, and any campaign parameters, then pass those values into your CRM when the form is submitted. This gives your sales team immediate context about where leads came from and what content they engaged with before reaching out.

A more advanced approach involves matching GA4’s Client ID to your CRM records, allowing you to build a full picture of the anonymous browsing sessions that eventually became customers. This is not trivial to set up, but for companies with average deal values above £10,000, the investment in connecting these systems typically pays for itself within a quarter because you stop spending money on channels that generate clicks but not revenue.

How Often to Review and What to Look For

Data is only valuable if someone looks at it and acts on what they see. We recommend three tiers of review:

Weekly: Check macro conversion volume and conversion rate. Look for sudden drops or spikes that might indicate technical issues (a broken form, a misconfigured redirect) or external factors (a campaign going live, seasonal demand shifts). This review should take 10 minutes.

Monthly: Review traffic source performance, micro conversion trends, and engagement quality across key pages. Compare against the previous month and the same month last year. Look for patterns. Is organic traffic growing but conversion rate declining? That suggests you are attracting the wrong audience or your landing pages are underperforming. This review typically takes 30 to 60 minutes and should involve both marketing and someone from the leadership team.

Quarterly: Step back and evaluate whether your tracking setup still matches your business objectives. New products, new markets, new campaigns, and website changes can all create gaps between what you are tracking and what you need to know. This is also the right time to audit your data quality. Are events firing correctly? Are UTM conventions being followed? Is any data looking suspiciously clean or suspiciously messy?

Prioritising Your First Tracking Setup

If you are starting from scratch or inheriting a messy analytics setup, do not try to implement everything at once. Prioritise in this order:

1. Macro conversions: Get your primary conversion actions tracked accurately with event-based measurement. Validate the data against your actual lead volume. If analytics says you got 50 form submissions last month but your inbox only has 30, something is wrong.
2. Traffic source hygiene: Set up UTM conventions, ensure your channel groupings are correctly configured in GA4, and audit your referral exclusion list so that payment gateways and authentication flows do not create false referral traffic.
3. Key page engagement: Add scroll tracking and CTA click tracking to your five most important pages. For most companies, that is the homepage, pricing page, primary service or product page, contact page, and one high-traffic blog post or resource.
4. Micro conversions: Add tracking for downloads, video plays, and other secondary engagement signals.
5. Advanced events: Site search, form abandonment, error tracking, and outbound link clicks.

Each tier builds on the previous one. Getting the first tier right is more important than having all five tiers partially configured.

Making Your Tracking Setup Last

The biggest threat to good analytics is not the initial setup; it is entropy over time. Websites change. New pages get added without tracking. Forms get rebuilt and the old event triggers break. Team members leave, and their knowledge of the tracking setup leaves with them.

Protect your investment by documenting everything. Maintain a living document that maps every tracked event to its business purpose, technical implementation, and the person responsible for reviewing it. When you make changes to the website, include “update tracking” as a step in the deployment checklist, not an afterthought.

Treat your analytics setup the way you treat your website itself: as a living system that requires periodic maintenance, not a project that is “done” after launch. The companies that get the most value from their data are the ones that build measurement into their workflow rather than bolting it on as a separate initiative. Start with the metrics that connect to real decisions, build outward from there, and audit regularly. That is the difference between a dashboard full of numbers and a tracking system that genuinely helps you grow the business.