The comfortable metrics and the predictive ones
Website analytics dashboards in 2024 are full of data. Session counts, page views, bounce rates, time on page, device breakdown, scroll depth, the volume of available metrics is not the problem. The problem is knowing which of those metrics have a genuine relationship to the commercial outcomes the website is supposed to support, and which are reported because they are available rather than because they are useful.
The transition from Universal Analytics to GA4 in 2023 disrupted many teams' historical data continuity and forced a re-evaluation of which metrics were worth reconfiguring. This was, in many cases, a useful forcing function, an opportunity to ask which metrics the team had actually been using to make decisions and which had been reported habitually without driving any consequential action. The answers were often surprising.
Why bounce rate was retired (and what replaced it)
Google's decision to retire bounce rate in GA4 in favour of engagement rate was not arbitrary. Bounce rate, the percentage of sessions where users visited only one page, was increasingly not predictive of commercial outcomes in a world where single-page applications, news articles, and long-form content could all produce high bounce rates for completely different reasons. A user who reads a 2,000-word article to completion and then leaves counts as a bounce. A user who opens the homepage and immediately closes it also counts as a bounce. The metric treats them identically.
Engagement rate in GA4 measures sessions with at least one of three engagement signals: a minimum time on site (default: ten seconds), a conversion event, or at least two pageviews. It is a more meaningful measure of whether a session involved any genuine interaction with the content, though it is still not a direct commercial outcome measure.
The metrics that correlate with commercial outcomes
The website metrics that have the clearest relationship to commercial outcomes for B2B businesses are the ones that measure progress toward a conversion action, quality of the traffic, and engagement with commercially relevant content.
Pages per session for high-intent visitors: Users who visit multiple pages in a session are demonstrating active interest. For B2B websites, a session that includes the homepage, a solutions or services page, and a case study or pricing page is a meaningful signal of commercial intent, significantly more meaningful than a session that involves only one page or only blog content.
Conversion rate by traffic source: Which channels send visitors who actually convert? Traffic quality varies enormously by source. Branded search traffic typically converts at far higher rates than social media traffic. Referral traffic from relevant industry publications converts better than broad display-driven traffic. Understanding conversion rates by source allows budget and effort to be directed toward the channels that send commercially valuable visitors, not just large numbers of visitors.
Scroll depth on commercial pages: On key landing pages and conversion pages, the percentage of users who scroll to a significant depth (50%+, 75%+) is an indicator of whether the content is engaging enough to hold attention. Low scroll depth on a page that is supposed to build the case for a product or service is a signal that the content is not doing its job, visitors are not reading far enough to encounter the value proposition.
Return visitor behaviour: B2B purchase decisions typically involve multiple website visits over a period of time. The proportion of your traffic that returns, what they look at on return visits, and how their on-site behaviour changes over multiple sessions are meaningful signals about the quality of the consideration relationship. A visitor who returns three times and visits the pricing page on the third visit is a very different signal from a visitor who bounces once and never returns.
The metrics worth tracking are the ones connected to decisions you need to make. If a metric has never caused you to do anything differently, it is reporting overhead, not insight.
Connecting web data to CRM data
The most powerful upgrade to website analytics for B2B businesses is connecting web behaviour data to CRM data, creating the ability to see how customers and prospects who are in the pipeline interact with the website. Which pages do prospects visit before a sales conversation? Which content do customers engage with before a renewal? Which entry points are most associated with deals that eventually close?
This connection requires integration work, linking the analytics platform to the CRM via identifier matching, but the investment pays off in a significantly more commercially grounded understanding of the website's contribution than web analytics alone can provide. It also enables the sales team to see web engagement data for their prospects, giving them useful context for conversations about what the prospect has been researching.
The review rhythm that makes the difference
Having the right metrics is necessary but insufficient. The metrics earn their value when they are reviewed on a consistent rhythm, weekly for the leading indicators, monthly for the trend analysis, with a clear habit of connecting what the data shows to specific actions. A website metrics review that consistently ends with two or three specific improvements to test or questions to investigate will produce compounding improvement in website performance over time.