Skip to main content
Digital13 June 2026• 8 min read

Audience Analytics for UK Journalists: GA4, Chartbeat & Privacy-First Tools

Understanding your audience is not optional for modern journalists — it shapes commissioning decisions, headline testing, distribution strategy, and editorial priorities. This guide covers the analytics platforms used in UK newsrooms, how to read the metrics that matter, and how to stay on the right side of ICO rules on cookie consent and UK GDPR.

9 min read

Last reviewed: Next review due:

Quick answer

UK news publishers using GA4 or Chartbeat must comply with the Privacy and Electronic Communications Regulations (PECR) and UK GDPR. Analytics cookies require prior, freely given consent unless you switch to a cookie-free tool such as Plausible or Fathom. Key metrics worth tracking are engagement rate, scroll depth, and session source — not raw pageviews or bounce rate alone. Attribution defaults to last-click; data-driven attribution gives a more accurate picture of what actually drives traffic.

This guide is for journalists, digital editors, and editorial staff who want to interpret audience data confidently and set up analytics correctly without breaching UK privacy law. It covers both the technical basics and the regulatory framework, and is relevant to staff at national publishers, regional titles, and independent news sites alike.

Setting Up GA4 for a News Site

Google Analytics 4 (GA4) replaced Universal Analytics (UA) as Google's primary analytics platform in July 2023. To set up GA4 on a news site, create a property in your Google Analytics account, then install the tracking code either directly in your site's <head> or via Google Tag Manager (GTM). GTM is strongly recommended for newsrooms: it lets editors deploy and update tags without developer involvement and keeps all tag logic in one place.

For news sites, configure the following GA4 settings at the outset:

  • Data retention: Set to 14 months (the maximum available in GA4) to allow year-on-year comparisons in Explorations reports.
  • Enhanced measurement: Enable scroll tracking (fires at 90% scroll depth by default), outbound click tracking, and file download tracking. These are set in the data stream configuration within your GA4 property.
  • Internal traffic filter: Create an IP filter to exclude traffic from your own offices, editorial team, and developer IP ranges. This prevents your own browsing from inflating engagement metrics.
  • Cross-domain measurement: If your site spans multiple subdomains or related domains (e.g. a separate subscription portal), configure cross-domain tracking to avoid sessions being broken when readers move between domains.
  • Consent mode: Implement Google's Consent Mode v2 if you use a consent management platform (CMP). This allows GA4 to model behaviour for users who decline cookies, reducing the gap in your data. Consent Mode is a regulatory requirement for UK publishers using GA4 under PECR (see below).

Key GA4 Reports: Engagement Rate, Sessions, Scroll Depth

GA4 surfaces several reports that are immediately useful for editorial teams. The most important for newsrooms are:

  • Engagement rate: The percentage of sessions that were “engaged” — defined as sessions lasting more than 10 seconds, resulting in a conversion event, or containing at least two pageviews. This is GA4's replacement for the UA bounce rate (inverted). A news article engagement rate above 50–60% is generally healthy; very short pieces or traffic from social media previews often sit lower.
  • Sessions by traffic source: Found in the Acquisition > Traffic Acquisition report. Breaks down sessions by channel group (Organic Search, Direct, Organic Social, Referral, Email, etc.). Use this to understand whether your audience is driven by SEO, social distribution, newsletters, or direct readership.
  • Scroll depth: The enhanced measurement scroll event fires when a user reaches 90% of a page. For longer articles, create custom events in GTM at 25%, 50%, and 75% to build a fuller scroll depth picture. High scroll depth on long-form investigations indicates genuine readership, not just landing and leaving.
  • Pages and screens report: Shows which articles or pages receive the most views and sessions. Combine this with engagement rate data — a high-pageview article with low engagement may be attracting clickbait traffic; a lower-pageview article with very high engagement may be your most genuinely impactful journalism.
  • Explorations: GA4's free-form analysis tool allows you to build custom reports. Useful configurations for newsrooms include funnel explorations (tracking subscriber conversion paths) and cohort analyses (measuring whether readers who arrive via search return directly).

Universal Analytics vs GA4: What Changed

The shift from UA to GA4 introduced significant conceptual changes that still trip up journalists and editors who learned analytics on the older platform:

  • Session definition: In UA, sessions reset at midnight and on campaign change. In GA4, sessions are defined solely by a 30-minute inactivity timeout. This produces different session counts for the same traffic; do not compare UA and GA4 session numbers directly.
  • Bounce rate replaced by engagement rate: UA's bounce rate (single-page sessions with no interaction) is gone. GA4's engagement rate is the inverse of its bounce rate, but the definitions differ. A UA bounce rate of 70% does not equal a GA4 engagement rate of 30%.
  • Event-based model: UA collected hits (pageviews, events, transactions). GA4 treats everything as an event, including pageviews. This makes custom event configuration more flexible but requires re-learning how to find basic data.
  • Historical data: UA data was deleted by Google in July 2024. If your organisation did not export UA historical data before this date, it is gone. GA4 has no record of pre-July 2023 behaviour (unless you migrated early).
  • Sampling: GA4's standard reports are unsampled for most publishers. Explorations use sampled data for high-traffic properties. If you see “data is sampled” warnings in an Exploration, reduce the date range or apply filters to reduce data volume.

Chartbeat: Real-Time Analytics for Newsrooms

Chartbeat is a paid real-time analytics platform used by many UK national and regional publishers. Unlike GA4 (which processes data with a 24–48 hour delay in standard reports), Chartbeat shows what readers are doing on your site right now, making it the tool of choice for editorial desks managing homepage placement, social distribution, and breaking news.

Chartbeat's core metrics differ from GA4 in important ways:

  • Engaged time: Chartbeat's signature metric. Unlike GA4's session duration (which measures wall-clock time between events), Chartbeat measures active attention — time spent with the browser tab active and the user interacting with the page (scrolling, clicking, or typing). Engaged time is widely considered the most reliable proxy for genuine readership quality. Industry benchmarks vary by content type, but 60–90 seconds of engaged time on a 1,000-word article is considered healthy.
  • Concurrent visitors: The real-time dashboard shows concurrent visitors (readers on your site at that exact moment), broken down by article. This drives homepage placement decisions — if an article is attracting readers rapidly, editors move it higher; if a story stalls, it is replaced.
  • Scroll depth: Chartbeat tracks scroll depth independently of GA4, showing the percentage of readers who reach each quartile of an article. This data is valuable for editorial feedback: if readers consistently abandon after the first quarter of long investigations, the opening may need reworking.
  • Recirculation: The percentage of readers who, after finishing one article, click through to another on your site. High recirculation indicates effective related content recommendations and strong site architecture. Low recirculation suggests readers arrive, consume one piece, and leave — a sign that your linking strategy or content recommendations need work.
  • Traffic sources in real time: Chartbeat shows which referrers are driving current traffic — useful for understanding whether a social post, newsletter, or search result is the source of a traffic spike.

Attribution Models: Last-Click vs Data-Driven

Attribution models determine which traffic source receives credit for a conversion (a subscription sign-up, a registration, or a click on a target action). Understanding the model your analytics uses is essential for making correct editorial and marketing decisions.

  • Last-click attribution (GA4's default for most reports) gives 100% of the credit for a conversion to the final touchpoint before the conversion occurred. If a reader finds an article via Google, returns directly the next day, and subscribes on that second visit, last-click attribution credits “Direct” — obscuring the fact that organic search drove the initial discovery. For newsrooms, last-click attribution systematically undervalues SEO and social media as acquisition channels.
  • Data-driven attribution (available in GA4 for properties with sufficient conversion volume) uses machine learning to distribute credit across all touchpoints in a conversion path in proportion to their contribution. This gives a more accurate picture of which channels genuinely drive subscriber acquisition, though it requires a minimum volume of conversions to work reliably (Google recommends at least 50 conversions per month from a given channel).
  • Understanding traffic sources: GA4's channel grouping combines traffic by type: Organic Search (unpaid Google/Bing results), Paid Search (PPC ads), Organic Social (unpaid social posts), Paid Social, Direct (no referrer or bookmarked), Referral (links from other sites), Email (links from newsletters), and others. Each behaves differently — social traffic tends to spike and fall quickly; organic search traffic is more stable; newsletter traffic arrives in waves corresponding to send times.

ICO Cookie Consent Requirements Under PECR

The Privacy and Electronic Communications Regulations 2003 (PECR), which implement the EU ePrivacy Directive in UK law and are enforced by the Information Commissioner's Office (ICO), require prior, freely given, specific, and informed consent before placing analytics cookies on a user's device. This applies to all UK-based publishers and to publishers targeting UK readers, regardless of where the publisher is based.

The ICO's guidance makes the following clear for analytics tools:

  • GA4 requires consent: Google Analytics 4, when it sets cookies on users' devices, requires consent under PECR. Legitimate interest cannot be used as a basis for analytics cookies under PECR (unlike UK GDPR, where legitimate interest is available for some processing). Consent must be obtained before the GA4 tag fires — not after the page has loaded.
  • Consent must be freely given: Cookie walls — which block access to content unless a user accepts analytics cookies — are not considered freely given consent by the ICO. Readers must be able to access your journalism without accepting tracking.
  • Withdrawing consent: Users must be able to withdraw consent as easily as they gave it. A consent management platform (CMP) must include a clearly accessible method for revoking consent at any time.
  • No pre-ticked boxes: Analytics consent cannot be pre-ticked or bundled with necessary cookies. Analytics must be presented as a separate, optional category in your CMP.
  • Consent records: Publishers must keep records of when and how consent was obtained. Most CMPs provide this automatically. The ICO can request these records during an investigation.

Practical note: Many UK news publishers use a consent management platform (CMP) such as Sourcepoint, OneTrust, or Didomi to manage cookie consent. When implementing Consent Mode v2 with GA4, test that the GA4 tag does not fire before a user has made a consent choice. Use Google Tag Manager's preview mode and the browser's network panel to verify that no GA4 requests are sent on page load before consent.

UK GDPR for Analytics: Lawful Basis, Data Minimisation, and Retention

Even if analytics data does not directly identify individuals, it is personal data under UK GDPR where it can be linked (directly or indirectly) to an identifiable person. GA4 collects IP addresses, device identifiers, and behavioural data that constitute personal data. The following UK GDPR principles apply:

  • Lawful basis: For analytics cookies, the lawful basis must be consent (as required by PECR). For server-side analytics that do not set cookies, legitimate interest may be available — but you must conduct and document a legitimate interest assessment (LIA) demonstrating that your interest in audience insight is not overridden by readers' privacy rights.
  • Data minimisation: GA4 should be configured to collect only the data you need. Enable IP anonymisation (on by default in GA4, but verify in your data stream settings). Avoid sending personally identifiable information (PII) — email addresses, names, subscriber IDs — to GA4 as custom dimensions. Sending PII to GA4 violates Google's Terms of Service and UK GDPR simultaneously.
  • Retention limits: Set GA4 data retention to the shortest period that meets your editorial needs. The default is two months; 14 months is sufficient for most newsroom purposes. Data held beyond your stated retention period without justification breaches the UK GDPR storage limitation principle.
  • Data processor agreements: Google is a data processor for your GA4 data. Ensure you have accepted Google's Data Processing Terms, which establish the legal relationship required by UK GDPR when using a third-party processor.
  • International transfers: GA4 data may be processed in the United States. The UK-US Data Bridge (the UK's equivalent of the EU-US Data Privacy Framework) provides a transfer mechanism for data sent to certified US organisations. Verify that Google is certified under the relevant framework when reviewing your data protection impact assessment (DPIA).

Privacy-Respecting Alternatives: Plausible, Fathom, Matomo

A growing number of UK publishers are replacing or supplementing GA4 with privacy-respecting analytics tools that do not use cookies and do not collect personal data as defined by UK GDPR. These tools offer significant compliance advantages and, for smaller publishers, simpler infrastructure:

  • Plausible Analytics: An open-source, EU-hosted tool that tracks pageviews, referrers, and geography without cookies or persistent identifiers. Because it does not set cookies, PECR consent is not required. Plausible's data is aggregated and cannot be linked to individuals, placing it outside the scope of UK GDPR personal data processing. It offers a script-based integration similar to GA4 and a clean dashboard covering top pages, referrers, countries, and devices.
  • Fathom Analytics: A similar cookie-free tool to Plausible, hosted in the EU and Canada. Fathom processes all data without cookies, is GDPR- and PECR-compliant without a consent banner, and offers conversion tracking (for newsletter sign-ups, subscription clicks, or other goals) without personal data collection.
  • Matomo: An open-source analytics platform that can be self-hosted on your own server, keeping all data under your direct control. Matomo can be configured in “cookie-less” mode (using IP anonymisation and session hashing) which the ICO has acknowledged may not require PECR consent, though publishers should take their own legal advice. The self-hosted version is free; Matomo Cloud is a paid hosted option. Matomo's feature set is closer to GA4 (including ecommerce, goals, heatmaps, and session recordings) than Plausible or Fathom.

For publishers with existing GA4 deployments, running a privacy-first tool alongside GA4 is common practice: the cookie-free tool provides a reliable baseline dataset unaffected by consent refusals, while GA4 provides richer behavioural data for the subset of readers who consent.

Interpreting Scroll Depth, Bounce Rate, and Engagement Rate

Individual metrics rarely tell the full story. The most useful editorial insight comes from combining metrics to understand reader behaviour:

  • Scroll depth + engaged time: An article with high scroll depth but low engaged time suggests readers are skimming (scrolling quickly without pausing to read). This might indicate the writing is too dense, the article is over-long, or readers came with a specific fact-finding intent and left satisfied. An article with low scroll depth but high engaged time suggests readers are absorbed in the opening but not finishing — often a sign the piece is too long for its topic, or that the best content is buried.
  • Engagement rate by traffic source: Organic search traffic typically produces higher engagement rates than social media traffic, because search readers have expressed an intent. Social readers often arrive via a headline that intrigued them and leave quickly if the full article does not match expectations. If your social-sourced engagement rate is very low, review whether your social headlines are accurately representing the content.
  • Bounce rate interpretation: In GA4, the bounce rate (the inverse of engagement rate) is not always a negative signal. A reader who arrives on a brief news bulletin, reads it in full in 20 seconds, and leaves has bounced in the technical sense — but they got what they needed. Context is everything; compare bounce rates within the same content type, not across the whole site.
  • New vs returning readers: GA4's Retention report shows the ratio of new to returning users. A high proportion of returning direct readers indicates a loyal core audience. A site that is overwhelmingly driven by new readers from social media has a fragile audience base dependent on platform algorithms.

Audience Segmentation for Editorial Decisions

Segmenting your audience — dividing readers into meaningful groups based on behaviour, source, geography, or device — allows editorial teams to make better decisions than aggregate data alone permits. In GA4, segments can be applied in Explorations reports to compare behaviour across groups:

  • Loyal readers vs casual readers: Create a segment of users who have visited five or more times in the last 30 days. What content do they consume most? Which sections do they never visit? This tells you what your core audience values, which should influence commissioning priorities.
  • Mobile vs desktop: News site audiences are typically over 70% mobile. If your most important investigations are not optimised for mobile reading (long paragraphs, no subheadings, poor image sizing), mobile engagement rates will lag behind desktop. Compare engagement rate by device category on your top articles.
  • Geography: For regional titles, GA4's geographic reports reveal whether your readership matches your editorial focus area, or whether you are attracting audiences from outside your coverage area that you are not serving with relevant content. For national titles, international readership data can inform distribution and SEO strategy.
  • Newsletter subscribers vs general visitors: If your CMS captures subscriber status and passes it to GA4 as a custom dimension (without PII), you can compare the content engagement of newsletter subscribers against general readers. Subscribers typically show higher engagement rates, greater loyalty, and different content preferences — insight that justifies subscriber-first content strategies.

Practical Checklist

Use this checklist when auditing your analytics setup or reviewing audience data for editorial decisions:

Common Mistakes

  • Treating pageviews as the primary success metric: Pageviews are easily inflated by clickbait headlines, low-quality traffic, and bot activity. Engaged time and scroll depth are far more reliable indicators of genuine readership and editorial impact.
  • Comparing GA4 and UA figures directly: Session definitions, bounce rate calculation, and attribution logic differ substantially between the two platforms. Year-on-year comparisons across the UA-to-GA4 transition are unreliable unless you kept a UA export.
  • Firing GA4 before consent: The most common PECR breach in UK newsrooms. Verify with GTM Preview and browser network tools that no analytics requests fire on page load before a consent choice is made.
  • Using last-click attribution to evaluate newsletter performance: Email typically drives return visits that then convert. Last-click attribution credits “Direct” for these conversions, systematically undervaluing your newsletter as an acquisition and retention channel.
  • Assuming cookie-free tools require no legal review: Plausible and Fathom do not require PECR consent (no cookies, no personal data). However, if you self-host Matomo in cookie-based mode, consent is still required. Always verify the data collection mechanism, not just the brand name.
  • Ignoring the impact of consent refusals on data quality: In markets with high PECR consent refusal rates (sometimes 30–50% for UK news sites), GA4 data significantly underrepresents actual readership. Running a cookie-free tool alongside GA4 provides a more accurate baseline.
  • Making commissioning decisions on a single week's data: Single-article or single-week analytics are heavily influenced by platform algorithm changes, news cycles, and seasonal effects. Review trends over at least four to eight weeks before making structural editorial changes based on analytics.

Red Flags to Watch For

  • GA4 shows unusually high session counts with very low engagement rates — often a sign of bot traffic or a mis-configured internal traffic filter
  • A sudden drop in GA4 sessions with no corresponding drop in Chartbeat concurrent visitors — may indicate a consent mode misconfiguration or tag firing issue
  • PII appearing in GA4 event parameters or custom dimensions in the DebugView or real-time reports
  • Your CMP does not log a consent record for analytics, or the records are incomplete
  • Chartbeat engaged time dropping sharply across the site after a redesign — a signal that the new design is hurting readability or page performance
  • Recirculation rate below 10% — readers are leaving after one article, suggesting related content recommendations are irrelevant or invisible
  • A traffic source that contributes a large share of sessions but very low engagement rate and zero returning users — often a signal of referral spam or paid traffic without editorial value

Jurisdiction note: The regulatory framework described above applies in Great Britain (England, Scotland, and Wales) under UK GDPR and PECR as retained in UK law post-Brexit. Northern Ireland retains alignment with EU GDPR in certain areas under the Windsor Framework. For publishers with significant EU readerships, the EU GDPR and the ePrivacy Directive (as implemented in each member state) apply in parallel; the ICO's guidance does not cover EU obligations. Publishers targeting both UK and EU audiences should take advice on both frameworks separately, as enforcement approaches and consent standards can differ.

Primary Sources

Related guides

BlogPhotography Journalists UKRead →BlogBest Free Tools UK Journalists 2026Read →BlogBudgeting Tax Tips Freelance Journalists UKRead →BlogCopyright Law UK JournalistsRead →BlogDefamation Law UK JournalistsRead →

Related articles

Digital

Social Media Journalism for UK Reporters

Read article →

Investigative

Investigative Journalism Techniques for UK Reporters

Read article →

Legal

Covering Local Government in the UK

Read article →