The Edinburgh analytics problem: why GA4 undercounts your Festival season by half

Why Google Analytics 4 no longer answers Edinburgh SEO questions

There is a well-worn refrain in Edinburgh marketing circles about GA4: half of what used to matter for organic search has quietly disappeared, and the other half is now expressed in a dialect no in-house SEO speaks fluently. When Google closed Universal Analytics in July 2023, the assumption in most Charlotte Square boardrooms was that the reporting interface had merely been repainted. Two years on, the pattern is unmistakable and, one might argue, was there to be read from the outset: GA4 was never designed to answer SEO questions in the first place. It was built to log product events and simplify remarketing to Google Ads. Everything that used to matter for organic reach — traffic per landing page, organic keywords, behaviour by source — is either broken, hidden behind an anonymisation threshold, or gone altogether. For a Fringe venue selling last-minute tickets, a Blackford tutoring startup, or a Baillie Gifford-adjacent research shop trying to prove that its content programme is worth renewing, that is not a minor inconvenience; it is the difference between a signed retainer in September and a polite email the following spring.

Ask any in-house SEO working for an Edinburgh scale-up such as a FanDuel spin-out, a Skyscanner-adjacent affiliate, a Cirrus Logic supplier or a FreeAgent alternative. Open GA4, drill into Acquisition → Sessions, and more than half your organic queries land as the useless placeholder (not provided). Cross-check against Search Console and the totals never match. Try to see what someone who arrived on the article “best dentist Bruntsfield” actually did — and GA4 hands you an aggregated funnel that will not separate channels cleanly. A monthly SEO report that used to take forty-five minutes now eats half a day, and by the time the Wednesday client call starts nobody at the table trusts the numbers.

Then there is the legal angle. The Information Commissioner’s Office (ICO) has not gone as far as CNIL in France or the Austrian DPA in ruling GA4 illegal outright, but its 2024 guidance on international transfers and its ongoing enforcement of PECR against sites that fire analytics before consent have made the “install GA4 and forget it” playbook indefensible. For a New Town marketing consultancy or a Leith design studio, this means either you invest in a properly configured consent-management platform, or you accept that every visitor who taps “Reject All” immediately disappears from your reports.

What Semalt Analytics is and where the data comes from

Semalt Analytics is a data layer purpose-built for SEO, resting on three sources that reinforce one another:

1. Google Search Console API v3 — the newer endpoint that returns up to sixteen months of history with daily granularity, well past the top-1,000 query cap of the UI.
2. Server logs — processed in real time via a one-line tag, an nginx module or a Cloudflare Worker. This captures sessions GA4 misses because of blockers, rejected consent or private browsing on iOS 17.
3. Semalt’s own crawl — the rank-tracking engine cross-references your traffic, so every organic hit is tagged with the exact SERP position and features present at click time.

That interlock is what nothing else on the UK market delivers in the same form. GA4 tells you how many sessions arrived. Search Console tells you the average position of a keyword. Semalt tells you, for each individual session, which position the query held at that exact minute, which SERP feature stood above it (People Also Ask, video carousel, Google Business Profile pack), and how that affected real CTR by device and by day of week — a Saturday morning ahead of a Six Nations match at Murrayfield looks nothing like a wet Tuesday in February.

Five things you see in Semalt and never in GA4

1. Real organic traffic per URL, without sampling

GA4 begins sampling the moment you cross two dimensions and pass ten million events per month. For a mid-sized Edinburgh online shop — even one with modest traffic — that threshold is easy to trip once you count scroll, view_item and outbound-click events. Semalt applies no sampling. Every URL runs through a column-oriented index on ClickHouse that resolves queries such as “give me the five hundred landing pages with the most organic traffic in July, compared with last July, filtered to mobile only on EE and Virgin Media O2” in well under a second.

2. Keywords without (not provided)

The infamous (not provided) does not exist inside Semalt, because the sources are raw Search Console data plus the platform’s own crawl. For every URL you see the fifty keywords driving the most traffic with clicks, impressions, CTR and average position. More importantly, you see the keywords sitting in positions four to ten — the ones a content refresh can push into the top three and roughly double the traffic. This view is called Striking Distance and it is the first thing we open when auditing a new Edinburgh client. If you want to see it for yourself, sign in to the Semalt dashboard and connect Search Console.

3. Sessions GA4 loses to blockers

In the UK, desktop ad-blocker adoption sits at roughly 34% according to GlobalWebIndex Q2 2026 data, and for under-35s on EE or O2 UK the figure climbs to 43%. In Scotland, uSwitch surveys put Edinburgh a couple of points above the UK average, thanks to a technically literate audience across the universities and the fintech cluster on Lothian Road. Every one of those sessions is invisible to GA4. Semalt’s log-based tracker sees them because it does not depend on JavaScript: the second the server serves the page, the session is counted. Across our Edinburgh client base we have measured the gap between 18% and 63% additional sessions, depending on vertical — higher in fintech, lower in traditional bricks-and-mortar retail on George Street or Multrees Walk.

4. Cross-device journey without Google Signals

GA4 promises cross-device sessions via Google Signals, but only for users signed in to Google on both devices and who have accepted personalised advertising. According to the IAB UK 2025 consent tracker, only 41% of British users click Accept on a standard TCF v2.2 banner. Semalt uses a technique called server-side fingerprint fusion that respects UK GDPR (no ad profiles, no cross-site behavioural graphs) and still shows that a visitor read your article on mobile on a ScotRail journey home from Waverley and completed the booking on desktop later that evening.

5. Organic revenue per keyword, in real GBP

The crown jewel: connect your CRM or shop (WooCommerce, Shopify, Magento 2, BigCommerce, HubSpot, Pipedrive) via webhook and Semalt links every order to the first organic keyword that brought the visitor. You can see that “independent tax adviser Stockbridge” produced £19,700 in organic revenue last month, while a fatter head keyword generated £0. With that in hand, you stop chasing vanity metrics and prioritise content by actual GBP revenue.

Case study: Leith craft brewery, 22,400 sessions/month

A client running a Leith craft brewery with a taproom on Constitution Street and national online distribution — the kind of independent Scottish operator with monthly beer subscriptions, event covers and a small trade wholesale line — came to us in February 2026 with this picture:

• GA4 reported 22,400 organic sessions and 71 online orders for the month (0.32% CVR).
• Search Console showed 49,800 clicks in the same window.
• The gap — more than double — had no internal explanation. The team blamed GTM, Cloudflare’s bot fight mode and the CookieYes banner in rotation across the checkout.

We connected Semalt Analytics in thirteen minutes. First numbers landed by breakfast the next morning:

Real organic sessions: 41,300 (84% more than GA4). Orders correctly attributed to first organic touch: 192 (2.7×). Verified organic pipeline value based on average basket and subscription lifetime: £198,000 vs. the £34,000 the founders had been estimating — a difference of £164,000 that was invisible in the monthly report.

The most valuable insight arrived when we cross-referenced keywords with pipeline value: eight long-tail terms (each between 110 and 590 searches/month, all mentioning Leith, specific hop varieties or the wedding-taproom hire angle) drove 68% of pipeline, while the head keyword “craft beer Edinburgh” with 8,100 searches/month drove less than 4% of orders. Content strategy was re-prioritised inside a fortnight. By May, organic pipeline was up 79% and the co-founder quietly told us they could finally justify hiring a second sales manager.

UK GDPR, PECR and the ICO position on GA4

The reason many Edinburgh and Glasgow agencies switch to Semalt is legal defensibility. PECR regulation 6 requires prior consent for any storage of information on, or access to information from, a user’s device — cookies, localStorage, IndexedDB and similar. Server-log tracking touches none of that. It uses only the information the browser sends anyway as part of the HTTP request (IP, user-agent, referrer). Those are processed server-side under Article 6(1)(f) UK GDPR, with the IP truncated within milliseconds (last octet removed) and never persisted.

That gives Edinburgh marketers a rare advantage: Semalt Analytics counts users who reject the banner too. For a mid-sized Scottish shop with a 45% consent rate, that alone means a 122% more accurate reach measurement, without straying into ICO enforcement territory. Semalt provides a UK-law DPA that can be signed with one click per account, plus a TOM document aligned with ICO’s 2025 accountability framework and NCSC hardening guidance — the last one is what tips the scales for Scottish Government and NHS Scotland supplier reviews.

Setup in twelve minutes

Semalt Analytics vs. GA4 vs. Matomo vs. Search Console

FAQ

Does Semalt replace GA4 completely?

If your focus is purely SEO and content, yes. If you also run Google Ads and need to export audiences to Google, keep GA4 in parallel — Semalt does not replace the native Ads connector. For eight in ten Edinburgh clients we keep GA4 as an Ads source and use Semalt as the SEO source of truth.

Is it compliant with UK GDPR and PECR?

Yes. Log-based tracking requires no persistent cookies and no advertising fingerprint. Server-side fusion works with pseudonymised data, and sessions anonymise after 26 months by default (configurable down to six). Semalt provides a UK-law DPA, a TOM document and a DPIA template aligned with ICO guidance — the same document pack we ship to Edinburgh University microsites.

Does it run on typical UK stacks?

Yes. The tag exists as a WordPress plugin, a Shopify app, an Umbraco package (still popular with Scottish councils and NHS boards), a Magento 2 module and SDKs for Node.js, PHP, Symfony, Laravel and .NET. For cloud-native setups it’s two lines in a Cloudflare Worker deployed from your Fasthosts or Krystal environment. Log in and pick your platform.

How fast do you see value?

Historical GSC data is visible in five minutes. Live tracking shows full data after 24 hours. The Striking Distance insight (keywords in positions four to ten) is actionable on day one — useful when Fringe season starts and you have a fortnight to rank a new listing.

Does it work for small websites?

Yes. A Morningside personal blog with 1,500 sessions/month uses Semalt the same way a B2B portal with two million does. For small sites the value is in seeing which article is about to reach page one and optimising it before Google cements the position.

Does it integrate with Semalt AutoSEO?

Yes — that’s the real synergy. Analytics identifies URLs with striking-distance keywords, and AutoSEO executes the fixes on those URLs without a developer in the loop. Details in the AutoSEO guide.