SEO Tools With Causality Engine: What a Ranking Report Leaves Out

The short answer: when the SEO dashboard says organic traffic rose by a fifth and the store's orders did not move, the dashboard is usually not lying. It is reporting a different kind of number. Semrush calls its Traffic figure "estimated site visits based on the domain's organic search positions" (positions report guide), and most tools below build their headline the same way, from positions, search volume and an assumed click rate. None of them counts an order. What organic search caused in sales sits in the store's GA4 export, which Causality Engine reads, with a free first finding at app.causalityengine.ai/start.

Disclosure: I founded Causality Engine and own it, and this page recommends it. Every claim about it links to Causality Engine's own pages.

Before the SEO number goes into a budget meeting:

1. Do the click arithmetic yourself. Morningscore publishes its table: position 1 gets 15.00% of a keyword's search volume as traffic, position 10 gets 1.00% (traffic article). A 1,000-search keyword climbing from 10th to 1st adds 150 minus 10, so 140 estimated visits, whether anyone buys or not.
2. Open the Search Console you already have. Its clicks are counted by Google, not modelled by a vendor.
3. Split brand from non-brand. Wins on the store's own name are organic visits, but the demand existed already.
4. Put GA4's Organic Search purchases and the store's orders beside the dashboard, same dates, before reading any trend.

The number on each dashboard, and where it comes from

Each row was read on the vendor's own page on 11 November 2026, in the vendor's own term for the metric.

Suites and research tools

Rank trackers

Content optimisers and AI visibility

WordPress plugins

So three kinds of number: an estimate of visits (position, volume, click table), a position or visibility share (the same inputs before the last multiplication), and a score that grades a page against a checklist or against pages that already rank. The only real counts of people are the Search Console figures the plugins and connected suites display, and those are Google's numbers shown inside another tool. Frase says as much: without Search Console it "can't see how anything you've already published is doing" (Connect Your Data).

Search Atlas is the plainest about the gap. Its help centre shows a homepage estimate of 7 beside 58 Search Console clicks over 28 days and calls Search Console the source of truth for your own clicks (help article). Seobility adds a warning of its own: adding monitored keywords that rank raises the visibility value (FAQ), so a rising line can mean a longer keyword list.

How search visits land in GA4

None of these tools tags a link. A visitor from a non-ad search result lands in Organic Search because the source matches Google's list of search sites, and Google's default channel group page now counts AI Overviews and AI Mode there too.

Search Console clicks and GA4 sessions are counted differently, and Google says so. Its guide to using the two together lists why: GA4 depends on its tag and on consent, Search Console reports in Pacific Time, Search Console counts every click while GA4 applies an attribution model, and Search Console groups data under canonical URLs. Linking them in GA4 adds a Google Organic Search Traffic report by landing page, with up to 16 months of data (Connect Search Console).

Brand needs its own line. Semrush splits branded from non-branded traffic in its overview; GA4 does not, and a visit from someone typing the store's name into Google is Organic Search all the same. The branded search incrementality guide covers the paid side of that question.

What the causal read sees, and what it cannot

Causality Engine takes one file, the GA4 Attribution paths export covering 12 months or more (how it works). In that file Organic Search is a single channel. The read cannot see a keyword, a page, a position or a score.

What it adds is at channel level: per the same page, it splits Direct back to the channels that sent those buyers, separates what each channel caused from what it only touched, and sets each one beside a last-click rebuild from the same file, with a data-health score from 0 to 100 and a next step tested in one region first. The integrations page says it measures every channel visible in GA4, organic included. It has no connection to any tool in the tables.

A worked example: estimated traffic up 18.4%, orders flat

For example, with every number invented for the arithmetic and no store behind it, a homeware shop compares September and October:

• The suite's estimated organic traffic goes from 41,300 to 48,900: 7,600 ÷ 41,300 = 18.4%. Traffic cost goes from $27,450 to $32,180.
• Search Console clicks go from 19,870 to 20,310, up 440, or 2.2%. Branded queries account for 380 of the 440, so non-branded clicks rose by 60.
• GA4 Organic Search sessions go from 22,140 to 23,060: 920 ÷ 22,140 = 4.2%.
• GA4 last-click purchases from Organic Search go from 412 to 418, which is 1.5%.
• The store books 3,152 orders, then 3,184.

The suite moved most because it multiplies new positions by volume and a click table; a few high-volume keywords climbing from 12th to 6th will do that with few extra visits. The $32,180 is what those estimated visits would cost as Google Ads clicks. It is not revenue, and nobody saved it.

Now October's ad claims: Meta 1,920 orders, Google Ads 1,240. 1,920 + 1,240 = 3,160 of 3,184 booked, leaving 24 for everything else. Add GA4's 418 for Organic Search: 3,578 claimed, 394 more than the store took in. No SEO tool is in that overlap, because none claims a sale; the overlap belongs to the ad platforms and to last click. A causal read would set Organic Search's caused share beside its 418 last-click purchases and show how much of Direct came from search. This page does not guess what it would find. Where the read and the ranking report disagree, change one thing in one region before moving budget.

The free check, and why it is weaker for search

For most channels the clean test is a holdout. Search has none: nobody can stop ranking for a tenth of visitors. These by-hand checks cost nothing and show disagreement and direction, not cause.

• Clicks, sessions, orders by landing page. Search Console clicks to product and category pages, GA4 Organic Search sessions and purchases for the same landing pages and dates, then the store's orders for those products. Match the time zone first.
• Branded against non-branded. Filter Search Console queries containing the store's name and misspellings, and read the two trends apart.
• Before and after a change. Compare the four weeks either side of a rewrite or fix. Season, promotions and other channels move in the same weeks, so label it weaker than a holdout.

Then the paid path. Export the Attribution paths report (how to export GA4 data) and drop it on the start page, where the first finding is worked out in the browser for free. Causality Engine, which I founded, charges €99 once, excluding VAT, for the full read, refundable within 30 days; Pro is €299 a month (pricing, checked on 11 November 2026).

Where the read stops

• It needs a year. The how-it-works page asks for 12 months or more of GA4 data.
• It needs a signal. Where spend is too small for one, the homepage says it declines.
• It stops at the channel. Which keyword or page earned the visits stays with the SEO tool and Search Console, which also find keywords, crawl sites and grade drafts, none of which the read does.

The plans are sold by the same unit the chart counts. Semrush's SEO plan is $139 a month for 500 keywords tracked daily (pricing), and Nightwatch tells buyers to "pay for keywords tracked", Starter being €79 a month billed yearly for 500 keywords (pricing), both checked on 11 November 2026. A visibility line that climbs as keywords are added is the line that makes the next allowance look necessary.

See also

• Causality Engine: the read, its prices and its limits
• How to use Causality Engine: from a GA4 export to a regional test
• Causality Engine on the marketplace: its affiliate terms, stated as the owner's
• Every worked example, by tool: the other tools set beside the read
• How to find what Direct traffic is hiding: where unattributed buyers came from
• Semrush vs Ahrefs: two suites and their traffic estimates
• Similarweb vs Semrush: panel data against keyword-based estimates
• Frase vs Surfer: two content scores compared

FAQ

Does any of these SEO tools connect to Causality Engine?

No. None passes data to it, and it needs none of them: the read works from the GA4 Attribution paths export, where search visits already sit as Organic Search (integrations).

Why does the tool's traffic differ from Search Console clicks?

One is modelled and one is counted. The suites multiply search volume by a click rate per position; Search Console reports the clicks Google recorded. Search Atlas tells users to compare its estimate with clicks, not impressions (help article).

Does a higher Content Score mean more organic sales?

The vendors do not claim it. Surfer, Frase and NeuronWriter describe their scores as fit with competing pages or on-page practice (Surfer, Frase, NeuronWriter). Whether a page sells shows in its Search Console clicks, its GA4 purchases as a landing page, and the store's orders.