Agency Said Traffic Is Up But Revenue Is Down

Published Read 6 min read

On this page

“Traffic up, revenue flat” is almost never a single problem. It is two stacked on top of each other: an attribution blind spot that hides organic’s contribution to conversions it assisted but did not close, and a targeting failure where the easiest-to-rank keywords got chased because they were easy, not because they convert. Before you decide the agency is failing, run the diagnosis in a fixed order, because the order is what tells you which of the two problems you actually have, and whether the relationship is fixable or replaceable.

The discipline here is sequencing. Pull the assisted-versus-last-click read first to find hidden value, then segment by landing page to test whether the traffic is even commercial, then ask the incrementality question, and only then judge the agency. Skip a step and you will misread the data: declare organic dead when it is quietly assisting half your conversions, or congratulate an agency for traffic that has no path to revenue.

Step one: read assisted versus last-click before judging anything

Last-click attribution gives all the credit to the final touch before conversion. Organic search frequently opens the journey rather than closing it: someone searches an informational query, lands, leaves, and returns weeks later through a branded search or a paid ad that gets the last-click credit. Under a last-click view, that organic session looks worthless. It was not.

In GA4 the place to look is the Advertising section’s attribution and conversion-paths reports, comparing how organic’s credit shifts between the last-click model and data-driven attribution. Note the current GA4 reality before you instruct anyone: Google renamed “conversions” to “key events” in 2024 (the word “conversion” now refers to the Google Ads side), and in 2024 it removed the first-click, linear, time-decay, and position-based rules-based models from cross-channel reporting. What remains for cross-channel comparison is data-driven attribution and paid-and-organic last-click. So the practical move is to compare organic’s credit under data-driven against its credit under last-click. A large gap between the two is the signal that organic is doing assisting work the last-click number erases.

Read it as a ratio. Suppose organic shows many key events where it participated somewhere in the path but few where it was the last touch. That pattern says organic is top-of-funnel: it opens journeys other channels close. A high assisted-to-last-click ratio means the value is real but mislocated in the report, not absent. The “revenue down” complaint may partly be a measurement artifact, and you cannot tell until you have this number.

Step two: segment organic traffic by landing page

The assisted read can rescue an agency unfairly blamed by last-click. The landing-page segmentation is what convicts one that earned the blame. Split organic sessions by the page they entered on and ask one question: did the traffic grow on commercial, product, and category pages, or did it grow on blog and informational pages while the money pages barely moved?

This is the targeting failure that produces the exact symptom. Informational keywords are higher volume and easier to rank, so an agency measured on traffic will pursue them and the chart will climb. But informational visitors are early-stage and rarely buy on that visit. If the growth is concentrated in blog and guide content while product-page entrances are flat, you have found the real problem: the work optimized for the volume metric, not the revenue one. The traffic is technically up and commercially irrelevant.

Step three: separate participation from causation

Assisted conversions show participation, not causation. They tell you organic was present in the path; they do not prove the conversion would not have happened without it. That distinction matters before you credit organic with revenue in a board conversation, because the obvious objection is “they would have found us anyway.”

The crude read is the branded-versus-non-branded split. Organic conversions concentrated in branded queries mean other channels created the demand and organic merely caught people already looking for you. Non-branded commercial queries are far more defensible as organic-created demand. The rigorous read is a holdout or geo test that suppresses organic in one segment and measures the difference, which is the only method that actually answers the causation question. Most organizations will not run one because it means deliberately turning off a channel. Acknowledging that limit honestly is more persuasive to a skeptical executive than overclaiming, so state which read you used and what it can and cannot prove.

Step four: check whether the content has a conversion path at all

There is a failure mode that no attribution model will surface: traffic that lands, reads, and leaves with no mechanism to capture or convert it. If the informational content has no email capture, no relevant product recommendation, no remarketing tag, and no next step, then even genuinely valuable visitors are unmeasurable brand awareness at best. The agency may have built traffic into a dead end.

This is worth checking directly. Walk the top organic landing pages as a visitor and look for the path: is there anything that moves a reader toward a measurable action, or does the page simply end? If it ends, the fix is conversion architecture, not more traffic, and that reframes the whole conversation.

The root cause is the incentive, not malice

Step back and the pattern is rarely an agency cheating. It is an agency optimizing the scorecard you gave it. A contract or KPI that says “increase organic traffic” gets you traffic, including the cheap informational kind that does not convert. The agency optimized its measured objective; the objective was the wrong one. Diagnose this neutrally, because framing it as dishonesty closes the door on the fix.

The renegotiation is to change the metric. Replace “increase organic traffic” with “increase revenue-attributable organic traffic” and “rankings and traffic to commercial and product pages.” Once the scorecard rewards revenue-relevant outcomes, the strategy follows the scorecard, the same way it did before.

Fixable versus replaceable

Bring the data to the agency: here is organic’s assisted contribution under data-driven attribution, here is the landing-page split showing where the growth actually went, here is the conversion-path gap. Then propose the shift to commercial-keyword targeting and conversion architecture, and watch how they respond.

The response is the real signal. An agency that engages the revenue framing, proposes a commercial-keyword and conversion-path plan, and accepts being measured on revenue-attributable organic is fixable. One that defends “we are building authority” with no articulated path from that authority to revenue, or insists the traffic growth is self-evidently good, is telling you the misalignment is structural. At that point the decision moves from diagnosis to selection, which is its own disciplined process with its own evaluation criteria, and you route it there rather than improvising a replacement.

Frequently Asked Questions

Does “traffic up, revenue down” always mean the agency failed?

No. It can be a measurement artifact: last-click attribution erasing organic’s assisting role. That is why the assisted-versus-last-click read comes first. Only after you confirm the traffic is informational and the money pages are flat have you found an actual targeting failure rather than a reporting one.

Which GA4 report shows organic’s assisted contribution?

Use the attribution and conversion-paths reports under the Advertising section, and compare organic’s credit under data-driven attribution against paid-and-organic last-click. Those two are the cross-channel models GA4 still offers after the 2024 removal of the rules-based models. A wide gap between them is the hidden-value signal.

Should I demand a holdout test to prove organic causation?

Ideally yes, because only a holdout or geo test answers the causation question, but most organizations will not suppress a live channel to run one. The realistic path is to use the branded-versus-non-branded split as a directional read and state its limits openly rather than overclaiming causation you cannot prove.

Sources

Related posts:

  1. Your Breadcrumbs Are Confusing Google’s Site Hierarchy When your BreadcrumbList structured data, your URL path, and your internal-link patterns disagree about a page's parent, Google receives contradictory signals about where the page sits in your site and......
  2. Your Contact Page Ranks Higher Than Your Service Pages A contact page outranks your service pages for the queries you actually care about when two things stack: it carries more internal-link weight than the service pages, and it contains......
  3. Your Mega Menu Is Leaking PageRank to Low-Value Pages A sitewide mega menu and a bloated footer put the same large set of links on every page of your site, so the link equity and crawl signaling flowing out......
  4. Why Changing Your CMS Broke Links You Never Touched A platform migration changes URL structure in ways you did not plan, and that is why URLs you thought you preserved now return 404. Trailing-slash handling, default permalink and path......
  5. Your Chatbot Widget Is Injecting Thousands of Hidden Links A third-party widget (live chat, social embed, review badge, support tool) loads its own markup into your DOM, and that markup often includes links to the vendor and to the......
  6. Google Demoted Your Page for a Schema Error You Can’t Find Your schema passes every validator, the Rich Results Test reports no errors, and yet the rich results are gone and there is no manual action in Search Console to appeal.......
  7. Why Removing Dates from URLs Killed Your Traffic Stripping the date segment out of your URLs is a full URL migration, and the traffic drop almost never comes from "losing the date" itself. It comes from three stacked......
  8. Why Query Parameters After Hash Symbols Don’t Exist to Google Everything after the # in a URL is a fragment, and the browser never sends it to the server. So when your filter or state lives after the hash, /properties#bedrooms=3......
  9. Why Your Staging Site Is Ranking Instead of Production Staging out-ranking production is a canonical signal inversion, not a simple case of an exposed test server, and treating it as the latter is why the usual fixes fail. Production,......
  10. How to Reverse Engineer Google’s Quality Raters Guidelines for Your Site The Search Quality Rater Guidelines are a public document describing exactly what Google trains its algorithms to detect, which makes them a usable blueprint for self-auditing. You reverse-engineer quality by......
  11. Neighborhood-Level SEO: Why “Local” Isn’t Local Enough Anymore TL;DR Local SEO strategies treating entire metro areas as single markets are leaving rankings on the table. Search behavior varies dramatically between neighborhoods within the same…...
  12. Informational SEO is Dead. Here’s What’s Left. Publishing top-of-funnel informational content primarily to capture search clicks has lost its economics. AI Overviews now answer "what is" and "how to" queries directly inside the results page, so the......
  13. Edge SEO: Implementing Changes at the CDN Layer Edge SEO ships header, redirect, hreflang, schema, and meta changes in hours by running serverless workers at the CDN layer, instead of waiting weeks for an engineering backlog to clear.......
  14. Redirects Are Working But Traffic Still Dropped A redirect that verifies clean in a crawler proves one thing only: the hop resolves with a 301 and lands on a live page. It does not prove that the......
  15. Google Keeps Choosing the Wrong Canonical A canonical tag is a hint, not a directive, and Google says so directly: it takes your declared canonical into consideration, but it can and does select a different URL......