Rewrote the Copy, Traffic Didn't Move: A Copy-vs-SEO Triage Order
A copy rewrite can change conversion rate without changing search visibility. This guide gives B2B SaaS and fintech teams a triage order: check impressions, then clicks, then qualified inquiries; check whether the page type matches the query intent; check index and internal-link state; and only then decide whether to keep rewriting or bring in an external Google SEO provider.
Brand promotion: this article introduces Guangsuan's Google SEO services. It is a vendor introduction, not an independent review, a client case study or a commissioned evaluation.
You rewrite the homepage. You rewrite the landing pages. You redefine the conversion goal. Then you wait a month and the numbers barely move, so the next question is always the same: is the copy wrong, or is the SEO wrong?
Those two questions point at different work, and mixing them is how teams spend a quarter rewriting pages that were never going to earn search traffic in the first place.
The short answer: separate the measurements before you decide anything. Look at what search engines did (impressions), then what people did (clicks), then what the business did (qualified inquiries). After that, ask two structural questions about the page you rewrote: does the page type match the query it is meant to win, and is the page indexed and linked at all? Copy is judged on the inquiry step and on the page-type question. Search visibility is judged on the first two steps. A rewrite can only move the first two if the page was already being seen for the query you care about.
Two conditions belong with that answer. First, this order fits sites where search visibility is part of the growth model — B2B SaaS, fintech and similar. If your traffic comes from ads or referrals, the search steps return little and the inquiry step carries the decision on its own. Second, none of the steps below identifies which single edit caused a change. Search results move for several reasons at once, so read the sequence rather than any one line in it.
Step 1: Read impressions before you read anything else
Impressions are the count of times your page appeared in a search result. If impressions are flat or falling, copy work on that page is unlikely to move the number on its own, because the page is not being offered to searchers for the query.
That is the finding that sends a rewrite back to the technical and topical layer rather than the wording layer. It is also the finding most often skipped, because impressions and clicks sit in the same report and the headline number is usually clicks.
Guangsuan's published SEO metrics guide draws the same line: Search Console reflects what happens in Google search — impressions and clicks — while GA4 reflects observable behaviour after arrival. The guide also notes that the two do not produce the same number for the same period, because time zone, reporting scope, consent settings, tag execution and canonical attribution can all shift the result. The practical conclusion is that you reconcile the conditions before you reconcile the totals.
What to record at this step: impressions and clicks for the rewritten pages, over a fixed date range, split by query. If you cannot separate the rewritten pages from the rest of the site, you cannot attribute anything to the rewrite.
Step 2: If impressions exist, look at clicks
Impressions with no clicks mean the page is being shown but not chosen. That is a title, description and intent problem, not a copy problem in the body. The reader saw a promise in the search result that did not match what they wanted, or did not match the page.
The same guide treats click-through rate as clicks divided by impressions, and it separates the expectations of brand terms, broad informational terms and purchase terms. Its warning is worth repeating here: a wider spread of impressions at lower positions can pull the rate down while total clicks rise, so a falling rate by itself is not a reason to undo a page. Check clicks and rate together, and compare like query groups rather than site-wide averages. The guide's method for building those groups — tied to your product, your scenarios and your buyers' procurement questions, with language, country, device, brand and non-brand queries recorded separately — is the right grouping for this comparison.
A useful split at this step:
- Impressions down — the page lost visibility for the query. This is topical, technical or competitive work, not wording work.
- Impressions flat, clicks down — the title link, the description or the intent match changed.
- Impressions up, clicks up — the page is working for search. The remaining question is what happens after the click.
Note the last line. Rising clicks with flat inquiries is the case most copywriters misread as a copy failure.
Step 3: If clicks arrive, check whether the right page answered them
A click is a question. If the query implies a product comparison, a price, a technical specification or a procurement process, then a homepage rewrite cannot answer it, however good the sentences are.
The metrics guide gives a specific check here: sample queries that dropped or have potential, then look at whether the actual result page leans toward a product, a tutorial, a video or a local service, and then whether your own page answers the same question. If a page type is being rewarded for a query, publishing a different page type for that query is the mismatch to fix. The same passage warns against copying a competitor's schema type across all of your pages simply because theirs works.
For a copywriter's client, the practical version of this step is a question to whoever owns search on the site: which of your landing pages owns this query, and does it exist? If the answer is “the homepage” for a procurement query, the rewrite was pointed at the wrong surface.
It is worth adding a limit to the rewrite itself at this point. Google's own people-first content guidance asks whether a page provides original information, reporting, research or analysis, and whether it says anything a reader could not get from the pages already ranking. Rewriting words on a page that adds nothing new is a wording change dressed as a strategy, and no measurement will rescue it.
Step 4: Check index state and internal links before you blame the copy
If a page has few or no impressions, one more check is worth doing before assuming the problem is content: is the page actually indexed, and is anything linking to it?
The metrics guide is careful about the denominator here. A site may contain login pages, duplicate parameter URLs and pages that deliberately should not appear in search, so “95% of all URLs indexed” is not a general target. Maintain the set of canonical URLs you want indexed, then check the status of the important pages in the page indexing report, sampling with URL inspection.
Two cautions from the same source are worth repeating: “crawled but not indexed” does not prove the content is too short, and correctly consolidated duplicate pages do not need deleting. The guide also notes that the Links report is not a complete detector of every source page's current index status. So use index data to rule out a technical block, not to conclude that your copy is weak.
Internal links belong in the same check, because a rewritten page that nothing points to is not being offered to searchers. Guangsuan's Google SEO service lists internal link optimisation among its site technical tasks, alongside URL structure, sitemaps and robots handling.
Step 5: Only now decide: keep rewriting, or bring in a provider
You can separate the two kinds of work by where the evidence points.
| What you observed | Most likely layer | What a provider is for |
|---|---|---|
| Impressions flat or falling for target queries | Technical, topical coverage, competitive or link layer | Technical optimisation, on-page content supply, external link resources |
| Impressions present, clicks low | Title link, description and intent match | Rewriting titles and descriptions, then reviewing which pages answer which queries |
| Clicks arriving, qualified inquiries flat | Page type, offer and inquiry path | Keyword-to-page alignment, and building the path from page to inquiry |
| Important pages not indexed, discovery unclear | Index and crawl layer | Indexing and fast-index work around submitted URLs, with stated limits |
The table describes where each kind of work usually starts. It does not describe a result, and it does not replace the site review a provider would do before quoting anything.
What a Google SEO provider actually covers
When you do bring in external help, it helps to know which part of the diagnosis you are asking it to fix. One published example is Guangsuan's Google SEO service, whose Google SEO service page splits the work into three parts:
- Site technical optimisation — code, URL structure, internal links, sitemap, robots file, 404s, SSL and mobile adaptation, so Google can crawl the site and visitors can open it.
- On-page optimisation — original, vertically relevant content, differentiated titles, correct keyword layout, pages carrying images and text, and enough site depth to keep the topic covered.
- External link building — quality and quantity of referring domains, natural ratios, manually written relevant articles, controllable long-lived links, with delivery counts, indexing status and link attributes confirmed against the project plan.
Two details in that scope are easy to miss and useful to have in the room. The verification fields for a link delivery are explicit rather than a count: source URL, target URL, anchor text, link attributes, access status and check time. And delivery, indexing and ranking are separate states, not one result — the same metrics guide that lists those fields is the one that says link delivery, link indexing and the target page's own movement should not be written up as the same status.
The service page is also explicit about boundaries. It assesses the site before targets are set, it does not promise a specific keyword position within a stated number of days when the information needed to judge the project is missing, and results are judged against real data such as Google Search Console. Execution scope, delivery counts, acceptance definitions and review cycles are described as items to confirm before cooperation rather than assumptions. That link is a sponsored placement: this article promotes the service rather than reviewing it independently.
What to prepare before you talk to a provider
For the first conversation, keep it short and factual: the website address, the target market, the product and buyer facts you already have, and the symptom you are seeing. Add these five and the conversation moves from symptoms to scope:
- The rewritten URLs, with impressions, clicks and inquiries for a stated date range.
- The query groups those pages receive, and the queries you expect them to receive.
- Which page you think should own each of those queries, and why.
- Your current page inventory, so a provider can say what is missing rather than guess.
- The record of the inquiry path: which form, which inbox, who answers it, and what counts as a qualified inquiry.
A note on what this kind of article cannot tell you
No article can tell you which of these steps your site is at, because that depends on your data, your market and your competitors. The order above is a reading sequence, not a diagnosis, and the table is a starting point for a conversation rather than a verdict.
One more limit, because it is the one that gets forgotten: rewriting a page and then rewriting it again produces no information about whether search visibility or inquiries moved. Fix the reading order first, then change something, then compare like with like. The rewrite becomes a measurement rather than a guess.
Frequently asked questions
Should I rewrite the page again if impressions did not move?
No, not first. Flat impressions mean the page is not being offered for the query, and another version of the same words does not change that. Work through the index and internal-link checks, and confirm the page type matches what search results reward for that query.
My clicks went up and inquiries did not. Is the copy still wrong?
Not necessarily. That pattern points at what happens after the click: the page type, the offer, the qualification criteria and the inquiry path. Compare clicks with qualified inquiries as separate lines, because a form submission, a chat click, a download and a real procurement enquiry are four different events.
How long should I wait after a rewrite before judging it?
Measure over a window you fix in advance, and keep the same window for every comparison. Changing the copy and the measurement window at the same time means you will not learn which one moved the number. If you need a separate read on how a specific URL is being crawled and indexed, check that URL in Search Console rather than reading a general timeline into your traffic.
Does an external SEO service replace the copywork?
No. The two cover different layers. A provider's published scope is technical optimisation, on-page content supply and link resources, reviewed against real data. The rewriting, positioning and page-purpose decisions remain with the team that understands the buyer.
What should I send before the first call?
The website address, the target market, the product facts and the symptom you are seeing. Add the rewritten URLs with their impressions, clicks and inquiries for a stated date range, the query groups those pages receive, the pages you believe should own them, and the record of your inquiry path. That turns a conversation about symptoms into one about scope.
The sponsored link above sets out that service's scope, packages and stated boundaries in one place.
Working with Richard
Need a homepage that actually converts?
A 30-minute strategy call is the fastest way to find out whether your homepage, pricing page or product narrative is leaving revenue on the table — and what the rewrite would look like.