In 28 days, this site went from 20 to 161 clicks coming from Google. Impressions went from 1,884 to 12,268, and average position dropped from 18.5 to 10.8 — on Google, lower is better.
An honest comparison between two 28-day blocks, measured in Search Console: June 25 to July 22 versus July 23 to August 19, 2026. Not a single cent spent on ads.
This post is an account of what we did, including what didn't work — because the part that didn't work usually teaches more than the part that did.
What worked
1. Writing for a specific question, not a topic
"Augmented reality" is a topic. "Why does my 3D model appear white in augmented reality" is a question. The second one has people searching for it and has a verifiable answer; the first competes with the whole world and tells no one anything.
2. Translating everything into four languages
This was the least glamorous and most efficient multiplier. A concrete example: the Spanish version of our post about the 3D viewer brought 24 clicks in the period, compared to 18 for the Portuguese version and 7 for the English one. The same research and writing work, distributed across four markets.
3. A supporting post for each tool
We have two free tools on the site. One of their pages had two impressions in the month — practically invisible. The post explaining that same tool became the site's biggest entry point: 49 of the 161 clicks, combining all languages.
The lesson, which applies to any business: a product page doesn't rank on its own. What ranks is the text that teaches — and it's that text that leads the person to the product. The pair that works is post → tool → service, in that order.
4. Real frequently asked questions, properly marked up
Each post ended with real questions and objective answers, marked up in the code so that search engines and AI assistants can display them. It's the format AIs cite most, because it already comes in a short, verifiable block.
5. Fixing the invisible basics
We found 17 pages whose main title, in Google's eyes, was a gray 12-pixel label — the large text the reader saw was marked up as a plain paragraph. A fix with no visual effect at all, and one of the most profitable of the period.
What did NOT work
A free tool, on its own, brings no one
We published a free 3D file converter, useful and with no sign-up required. Impressions in the period: zero. A good tool is not content — no one searches Google for something they don't know exists.
Traffic is not a lead
This is the uncomfortable figure: 12,268 impressions and 161 clicks yielded a single real contact through the form. One. Everyone else the site received came by, read, and left without a trace.
The conclusion isn't that SEO doesn't work — it's that growing traffic and capturing contacts are two different jobs, and we only did the first one. Anyone with a download, a calculator, or gated content who isn't asking for an email in exchange is making the same mistake.
A topic can inflate impressions without generating clicks
One of our posts accumulated 2,254 impressions and zero clicks at position 8. On investigation, the cause was something else: Google was still showing the text's old URL, with the old title, which didn't promise the answer the person was searching for. Impressions without clicks almost never mean "Google doesn't like me" — it's usually a title that doesn't match the search.
Internal linking gets forgotten exactly where it matters
We discovered that the page with the most impressions on the site received not a single link from the other 240 posts. Orphaned content takes much longer to consolidate.
What we started measuring
- Clicks before impressions. Impressions inflate easily; a click is a choice. In our case clicks grew more than impressions (+705% versus +551%), and that's what indicates the content is truly answering the question.
- Average position by commercial term, separated from informational terms. These are different audiences, and one hides the other in the average.
- Contacts, not visits. After this review, it became a priority — because it was the number that exposed the flaw.
The summary, if your site is stalled
Publish content that answers specific questions; translate it, because the marginal cost is low and the reach doubles; create the post that explains each thing you offer; fix the technical basics no one sees; and — the part we got wrong — build the capture mechanism before bringing in the traffic, not after.
If you want to understand how to appear in AI answers too, not just on Google, start with AEO: how to get ChatGPT and AIs to recommend your business. And if you'd like this work running on your own site, that's what we do in website creation and digital marketing.
All figures came from the Google Search Console for agenciaprimeirapagina.com.br, comparing 06/25–07/22 and 07/23–08/19, 2026.
Perguntas frequentes
How long does it take for a site to start getting clicks from Google?
In our case the difference showed up when comparing two 28-day blocks: clicks went from 20 to 161 and impressions from 1,884 to 12,268. Worth noting that this followed months of continuous publishing; new content usually takes weeks to settle into position.
Is translating the blog worth it for SEO?
In our case it was, and it proved the most efficient multiplier. The Spanish version of one post brought 24 clicks over the period, against 18 for the Portuguese version and 7 for the English one — the same research and writing, spread across four markets.
Why doesn't my product page show up on Google?
Because a product page rarely ranks on its own: people do not search for what they do not know exists. What ranks is the content that teaches and leads to the product. On our site, one tool's page got two impressions in a month, while the post explaining that same tool brought 49 of the 161 clicks.
I have plenty of impressions and no clicks. What should I do?
It is almost always that the title and description shown in search do not promise the answer the person is looking for, rather than a penalty from Google. Check whether the URL Google shows is the current one and whether the title matches exactly the question driving those impressions.
Does growing traffic automatically bring customers?
No. In our case, 12,268 impressions and 161 clicks produced a single real enquiry through the form. Driving traffic and capturing contacts are different jobs: without a clear reason to leave an email, most visits pass through and leave.
What should I measure first in Search Console?
Clicks before impressions, because impressions inflate easily while a click is a choice. Then average position, separating commercial from informational terms, since one hides the other in the average. And finally enquiries generated — the number that shows whether traffic became business.