FloxoLab

@floxolab.com

I do technical SEO and automate the grunt work. Crawlers, GSC, Apps Script, n8n. Code included. → https://floxolab.com/blog

Automation check 01: check the HTTP status before rebuilding the workflow. A 404, 401, 405, and 500 point to four different problems. “Webhook broken” tells you nothing.

An n8n workflow can work locally and still advertise the wrong public webhook URL. Behind Cloudflare or another reverse proxy, verify the public webhook base URL, HTTPS scheme, host, and forwarded headers before changing the workflow.

Most technical advice skips the exact screen, URL, or setting that proves the fix. FloxoLab is building the opposite: four inspectable SEO and automation notes a week. Follow if practical evidence is your kind of feed.

The fastest n8n webhook debugging order is URL mode, HTTP method, workflow state, proxy, then sender. Starting inside the workflow wastes time when the request never reached n8n.

An n8n webhook can return 404 even when the workflow is fine. Check the URL before touching the nodes: `/webhook-test/` only listens during a test, while `/webhook/` is the production endpoint. Same workflow, different door.

Half of technical SEO is deleting things. A robots rule blocking the wrong folder, a canonical pointing nowhere, a redirect chain three hops deep. Nobody sells subtraction, so nobody does it. That gap is the opportunity. #SEO

High impressions, low CTR is a title problem, not a ranking problem. Sort GSC by impressions, find pages people see but skip. The rank is fine, the result is boring. Rewrite the title to the intent. Cheapest upgrade there is. #SearchConsole

You can win the answer box with 45 words. Google lifts a short answer above the results for question queries, and its AI Overview quotes the same tight paragraph. Put an H2 with the exact question, answer in 40 to 55 words underneath, detail below for humans. Format is half the win. #SEO

"Crawled, currently not indexed" is not a bug, it is a verdict. Google read the page and decided it was not worth keeping. Resubmitting does nothing. Earn it: real content, a reason to exist, internal links pointing in. #SearchConsole

The cheapest ranking lever nobody pulls: internal links. You publish a strong page, then orphan it, and Google reads the silence as low priority. Three links from pages that already rank and it moves. Ten minutes, real effect. #SEO

You probably have backlinks pointing at pages that no longer exist. Someone linked to a URL you deleted, now it hits a 404 and the authority is wasted. 301 the dead URL to the closest live page, reclaim the link with one redirect. #SEO

More pages is not more SEO. Ten thin pages on one topic is one topic split ten ways, none strong enough to win. Concentrate the relevance into one page that can own it. Subtraction beats publishing more often than people admit. #SEO

Everyone rewrites the title and ignores the meta description. So Google writes it for you off a random line. Write the pitch under every result or let a robot pick it. Two minutes a page, free CTR. #SEO

Two of your pages ranking for the same query are not helping each other, they are splitting the vote. Google picks one, flip-flops, both underperform. Sort GSC by query, find the pair, merge, redirect the weaker. One strong page wins. #SEO

Updating an old page usually beats writing a new one. Pages decay. Compare last 3 months in Search Console to the previous 3, find what dropped, refresh those first. Age and links already earned. New content starts at zero. #SEO

Ranking is not the finish line, it is the door. Position 1 with a title that reads like a filing cabinet still loses the click. Traffic is won after you rank, in the title and description, not at the ranking itself.

Your fastest SEO wins are stuck on page 2, where nobody clicks. Filter Search Console to positions 11 to 20. Google already ranks those pages, they are one push from page one where the clicks are. Better title, one section, three internal links, they cross over. Stop chasing rank 50.

The demo always works. The 200th run is where automation lives or dies. A rate limit, a timeout, one weird input. A flow that passed once and broke at scale was never finished, it was auditioned. Build for run 200, not run 1.

Automation does not fix a broken process, it runs it faster. If the steps are a mess by hand, automating just makes a mess at scale and on schedule. Clean the process first. Then automate the clean version.

Duplicate title tags are the most common issue I find and the easiest to ignore. Every page claiming the same title tells Google they are interchangeable. They are not. One title per page, written for the query, not the brand.

Webhooks beat polling when you can use them. Polling every minute burns runs and still lags. A webhook fires the instant something happens and costs nothing while idle. Check if the service offers one before you schedule a loop.

You can rank well and still get no clicks. Position 3 with a dull title loses to position 5 with a sharp one. Open Search Console, sort by impressions with low CTR, rewrite those titles first. Cheapest traffic you will find.

The fixes a dashboard-only agency never touches are template-level: duplicate titles, wrong canonicals, parameter URLs, a bloated sitemap. Boring, and they move more than another blog post. Code for each: floxolab.com/seo/guides/t...

The Technical SEO Fixes Agencies Won't Touch | FloxoLab

Duplicate titles, wrong canonicals, parameter URLs, broken internal links, and a bloated sitemap. The template-level fixes a dashboard-only agency cannot make, with the code for each.

floxolab.com