We recently audited our own website. Not a client's — ours. It had been live for years, it looked fine, the content was written by people who knew the subject, and it was barely being indexed. Google had crawled most of it and decided, page by page, not to keep it.
That is a more common situation than the SEO industry likes to admit, and it is almost never fixed by keyword research. What follows are the four problems we actually found, in the order of how much damage they were doing. If your site is good and still invisible, it is probably one of these.
1. Doorway pages: 245 URLs that should never have existed
Somewhere along the way, a route had been added that generated pages automatically from a configuration file: seven services multiplied by thirty-five locations. Two hundred and forty-five URLs, each one a template with a city name swapped in. "Web development in Lagos". "Web development in Accra". And so on, through every combination.
The intent was reasonable — appear in local searches. The execution is something Google names explicitly in its spam policies: multiple pages targeting specific regions or cities that funnel users to one page. These are doorway pages, and the penalty is not limited to the generated pages. It attaches to the site.
What makes this failure mode dangerous is that it is invisible from the front end. Nothing linked to those URLs. They were not in the sitemap. Nobody browsing the site would ever have seen one. They existed only because the route existed, which means they can sit there for years generating risk that never shows up in any report you routinely read.
The fix was deletion — the route removed, and the command that generated them disabled outright, so they cannot be recreated by someone running it out of habit. Location targeting still matters, but it means a small number of genuinely distinct pages for places where you have real clients and real work to describe. Three good ones beat two hundred and forty-five permutations, and it is not close.
2. Duplicate URLs: every page competing with itself
The site exposed most of its content at two addresses. A blog post lived at both /blog/some-article and /blog/details/47. Products, services — the same pattern throughout. The second form was a legacy of an older URL scheme that had never been retired.
Both URLs returned the identical page with a 200 status. To a crawler, that is two separate documents that happen to be the same. Every signal a page earns — links, engagement, time on page — gets divided between the two addresses instead of accumulating on one. Neither version ends up strong enough to rank, and Google frequently resolves the ambiguity by indexing neither.
This is the single most common technical problem we find on sites built over several years, because it is created by the most reasonable of decisions: not breaking old links. But keeping an old URL alive and having it serve content are different things. The old URL should respond with a 301 redirect to the canonical one. That preserves the inbound link and consolidates the signal, which is the entire point.
Worth noting how partial these fixes tend to be. On our site, the blog and services routes had been corrected at some point. Software had not. Nobody had noticed, because the pages all worked perfectly from a visitor's point of view.
3. Thin content: half the pages had nothing to say
We measured every page against a simple bar: does it carry at least 300 words of substantive content? Eleven of twenty-five did not.
These were not abandoned drafts. They were listings for products, each with a title, a category, a stock image and roughly ninety words of description. Several described software that had never actually been built — placeholders added during a redesign that were never filled in and never removed.
The instinct is to leave them alone on the theory that a weak page is better than no page. That is backwards. Thin pages do not merely fail to rank; they contribute to how the whole domain is assessed. If half your site is filler, that is a fact about your site, not about those individual pages. Google's own guidance on site-wide quality is unambiguous on this point, and it is the same judgement that gets applied when a site is reviewed for advertising programmes.
So the fix is uncomfortable but simple. Either write the page properly or remove it. We removed six outright and rewrote the rest. Removing a weak page is a legitimate optimisation, and it is one most agencies will not propose, because deleting things does not look like work.
4. A sitemap that advertised the worst pages on the site
The sitemap had guards — someone had clearly thought about this. Services were included if their description exceeded 200 characters. Products were included if they had a meta title, or appeared on the homepage, or had an image.
Both tests were useless. Two hundred characters is about thirty-five words. And every thin placeholder product had a meta title, because filling in the SEO fields is exactly the sort of tidying that gets done on pages nobody has written yet. The check passed precisely the pages it existed to exclude. Blog posts had no length check at all.
A sitemap is not a list of pages that exist. It is a submission — a statement that these URLs are worth Google's attention. Submitting a ninety-word stub is actively inviting a search engine to judge you on your weakest work. We replaced all three guards with one honest word count, applied identically everywhere, and twelve URLs immediately stopped being advertised.
What to check on your own site
None of the above requires a tool subscription.
- Look at your route list. Not your navigation — your actual routes. Anything that generates URLs from a loop or a config file deserves scrutiny.
- Search for id-based URLs. Anything matching
/details/{id}or similar. Confirm it returns a 301, not a 200. - Count words on your twenty weakest pages. Not characters. Then decide, page by page, whether to write it or remove it.
- Open your sitemap and read it. Every URL in there is one you are vouching for. Would you send a prospective client to each one?
- Check the "crawled — currently not indexed" report in Search Console. That group usually has something in common, and the pattern is the diagnosis.
The uncomfortable part
Every one of these problems was created by someone trying to improve the site's SEO. The location pages were built to capture local search. The id-based URLs were kept to avoid breaking links. The placeholder products were added to look established. The sitemap guards were written specifically to exclude weak pages.
Good intentions, each one, and collectively they were the reason the site could not rank. Which is the actual lesson: search visibility is far more often lost to structural decisions made years ago than won by tactics applied this quarter. Before commissioning anyone to build you more pages, it is worth finding out what the pages you already have are doing to each other.
Common questions
How long does it take to see results from SEO?
Most businesses begin seeing noticeable improvements in organic traffic within 3-6 months of implementing a comprehensive SEO strategy. However, SEO is a long-term investment. Significant ranking improvements and traffic growth typically occur over 6-12 months, with continued growth as your site builds authority. Competitive industries may take longer, while businesses in less competitive niches might see faster results.
What is the difference between SEO and paid search advertising?
SEO (organic search) involves optimizing your website to rank naturally in search results without paying for each click. Results take time to build but provide lasting value. Paid search (PPC) involves paying for ad placement at the top of search results, providing immediate visibility but requiring ongoing investment. SEO typically delivers higher long-term ROI, while PPC offers quick results and precise targeting.
Can I do SEO myself or should I hire an expert?
Basic SEO tasks like optimizing meta tags and creating quality content can be done in-house with proper guidance. However, comprehensive SEO requires technical expertise, specialized tools, and significant time investment. Most businesses find that hiring SEO professionals delivers better results faster, as experts stay current with algorithm changes, have access to premium tools, and can implement advanced strategies that typical business owners lack time or knowledge to execute effectively.
How much should I invest in SEO?
SEO investment varies based on your industry competitiveness, current website status, and goals. Small businesses might invest $1,000-$3,000 monthly for basic SEO services, while competitive industries may require $5,000-$15,000+ monthly for comprehensive strategies. Consider SEO a long-term investment with compounding returns. Even modest SEO budgets typically deliver positive ROI within 6-12 months, with returns increasing over time as your site builds authority.
