- Most waste is not fraud. It is work done in the wrong order or aimed at the wrong queries.
- Publishing onto an unfixed site is the most expensive habit in SEO.
- Traffic that cannot convert costs the same to produce as traffic that can.
- Review the spend against enquiries quarterly, not annually.
Five patterns account for most of the money wasted on SEO by businesses in Madrid. None of them require a dishonest supplier - they are ordinary work, done in an order that guarantees a poor return. All five are visible from the outside if you know where to look, and each has a check you can run yourself in under an hour.
The five leaks at a glance
| Leak | What it looks like in the report | The check that exposes it |
|---|---|---|
| Content onto a broken site | Steady publishing, flat rankings | Search Console coverage: how many published URLs are indexed |
| Traffic that cannot buy | Sessions rising, enquiries flat | Share of clicks from queries a buyer would type |
| Bought links | A rising domain metric | Referring domains added last quarter, opened one by one |
| Pages competing | Page count rising, positions stuck | Site search for your main query: how many of your own pages appear |
| Reporting instead of doing | A polished monthly dashboard | The change log: URLs edited, with dates |
1. Content published onto a broken site
Pages published before indexing, speed and duplication are fixed underperform, and you cannot tell whether the content or the site was the problem. Two weeks of technical work first changes the return on everything that follows.
The tell is simple: compare the number of pages published in the last year against the number Search Console reports as indexed and receiving impressions. A gap of more than a fifth means you are paying writers to produce pages the crawler never properly took in.
Flat results after six months of publishing are almost always read as "the content is not good enough", and the answer is more content. If the site could not serve the first forty pages properly, it will not serve the next forty either.
2. Chasing traffic that cannot buy
Informational terms with high volume bring visitors who will never become customers. They flatter the report and cost the same to produce as the narrow commercial terms that would have brought enquiries.
Check what share of your traffic comes from queries a buyer would type. The distinction is not about volume, it is about where the searcher is standing:
| Query type | Example shape | Volume | Worth paying for |
|---|---|---|---|
| Commercial | service + city, service + price, service + provider | Low | Yes - this is the whole point |
| Comparison | X vs Y, best X for Z, alternatives to X | Medium | Yes - buyers arrive here late |
| Problem-aware | why is my X doing Y, how to fix X | Medium | Sometimes - only if your service is the fix |
| Purely informational | what is X, history of X, X statistics | High | Rarely - it fills the chart and nothing else |
A page ranking first for a term with two hundred searches a month and buying intent is worth more than a page ranking first for one with twenty thousand and none. Reports rarely present it that way, because the second number looks better.
3. Links bought instead of earned
Volume link packages produce a profile that either does nothing or has to be cleaned up later, at a cost. The money would have bought two properly researched pages, which compound instead of decaying.
You can usually tell within ten minutes. Open the referring domains added in the last quarter and read them:
A trade body, a supplier, a local paper, a client case study - all worth having in their own right. A site of unrelated articles in four languages is not.
If the only content around your link is other paid links, the value transferred is close to zero and the risk is not.
Earned links have a human on the other end. Bought ones arrive as a spreadsheet row.
“The cheapest link package and the most expensive cleanup are usually sold to the same business, eighteen months apart.”
4. Pages that compete with each other
Publishing without a structure produces several pages targeting one query, each too weak to win. The site grows, the rankings do not, and the maintenance burden increases. Consolidation typically recovers this faster than any new work.
The check takes a minute: search your own site for your main commercial query and count how many of your pages could plausibly answer it. Two is a warning. Four is the reason you are stuck.
List every page that could answer the same query. Group them by the intent behind the query, not by the words in the title.
The one with the best links and the clearest URL, not the longest or the newest.
Move what is genuinely unique into the survivor. Most of it will be repetition you can drop.
Every retired URL to its survivor, one hop, no chains. Then remove them from the sitemap and the internal links.
5. Reporting instead of doing
A retainer that produces a monthly dashboard and no deliverables is not cheap, it is total waste. Ask for the change log: which pages were edited, what shipped, what was fixed. Two consecutive months without one is the answer.
A useful report and a decorative one are easy to tell apart:
| A useful report says | A decorative report says |
|---|---|
| These six URLs were edited, on these dates | Organic visibility improved this month |
| This is what we fixed and what broke it | We continued to optimise on-page factors |
| These queries moved, and here is why | Keyword rankings are trending positively |
| This is what we are dropping and why | We will continue to build authority |
| This is what we recommend you stop paying for | Recommendations: improve internal linking |
A quarterly check that finds all five
- What shipped, as a list with URLs and dates.
- What share of organic traffic came from commercial queries.
- Enquiries from organic, against the previous quarter.
- Cost per enquiry, trending in the right direction or not.
- Any new pages competing with existing ones.
- Referring domains added, read individually rather than counted.
Five of the six are answerable from Search Console and your own CRM. If your provider cannot produce the sixth, that itself is the finding.
Frequently asked questions
How do I know if my agency is doing real work?
The change log and the site itself. If neither changed in two months, nothing was done regardless of the report.
Is a big budget wasted on a small site?
Often, yes. A twenty-page local site cannot absorb 4,000 euros a month of useful work. Spend it on ads or on the offer instead.
When should we cut spend?
When the leading indicators - impressions for target clusters, pages shipped - stay flat for two consecutive quarters with no explanation.
Is cheap SEO always wasted?
Not always, but a low fee buys few hours, and few hours spent on the wrong thing is the most common shape of this waste. Ask how many hours the fee represents and what they will be spent on before judging the price.
We paid for links a year ago. Should we disavow them now?
Almost certainly not. Without a manual action, disavowing is more likely to remove value than to recover it. Stop buying, and spend the same money on pages that earn links instead.
How long before we can judge whether the budget is working?
Two quarters for leading indicators - pages shipped, impressions for the target cluster. Three to four for enquiries. Anything judged sooner is being judged on noise.
Find out where your budget is going
An independent review of the work being done and what it has produced.
Get a second opinionRelated: SEO services.