The hardest thing about an underperforming SEO engagement is that it does not look like failure. The reports arrive on time. Work is visibly happening. Everyone is busy.
Six months in, organic enquiries are flat, and nobody can say precisely why.
The failure mode is almost never fraud. It is a provider optimising for what you can see, because in a market where quality is unverifiable to the buyer, visible output wins over invisible diagnosis. Here are the eleven patterns that indicate it is happening to you.
1. The deliverables are counted, not chosen
Twelve blog posts. Fifty directory submissions. Four hours of link building. Ten optimised pages.
A quantity is a menu item, not a plan. It tells you nothing about whether any of those units addresses the actual reason your site underperforms.
Two businesses with identical symptoms can need completely different work. One needs three months of content restructuring, the other needs four hours fixing a stray noindex left over from staging. Both will receive the same twelve-posts-a-month proposal if it was written before anyone looked.
The test: ask what would change in the plan if your biggest constraint turned out to be different. If the answer is “we’d adjust the content topics,” the plan is not responsive to diagnosis.
2. Nobody has ever recommended deleting anything
Healthy SEO work involves subtraction. Thin pages get consolidated. Duplicate pages competing for the same query get merged. Old content that no longer serves anyone gets removed or redirected.
A provider whose recommendations are exclusively additive has a business model reason for that: adding is billable and subtracting is not.
This matters more now than it used to. Content that exists in many places and can be assembled from general knowledge has lost most of its value as AI systems answer those queries directly. A site carrying forty pages of generic advice is not neutral. It is diluting the pages that could actually rank.
3. The report shows positions but not clicks
This is the single most reliable tell.
Positions can improve while traffic falls. GrowthSRC’s 2025 study of more than 200,000 keywords found average click-through rate across positions one to five declined roughly 17.92% year over year, as AI Overviews and SERP features absorbed clicks above the organic results.
So a report showing green arrows on rankings and no click data is showing you the one number that can rise while the business gets nothing.
What a real report contains: clicks per keyword cluster, impressions-to-clicks ratio against the benchmark for the position you hold, and movement on the specific queries that produce customers.
4. You have never been told a keyword is not worth chasing
Ranking fifth on a hard, high-volume term feels like progress toward first. It usually is not. It is a plateau, because the sites above have advantages that another six months of the same activity will not close.
The click curve makes this concrete. Position one takes somewhere between three and five times what position five takes, depending on whose study you read. FirstPageSage puts position one at 39.8%, Backlinko’s four-million-result analysis at 27.6%, Semrush at around 22%. They disagree on the number and agree completely on the shape.
The same budget aimed at three specific queries you could hold at position one would produce more traffic than fifth place on the flagship term.
A provider who has never suggested abandoning a keyword has never done that arithmetic in front of you.
5. Internal linking has never come up
Internal links are the only significant ranking input under your complete unilateral control. You choose the source, the destination, the anchor text and the position.
Most sites waste them entirely: “click here” anchors that describe nothing, links only from navigation templates, orphan pages nothing points at, and every link aimed at a homepage that was already the strongest page.
An afternoon with a crawler typically surfaces twenty to eighty fixable issues on a mid-sized site. No content commissioning, no budget approval, no developer.
It is free, it compounds with everything else, and it is consistently skipped because it produces no visible deliverable to put in a monthly report.
6. Directory submissions are still on the invoice
In most markets, bulk directory submission has been worthless for a decade. In thin markets it is worse than worthless.
Where the local directory landscape is largely low-quality or outright spam, which describes a great many smaller search markets, submitting to two hundred of them manufactures precisely the footprint that link spam systems are built to detect: sudden velocity, topical irrelevance, template patterns, and links from pages nothing else points to.
A provider enthusiastic about volume directory work in 2026 is either not paying attention or selling something countable.
7. The Google Business Profile is an afterthought
If you are a local business and the proposal leads with blog posts, whoever wrote it has not read the local search data.
Google shows a local pack in roughly 93% of local queries, and that pack takes about 42-44% of the clicks. Businesses inside the three-pack receive 126% more traffic and 93% more conversion-oriented actions than those in positions four to ten. Complete profiles get seven times more clicks than incomplete ones, and 56% of profiles are not optimised.
More than half your competitors have not done the free thing. Any provider skipping past that to sell content production has inverted the priority order.
8. You do not own the accounts
Search Console, Analytics, the Business Profile, the domain, the hosting.
All of it should have been created under your ownership with access granted to the provider. When it runs the other way, leaving costs you your entire measurement history, the exact data any future provider would need to evaluate what happened.
Some providers do this through carelessness. Some do it because it makes the relationship expensive to exit. Either way the fix is the same and it should happen this week.
9. Nobody has ever said “this isn’t working, here’s why”
Every engagement has months where something does not land. A content bet misses. A technical fix does not move what it was supposed to move. A competitor does something clever.
A provider who has never reported a miss is either extraordinarily lucky or is filtering what reaches you. The second is far more common, and it removes the entire feedback loop that makes the engagement improve.
The strongest signal of competence is a provider volunteering a failure before you notice it, with an explanation of what they learned. Any seo consultant in Nepal or elsewhere who does this consistently is worth keeping, almost regardless of their other results, because it means the reporting reflects reality.
10. Every month’s work looks like last month’s
Search work should change shape as it progresses. The first phase is remediation: indexing, titles, competing pages, technical debt. Once that is done, the work should move to content and authority, and later to conversion and defence.
A retainer where month eleven is indistinguishable from month two has stopped being diagnostic. Either the remediation phase never finished, which someone should be able to explain, or the engagement has settled into producing a fixed quantity of output because that is what the invoice describes.
The test: ask what the work will look like in six months and why it differs from now. A good answer describes a transition. A weak one describes the same activity continuing.
11. You have never seen a number that went down
Every engagement has months where something misses. A content bet does not land. A technical fix does not move what it was supposed to move. A competitor does something clever.
A provider who has never reported a decline is either extraordinarily lucky or filtering what reaches you, and the second is far more common. It also removes the feedback loop that makes the engagement improve. You cannot correct a course nobody admits has drifted.
Volunteering a miss before the client notices it is the single strongest signal of competence available, because it means the reporting reflects reality rather than the relationship.
What good actually looks like
For contrast, because “not this” is not a specification.
A healthy engagement has a written diagnosis that names the constraint and the evidence for it, produced in the first few weeks and revisited when it changes. It reports clicks on converting queries rather than positions on vanity ones. It includes months where the recommendation is to remove or consolidate rather than add. It shows the reasoning, so you could disagree with a decision if you wanted to. And it changes shape over time as the constraint moves.
None of that requires the provider to be a genius. It requires them to be doing diagnosis rather than production, and the difference is visible in the first conversation.
What to do if you recognised four or more
Do not necessarily fire anyone. Reset the engagement instead, with three specific asks.
Ask for the diagnosis in writing. One page: what is the single biggest constraint on this site’s organic performance, and what is the evidence. If nobody can produce this after six months of retainer, that is the finding.
Change the report. Clicks per cluster, movement on converting queries, and one paragraph explaining what was tried and what happened. Positions can stay in an appendix.
Reallocate one month to subtraction. Consolidate the competing pages, fix the internal links, and kill the thin content. It is the least billable month and frequently the most productive one.
If the provider resists all three, you have your answer. Plenty of independents (a seo expert pokhara, a specialist in your own city, whoever) will happily work this way, because diagnosis-led engagements are more interesting to run and produce results that renew themselves.
The one question worth asking any replacement, before anything else: what did you find when you looked at my site? A specific finding means they looked. A description of their process means they did not.
Disclaimer: The information provided in this article is for general informational and educational purposes only. It does not constitute professional SEO, marketing, or business advice. Search engine algorithms, click-through data, and local search behavior vary by industry and market. Readers should evaluate their specific provider relationships and performance data independently before making decisions. The mention of specific studies, tools, or providers is illustrative and does not imply endorsement. The author and publisher disclaim all liability for any financial losses, ranking changes, or other consequences arising from reliance on this content. Always base SEO decisions on your own verified data and consult qualified professionals where needed. This article does not guarantee specific traffic or ranking outcomes.
Discover articles that make you think differently—start reading our perspective-shifting resources today.
