Does SEO Actually Work? (And Why It's Not Working for You)
Does SEO actually work? Yes, with conditions. The real numbers from a B2B SaaS program, the five reasons it fails, and when not to bother at all.
Yes, SEO actually works, on its own schedule, for businesses with something worth ranking in Google and in ChatGPT answers. As for why SEO doesn’t work for a given company, the cause is structural: five reasons cover nearly every stalled program I see across the B2B SaaS accounts I run, and a few businesses should spend the money elsewhere first.
I sell SEO for a living, and the headline numbers below come from a legal SaaS program I run.
SEO works. It does not work on demand, on a deadline, or without something worth ranking.
Does SEO work, and what does it look like when it does?
SEO works when a page that deserves to be chosen is findable, and the business is patient enough to let the compounding happen. The shape is slow and then obvious.
On a legal SaaS account, an audit-first program rebuilt the priority pages around the queries that actually mapped to buying intent. Over six months, organic sessions rose 63% and organic became the largest influenced-demo channel on the account.
Two caveats on that number. It was an established domain, and the program included consolidation and technical work alongside the content, so no single change owns the result. What it demonstrates is the shape: the work went into pages built for buying queries (29 rebuilt platform pages and 24 new or refreshed bottom-of-funnel posts), and the number that moved was organic-influenced demos.
When the missing piece is ownership, a fractional SEO consultant fills it without a full-time hire.
Why is my SEO not working?
Your SEO is not working for one of five structural reasons: the page has no reason to be chosen, nobody internally owns the channel, the measurement is wrong, the timeline was never realistic, or the business problem is not a search problem. When a founder calls me after searching “SEO not working for my website,” the cause is almost never a meta description. These five are the part of SEO strategy that decides whether any of the tactics matter.
1. The page has no reason to be chosen
A page with no reason to be chosen is the biggest cause of SEO not working. If your page says what ten already-ranking pages say, neither Google nor ChatGPT has a reason to prefer yours, and no amount of on-page tuning creates that reason. Search engines and answer engines are both selection mechanisms, and selection requires a difference.
That difference is usually first-party data or a real point of view: a number only you have, a process you actually run, a tradeoff you will name. Publishing more of the same is the most common response to SEO not working and the least effective.
The AI-search version of this is starker. Across a six-account citation export, only 2.6% of the sources cited in AI answers were the brand’s own website. The rest was independent third parties and competitors. An independent dataset lands in the same place: Derivatex’s 2026 B2B SaaS AI citation study ran 40 categories through ChatGPT, 10 times each, and found that when ChatGPT recommended a tool it cited that tool’s own site 11.6% of the time and third-party sources 88.4% of the time. If your entire strategy is publishing on your own domain, you are competing for a small slice of the evidence pool.
2. Nobody internally owns it
Missing ownership is the quiet cause behind many failed SEO programs. The work was fine and nobody could approve it, implement it, or decide between competing priorities. Drafts sit in review for three weeks, technical fixes sit in a backlog for a quarter, and the agency gets blamed for recommendations that never shipped.
If that sounds familiar, the problem is the operating model, and it is covered properly in what to hand off and what to keep in-house.
3. The measurement is wrong
Wrong measurement is the most common false alarm in SEO. Rankings and raw traffic have decoupled from revenue, so a program can be genuinely working while the dashboard everyone watches looks flat or negative.
On an HR-tech SaaS account I ran until September 2026, impressions climbed by more than 650,000 in one month while clicks fell by 7,100, which is the Great Decoupling in one chart. Judged on sessions, that program looked broken: organic sessions fell 52% in the first half of 2026. Judged on demand, organic MQLs still reached 116% of goal. Before concluding SEO is not working, confirm you are reading the right metric at the right stage.
4. The timeline was never realistic
SEO compounds, which means it looks like nothing is happening right up until it does. Indexing and impressions move in weeks, rankings and traffic over months, and influenced pipeline over quarters. Calling it a failure at week six is reading a slow instrument on a fast clock. Ahrefs tracked about 1 million pages first seen in September 2023 and found that only 1.74% of newly published pages rank in the top 10 within a year, and that 72.9% of pages in Google’s top 10 are more than three years old. A brand-new page judged at week six is being judged before most pages ever rank. The realistic timeline is its own subject.
Impatience does real damage here, because the usual response is to change strategy every eight weeks, which guarantees nothing ever compounds.
5. The business problem is not a search problem
Sometimes SEO is not working because search was never the constraint. If the offer does not convert the traffic it already gets, more traffic will not fix it, and a sudden decline is a different diagnosis entirely, covered in why organic traffic drops and why rankings drop. If nobody is searching for the category, there is no demand to capture, only demand to create, and that is a different budget.
Does SEO work anymore, with AI answering everything?
Yes, though what it produces has changed. More questions get answered on the results page: SparkToro’s 2026 study found about 68% of US Google searches ended without a click from January to April 2026. The same visibility yields fewer clicks, and the value increasingly shows up as being named in an answer.
That is a measurement shift: the channel still works, and the scoreboard moved. What has genuinely stopped working is publishing commodity content at volume, because that is exactly what AI answers replace. The longer argument about the profession itself is in will AI replace SEO, and what the click loss means for your traffic is in your organic traffic is gone.
When should you not bother with SEO?
You should not bother with SEO yet if any of these five conditions is true:
- You do not have product-market fit. Search visibility for an offer the market has not accepted just gets you a faster no.
- Nobody searches your category. If the demand does not exist yet, you need demand creation, and SEO is the wrong first purchase.
- You need revenue in six weeks. Paid can buy that. SEO cannot, and anyone promising otherwise is selling. Google’s own guidance says it plainly: “No one can guarantee a #1 ranking on Google.”
- You cannot articulate what you do. If the positioning is unresolved, publishing pages at scale distributes the confusion faster and at higher volume. Fix the message first.
- Nobody will own it. See reason two. Without an internal owner, the budget buys activity.
Fix the one that applies, then come back to search. Each of these makes the money you spend on SEO buy activity, with no compounding behind it.
Frequently asked questions
Does SEO make a difference?
SEO makes a measurable difference when it targets the queries buyers use right before they talk to sales. On the legal SaaS account above, organic became the largest influenced-demo channel within six months, with 362 organic-influenced demos. On a site that publishes generic explainers, the difference is much smaller, because AI answers now absorb that content before anyone clicks.
How quickly does SEO work?
SEO shows up in indexing and impressions within weeks, in rankings and traffic over three to six months, and in influenced pipeline over quarters. Programs that change strategy every eight weeks never reach the compounding stage. The stage-by-stage timeline is in how long SEO takes to work.
Why is SEO not working now?
SEO is not working now for most sites because the content has no reason to be selected or the measurement is reading the wrong signal. Both increased in 2026: the bar for differentiation went up as AI answers absorbed commodity explanations, and the gap between impressions and clicks widened so that healthy programs can look flat. Check whether your pages add something specific, then check whether you are judging the program on rankings and sessions when it should be judged on conversions and pipeline.
Why does SEO usually not work for a website?
The most common structural reason is that the site publishes pages nobody has a reason to choose, and the second is that no one internally owns the decisions the work depends on. Technical problems and thin backlink profiles matter, but they are rarely the whole story on a site that has been trying for months. If a page would not be the best answer a reader could find, its ranking problem is a content problem.
How do you fix SEO that is not working?
Diagnose before you publish, because most SEO issues are not the ones a tool flags. Confirm the pages are indexed, confirm you are reading conversions and pipeline, then audit whether the pages carry anything a competitor could not copy. Fix the ownership gap if approvals and implementation are the bottleneck. Adding more content is the most common first move and usually the wrong one, because volume does not resolve a selection problem.
Is SEO outdated?
No, though a specific version of it is. The playbook of publishing high volumes of keyword-targeted explainers is outdated, because AI answers now absorb exactly that kind of commodity content. What is not outdated is earning selection: being the source with the data, the experience, or the point of view that a search engine or an answer engine has a reason to pick. The tactics aged. The mechanism still works.
Why do people say SEO does not work?
People say SEO does not work because it fails visibly and succeeds slowly. Why SEO fails is easier to see than why it succeeds. A program that quietly compounds over three quarters generates no stories, while one that spent six months producing nothing generates a lot of them. The failures are also real: plenty of SEO spend goes to commodity content, wrong-intent keywords, and programs nobody owns. Both things are true. The useful question is whether the conditions for SEO working are present in your business.
What changed
- November 19, 2026: Added Ahrefs data on how long new pages take to rank, Google's own guidance on ranking guarantees, and an independent AI citation study, and corrected the HR-tech example to the numbers the account reported.