How long does SEO take to work, and why is there no single answer?
Because SEO covers two very different jobs. One is repair: something on the site is stopping Google from showing pages it would otherwise rank, and once a technical SEO audit finds the fault and it's fixed, those pages can come back fairly quickly. The other is growth: the site's never ranked for the searches that matter, and it has to earn its way up against sites that already do. Those jobs don't run on the same clock.
Ask how long it takes to rank on Google and the stock answer online is somewhere between a few months and a year. It isn't wrong, it's just too wide to plan a budget around. Google's own starter guide puts the range wider still: some changes take effect in a few hours, others could take several months, and it suggests waiting a few weeks before judging whether a change helped. So I start with the site, because its Search Console history already says a lot about how long the climb will be. Whether that wait pays off for your business is a separate question, and I answer is SEO worth it on its own page.
Why does SEO take so long?
Because every change waits on Google. It has to recrawl the page, which Google says can take a few days to a few weeks, then re-evaluate it against pages that already rank. A new page starts with no history, and earning links and a place among established results takes months, not days. The work itself is rarely the slow part.
The waiting isn't the same everywhere. A page Google already knew and trusted, hidden by a fault, comes back as soon as it's recrawled. A page competing for a contested search has to beat whatever's there, and that's where the quarters go.
What does a site's starting position tell me?
More than anything I could ask the owner. Average position tells me how far the site has to climb: position 8 is on page one, position 36 is on page four, where almost nobody scrolls. Impressions a day tell me whether Google is showing the site at all. Then I split the queries into brand and non-brand, because a site only found by people typing its own name isn't being discovered, it's being looked up.
| Where the site starts | What it usually means | How I plan the timeline |
|---|---|---|
| Clicks fell suddenly on pages that used to rank | A fault: a stray noindex, a broken redirect, a blocked folder | Weeks after the fix, once Google recrawls the pages |
| Most target pages sit on page two, positions 11 to 20 | Close enough that page-level work can move them | A few months, judged page by page |
| Average position in the twenties or thirties | Seen, rarely clicked, not yet trusted for the terms | Quarters, planned in three-month blocks |
| Almost every click is a brand search | Found only by people who already know the name | Quarters, with new pages for buyer questions first |
| A new domain with no history | Nothing to read yet | Checkpoints instead of dates until the data exists |
Lux Jewels, a Canadian lab grown diamond store, started in the third and fourth rows at once: average position 36.0, about 240 impressions a day, and almost every click coming from people who already knew the name. That's a quarters job, and it's the kind of thing I'd rather say before a contract starts than halfway through it.
Rather have me do this? One person, whole job.
Book a call on UpworkWhat does a typical first year of SEO look like?
There's no typical site, but there's a typical order, and it follows the bands above. The table is the SEO timeline I'd sketch before seeing any data. Your own baseline moves each row earlier or later.
| Months | What I'm doing | What should show |
|---|---|---|
| 0 to 3 | Baseline, fault fixes, titles on pages already near the top, internal links to page-two pages | Pages a fault was hiding can come back; impressions start rising |
| 3 to 6 | Page-level work on positions 11 to 20, and new pages for missing buyer questions | Positions improve page by page; the first new pages get indexed |
| 6 to 12 | The twenties and thirties, and non-brand reach | Impressions well above the baseline, positions climbing, clicks following |
| 12 and on | Head terms against established sites | Movement I won't date, built on the narrower wins around them |
How I set a timeline for a new client
This is the order I work in before I'll put any timeframe in writing. It's the same order the rest of my process follows, because the timeline and the audit come from the same data.
- Step 01
Export the baseline
The Performance report by page and by query, as far back as Search Console goes, with the date range written down. Every later number gets compared with it.
- Step 02
Split brand from non-brand
If the non-brand group is nearly empty, the timeline is about building reach, not defending it.
- Step 03
Check for a fault first
A drop that lines up with a redesign, a plugin update or a host move usually has a cause I can find, and fixing it is the part that can move in weeks.
- Step 04
Sort the target pages into bands
Page-two pages go in the fast band, the twenties and thirties in the slow band, and searches with no page at all go last.
- Step 05
Set checkpoints, not a finish date
Impressions rising first, then positions, then clicks. A checkpoint that doesn't show its signal is a reason to change the plan, not to wait longer.
- Step 06
Say the long part out loud
If the honest answer is two or three quarters, I say that on the first call, before work starts.
Don't let anyone set your timeline from a sales call alone. Without the Search Console history, a date is a guess, and a guess sold as a plan is how owners pay for months of work that was never going to show results in that window.
What moves fast, and what takes quarters?
Fast and slow here are about how long Google takes to notice and re-evaluate a change, not how long the work takes me. Google says crawling can take anywhere from a few days to a few weeks, and that asking for a crawl doesn't mean the page will be included instantly, or at all.
- i.
Removing a fault that hides pages
An accidental noindex, a robots rule blocking a folder, or redirects lost in a move. Once fixed and recrawled, the pages can return within weeks, because Google already knew them.
- ii.
Getting stuck pages into the index
Pages Google has found but won't index usually need better internal links or a stronger reason to exist. Discovered, currently not indexed is the usual status, and clearing it tends to take weeks.
- iii.
Titles on pages already near the top
A top-ten page nobody clicks can often earn more clicks from a better title within a few weeks of Google picking it up.
- iv.
Pages sitting on page two
Positions 11 to 20 are where on-page work and internal links show up soonest in the ranking itself. I'd plan a few months, page by page.
- v.
New pages and non-brand reach
A page that didn't exist has no history. It needs crawling, indexing and then time to earn its position.
- vi.
Head terms against established sites
The broad, contested search is the last thing to move and the one I'm least willing to date. The honest route is winning the narrower searches around it first.
Two of my clients show the spread. Scrap a Vehicle went from about 10 clicks a day in June 2026 to about 70 in September, on a site built around location pages. Lux Jewels needed six months to move from position 36.0 to 27.3, starting on page four with almost no non-brand reach. Neither figure is a forecast for your site.
How can I get SEO results faster?
Work the fast rows first. Fix anything that's hiding pages, rewrite titles on pages already in the top ten, and give page-two pages better internal links, since those changes wait only on a recrawl. When a page changes, I ask for a recrawl through URL Inspection, which helps Google notice it but doesn't guarantee anything. What doesn't speed it up is publishing dozens of thin pages at once, which just gives Google more to crawl and less reason to keep it.
What should the first 90 days show?
Impressions before clicks, almost every time. When a site starts climbing out of page four, Google shows it for more searches well before it shows it high enough to be clicked. So I look for impressions a day rising against the baseline, then average position improving, and clicks last.
That order is plain in the Lux Jewels numbers. Over six months, daily impressions went from about 240 to about 1,200 while average position went from 36.0 to 27.3. That's real progress, and it's also still page three. I'd rather show a client that straight than dress it up.
A repair job looks different at 90 days. The pages a fault was hiding should be back in the index, with clicks heading toward where they were before the drop. If they aren't after a few weeks of recrawling, I've missed a second cause or the drop wasn't about the fault, and I go back to the data. I judge each change after 4 to 8 weeks, in impressions, positions and clicks for the pages I touched, never in tool scores.
If you'd like a timeline read from your own Search Console data rather than a stock range, that's where an audit starts. I'll tell you which band your site's in and what each checkpoint should show before any work is agreed.
Book a call on Upwork
